Buku Local Conflict and Community Development in Indonesia

Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized
40478
Indonesian Social Development Papers
Since 1998, Indonesia has been undergoing a momentous political and economic transition. The fall of the New
Order, the economic crisis (krismon), and radical decentralization have changed the political, economic and
social context. Within this new context, power relations are in flux, identities are being renegotiated, and insti-
tutions are changing. Changes in incentives, and in the role of formal and informal institutions at various levels,
have altered the ways in which individuals and groups relate to each other and the state. Understanding this new
context, and the ways in which various actors (national and international) can promote progressive social change
is important.
The Indonesian Social Development Papers series aims to further discussion on a range of issues
relating to the current social and political context in Indonesia, and to help in the generation of ideas on how
democratic and peaceful transition can be supported. The series will cover a range of issues including conflict,
development, corruption, governance, the role of the security sector, and so on. Each paper presents research on
a particular dimension of social development and offers pragmatic policy suggestions. Papers also attempt to
assess the impact of various interventions—from local and national actors, as well as international development
institutions—on preexisting contexts and processes of change.
The papers in the series are works in progress. The emphasis is on generating discussion amongst different
stakeholders—including government, civil society, and international institutions—rather than offering absolute
conclusions. It is hoped that they will stimulate further discussions of the questions they seek to answer, the
hypotheses they test, and the recommendations they prescribe.
Luthfi Ashari and Patrick Barron (series editors)
[email protected]
[email protected]
Local Conflict and Community
Development in Indonesia
Assessing the Impact of the Kecamatan

Development Program
Patrick Barron
Rachael Diprose
Michael Woolcock
July 2006
Indonesian Social Development Paper No. 10

Papers in the Indonesian Social Development series are not formal publications of the World Bank.
They are published informally and circulated to encourage discussion and comment between those
interested in Indonesian development issues. The findings, interpretations, judgments, and conclu-
sions expressed in the paper are those of the authors and should not be attributed to: the World
Bank and affiliated organizations; members of the World Bank’s Board of Executive Directors or
the governments they represent; the UK Department for International Development (DFID); or
the Decentralization Support Facility in Indonesia.
The full range of publications associated with the broader study of local conflict in Indonesia (of
which this report is a product) is available online at www.conflictanddevelopment.org.
Email addresses for correspondence:
[email protected]
[email protected]
[email protected]
Copies of this paper are available from:
Decentralization Support Facility

Jalan Diponegoro No. 8
Jakarta 10310
Indonesia
Tel: +62 (0)21 314 8175
Fax: +62 (0)21 3190 3090
Photographs by Poriaman Sittangang

This report is dedicated to the extraordinary team of field researchers
who conducted the qualitative components of this study.
In East Java: In NTT:

Novia Cici Anggraini Stanis Didakus
Luthfi Ashari Yan Ghewa
Saifullah Barnawi Agus Mahur
Endro Crentantoro Peter Manggut
Imron Rasyid Olin Monteiro
Mohammed Said Don dela Santo
All of them demonstrated patience and perseverance beyond the call

of duty, and in the process both generated a wealth of valuable knowl-
edge and afforded us a unique glimpse into the dynamics of social and
political change in rural Indonesia. It has been a pleasure and privi-
lege to learn from them, and to have them as our guides on this jour-
ney. We hope that the quality and usefulness of the findings (and their
implications) reported here do justice to their efforts and talents, and
those of the broader community of scholars, activists, local leaders,
public officials, journalists, development practitioners, and (most im-
portantly) everyday citizens committed to making democracy work in
Indonesia.
Local Conflict and Community Development in Indonesia
iv
One might well ask: Why are we here, in a village of no particular signifi-
cance, examining the struggle of a handful of history’s losers? For there
is little doubt on this last score. The poor of Sedeka are almost certainly,
to use Barrington Moore’s phrase, members of “a class over whom the
wave of progress is about to roll.” And the big battalions of the state, of
capitalist relations in agriculture, and of demography itself, are arrayed
against them. There is little reason to believe that they can materially
improve their prospects in the village and every reason to believe they
will, in the short run at least, lose out, as have millions of peasants before
them.
The justification for such an enterprise must lie precisely in its banal-
ity—in the fact that these circumstances are the normal context in which class
conflict has historically occurred. By examining these circumstances closely, it
may be possible to say something meaningful about normal class conscious-
ness, about everyday resistance, about commonplace class consciousness
where, as is most often the case, neither outright collective defiance nor rebel-
lion is likely or possible.
James Scott, Weapons of the Weak 1
1
Scott (1985, p. 27).
v
vi
Executive Summary
1. Introduction and Context
This study examines questions relating to the nexus between development projects and different
forms of local conflict. It does so by examining how the World Bank/Government of Indonesia’s
Kecamatan Development Project (KDP) interacts with social tensions and local conflict, and
how it affects the nature and extent of local conflict management. By local conflict we mean
disputes (violent and non-violent) that play out at the local level—that is, at the sub-district level
and below.
While KDP was not designed as a conflict reduction or management program, it provides a
particularly interesting venue for examining the relationship between development (projects) and
local conflict. First, at a cumulative cost of over a billion dollars, and having operated in over
28,000 villages (40% of the total) in Indonesia, KDP is the largest social development project in
Asia and one of the World Bank’s flagship “community-driven development” (CDD) programs.
While many claims, and counterclaims, are made regarding the efficacy of CDD projects generally,
and KDP in particular, the evidence base on which to assess any of them remains rather thin.
Second, as the “community-driven” banner implies, KDP is a decentralized project that affords
significant opportunities for discretion by local level staff and input by local communities at every
stage in the design, selection, and implementation process. Its primary objective is to help
participants secure small sub-projects (roads, bridges, water pipes) that accord with their needs,
priorities, and values. KDP attempts to realize this objective by applying the design principles of
participation, transparency, local choice and accountability to a competitive bidding process, in
so doing striving to help villagers (and especially the most marginalized groups) acquire new
civic skills and decision-making opportunities for realizing their interests and aspirations. Such
skills and opportunities are vital for mediating in constructive ways the conflicts—generated by
competitive bidding for finite resources, and the issuance of challenges to elite power—KDP
itself inevitably produces. These skills may also help in the management of other conflicts that
are not related to the project. Does KDP, in fact, achieve these goals of heightening conflict
mediation capacities? How does it compare with other development projects and the conflicts
they inevitably generate?
Third, KDP operates in Indonesia, a country in the midst of an ongoing and uneven democratic
transition, a process that has, at times, been accompanied by violence. In addition to outbreaks
of large-scale and violent communal conflict in a number of locations, and secessionist conflict in
two provinces, widespread (and often violent) local conflict has occurred across the country.
What are the strengths and limitations of projects like KDP in an unstable social and political
environment, where identities, rules, and group relations are being reconfigured, where long-
standing grievances now have the space to surface, and where access to power is being
renegotiated? Can outside interventions such as KDP support progressive social change in this
type of dynamic environment? If so, how?
vii
2. Research Questions
There are two primary questions we seek to answer in this paper:

• Does KDP generate fewer conflicts, or less serious conflicts, than other development
projects?
• Does participation in KDP help villagers find more constructive solutions to local
level conflict in general, and, if so, does it help resolve certain types of local conflict
more effectively than others?
These questions give rise to a related set of secondary questions:

• Through what mechanisms are any such positive outcomes achieved, and potentially
negative outcomes avoided or minimized? That is, how exactly do these outcomes
materialize?
• What elements of context are important in determining the effects of KDP? What factors—
either internal to the program or in the local environment in which it operates—influence
the extent to which and the ways in which the program has an impact?
3. Methods
Research Tools
Assessing the efficacy of social development projects is difficult because a defining feature of
many such projects is the non-standardized ways in which they seek to adapt to idiosyncratic
local circumstances and, in the process, generate outcomes (such as enhanced ‘participation’
and ‘inclusion’) that do not have an established or clear metric. For this reason, our strategy has
followed from a canonical, though too often ignored, research principle that the nature of the
problem should determine the choice of method(s) used, not vice versa. A methodological
strategy was developed that employs a number of different data sources and approaches.
A team of twelve researchers and supervisors conducted nine months of qualitative fieldwork in
41 villages. They developed 68 case studies of conflict pathways, which explored the evolution
of specific conflicts, some of which were violent, others not. The cases covered a wide range of
disputes including land and natural resource conflicts (which range from large ethnic conflicts to
private conflicts over inheritance), cases of vigilante justice (against thieves, witchdoctors, etc.),
gang fights, political disputes (e.g., over local elections and administrative boundaries), conflict
over access to and the management of development resources, and domestic and sexual violence.
The researchers also collected rich ethnographic material on 14 topics—ranging from how local
governments function, to local socio-economic conditions, to the role of traditional and religious
leaders, and so on—to allow for cross-village comparison. In all, over 800 interviews and
100 focus group discussions were conducted.
A key informant survey was conducted in the research villages to gather comparable responses
to perception questions relating to KDP, its effect on conflict, and processes of social change.
The full survey was conducted in areas that had had KDP; a shorter version was implemented in
control sites (shorter, since we obviously could not ask questions about the impact of KDP in
areas where KDP was not implemented). In order to assess patterns and forms of conflict,
viii
Executive Summary
and variations between areas, a dataset of conflicts as reported in local newspapers in the
research areas, and surrounding districts, was constructed (the KDP & Community Conflict
Negotiation dataset). Two other ‘larger-N’ quantitative surveys were analyzed as part of
the study: the Government’s Potensi Desa (PODES) survey, which provides information
on conflict for all 69,000+ villages in Indonesia; and the World Bank’s Governance and
Decentralization Survey (GDS). Together, these data sources provide the basis for a
comparative framework; utilizing the different data sources can help us control for
differences in conflict outcomes, in conflict mechanisms (e.g., conflict resolution attempts,
common escalation patterns), and in contexts. At different points the paper uses different
units for comparative analysis: sub-district, village and conflict case.
Site Selection
The qualitative and key informant survey research was conducted in 41 villages in sixteen sub-
districts in four districts in two Indonesian provinces: East Java and Nusa Tenggara Timur (NTT).
The provinces with the highest levels of conflict in Indonesia were excluded based on the
assumption that development projects are most likely to have a measurable effect (positive or
negative) at the level at which they are operational. Further, in areas of high-conflict, where
levels of violence are significantly affected by external actors and exogenous factors (e.g.,
military action), it would be much harder to separate out the potential impacts of a local level
report from other causal variables in the research sites. Both provinces are thus ‘medium-level’
conflict sites, with significant levels of local conflict. The provinces vary considerably in terms
of population size and density (high in East Java, low in NTT), ethnic homogeneity (homogenous
in East Java, heterogeneous in NTT), dominant religious group (Muslim and Catholic,
respectively), and provincial development (East Java is relatively rich, NTT extremely poor).
The rationale for selecting diverse provinces was that if we find similar patterns in very different
contexts, it is more likely that these findings will hold true across other locations.
Within each province, two districts were selected based on variations in local ‘capacity’; i.e.,
the ability of communities and/or the state to collectively solve or manage conflicts when they
arise. In ‘high capacity’ areas, emergent problems are usually handled early and effectively (by
formal actors, informal actors, or a combination of both) so that they do not escalate and/or
become violent. In ‘low capacity’ areas, conflicts tend to emerge more easily and escalate and/
or become violent. We chose high and low capacity districts in order to see how KDP works
and interacts with conflicts in a range of environments: our ‘high capacity’ districts were Ponorogo
(East Java) and Sikka (NTT); our ‘low capacity’ districts were Pamekasan (East Java) and
Manggarai (NTT). Districts were chosen based on interviews with a range of stakeholders at
the national and provincial levels. The newspaper mapping was conducted in the same four
districts plus surrounding districts (for a total of seven in each province) in order to determine
variation in conflict levels at the district level, and to see how representative our districts were in
comparison with their neighbors.
Fieldwork was conducted in a total of sixteen sub-districts. Within each district, four
sub-districts (kecamatan) were selected. Three were ‘treatment’ sites (i.e., they had received
KDP), and one was a ‘control’ site (i.e., it had not received KDP). For the first phase of the
study we selected locations that had had KDP for at least three years, and matched them—
ix
using propensity scoring to match on ‘observable’ variables and qualitative verification to

account for ‘unobservable’ variables (such as political connections)—with demographically
and socio-economically ‘identical’ non-KDP control locations. We later expanded the
sample to include a larger range of KDP locations, which had received KDP for varying
lengths of time. Villages were selected based on the location of “interesting” conflict
cases identified in the qualitative research, based on detailed criteria aimed at making it
easier to control for non-program effects.
For the qualitative fieldwork, sampling of informants was driven by two aims. First, we wanted
to talk with a wide cross-section of the population within each district, sub-district and village.
Special emphasis was placed on talking to non-authority figures, and women. Second, the
researchers used “snowball sampling” in order to find respondents who would be able to provide
insights into the cases being followed and, more generally, issues in the village. Focus group
discussions were conducted with particular populations groups: poor women; educated/elite
women; young men; and poor men. For the key informant survey, eight respondents were
interviewed within each village, with an additional three at the kecamatan level. Informants
were selected based on their knowledge of KDP, with a balance between elite and non-elite
respondents.
4. The Impact of KDP on Local Conflict and Conflict Management
Development projects and conflict inevitably go hand in hand, because development is an

inherently political and contested process. KDP, or other projects, may have negative impacts
on local conflict and conflict management capacity, in ways that are direct and/or indirect. The
introduction of new resources into poor areas, for example, can lead to inter-group tensions.
Programs like KDP, which aim to empower marginalized groups, also introduce new rules and
norms about decision-making procedures, and, in doing so, impact on local power balances and
structures. Resistance from elites to such changes is another basis for conflict.
Conversely, programs like KDP may have positive impacts on local conflict and conflict
management; these may also be direct or indirect. Direct impacts relate to the introduction of
facilitators and forums. Are these people and spaces used for managing conflict that is unrelated
to the program? KDP might influence local conflicts indirectly through changing the underlying conflict
environment, i.e., the structures and norms that make conflicts more or less likely to arise and/
or escalate. It may do so through three mechanisms. First, the introduction of collective
decision-making processes, that include involvement from different groups, may change inter-
group relations. Second, KDP encourages participation from marginalized groups and collective
decision-making. This may lead to behavioral changes and, in doing so, may reshape the
relationship between citizens and the state, and between ordinary villagers and elites. Third,
KDP may change norms, attitudes, and expectations regarding how disputes should be resolved.
KDP and Other Development Projects as Sources of Conflict

By introducing new resources and services into poor communities, development programs
inevitably shape local conflict dynamics, not only in areas that have experienced high levels of
violent conflict but elsewhere, too. Our evidence suggests that KDP and other development
x
Executive Summary
projects frequently trigger conflict, or interact with existing disputes, which can potentially
lead to conflict escalation. Importantly, however, KDP-related conflicts are far less likely
to escalate and/or turn violent than those relating to other programs.
The research found three forms of development-related disputes. First, KDP introduces competition
within and between villages over which proposals should be funded; this can and does lead to
tensions, in particular when groups feel that the decision-making process was not transparent or
fair. However, the research found that over time groups tend to accept the validity of the competition
process and, as a result, the outcomes it generates. Only where the program does not function as
intended (e.g., where one group has captured the process) do larger problems emerge. The second
form of conflict stems from these and other ‘program malfunctions’, which can be either problems
of omission or commission. The first is a result of poor socialization (public information aimed at
explaining the program) or implementation; the second where there has been deliberate and active
malfeasance from program staff or local elites, e.g., in cases of corruption. The latter is more
serious than the former, with cases of corruption providing a basis for larger community unrest.
A third form, interaction conflict, occurs when development projects (KDP or others) interact
with pre-existing local tensions, power structures or conflicts, triggering conflict escalation and,
in some cases, violence. Projects operate in contexts in which power relations are constantly
being negotiated; as such, development projects constitute a vital resource that can be utilized in
this process. In certain cases, interaction conflicts involve actors using the project for patronage
purposes, raising tensions between competing local elites. In others, elites attempt to capture
the project for self-enrichment. Other cases concern the resistance of elites to the norms of
widespread access to decision-making, transparency, and accountability that KDP brings. Where
resistance is greater, there is more potential for conflict.
Despite the numerous ways in which projects can trigger conflict, we found that KDP-related
conflicts almost never become violent. There was only one minor violent dispute relating to
KDP between 2001 and 2003 in the research areas; in contrast, there were 36 violent disputes
related to other government development programs and provision of government services. We
argue that the difference is largely because (a) KDP projects, by virtue of emerging from a
process whereby communities define their needs, are less likely to clash with local priorities, and
hence conflicts are less likely to emerge; and (b) KDP has a battery of in-built mechanisms
(people and procedures) that allow tensions to be addressed as and when they arise. Resolution
success rates of KDP-related problems are very high. This is in marked contrast to many other
development projects, which do not have such mechanisms; as a result, conflicts relating to
other governmental and non-governmental projects are far more likely to turn violent.
Direct Impacts of KDP on Local Conflict and Conflict Management

While KDP forums, facilitators and complaints mechanisms are used frequently and effectively
to deal with conflict related to the program, we find little evidence that KDP per se has a positive
impact on conflict at an aggregate level or a direct positive impact on non-project-related conflict
at the local level.
In both East Java and NTT, the direct impacts of KDP on conflict management are minimal
in the first three years of the program. KDP forums and facilitators are rarely used for
xi
addressing conflicts unrelated to the program; where they are used, it tends to be in an ad
hoc manner. In none of our research locations had KDP been institutionalized as a regular
(and/or more general) conflict resolution device.
There are a number of reasons why KDP does not frequently get used for non-program problem
solving. First, other institutions exist at the village level to deal with conflicts that are not related
to the program. In most cases, KDP is thus not seen as the appropriate mechanism through
which most forms of non-KDP-related conflict should be addressed, and facilitators, or other
program staff, are not seen as having jurisdiction over these issues. Where conflict management
institutions do not exist (for example, in lower conflict management capacity areas, and for
certain forms of conflict such as that relating to other development projects), KDP forums and
facilitators are sometimes used. This suggests that KDP can act as a complement to existing
conflict mediation mechanisms. It may be that in the parts of Indonesia with the highest levels of
violent conflict, where many local institutions have collapsed, KDP can play a role in resolving
certain kinds of disputes, such as those relating to administrative procedures and resource
allocation.
Second, in some cases facilitators are not perceived as having the personal legitimacy to handle
disputes. In particular, where they are seen to have been corrupt, or to have unfairly favored
one group over another, it is unlikely that they will be called on to mediate contentious issues.
For program facilitators to be effective, they need to be seen to be honest, independent, and
neutral. Indeed, the perceived and actual (political) neutrality of most KDP facilitators is a
defining virtue of the program in the eyes of participants.
Third, in many cases facilitators are unwilling to address conflicts and problems brought to them
that are not related to the program. In large part, this was because they are (not unreasonably)
risk averse. Many facilitators are conscious of the sanctions that might follow—for them and/or
for the communities in which they work—if they deviate from the PTO (Operational Technical
Guidelines). Our findings suggest that more flexibility needs to be built into these guidelines to
allow for “altruistic deviance” by facilitators.
Fourth, in some cases KDP facilitators are not called on to mediate disputes because of gaps in
their capacity. These gaps differ by the position of the facilitator. Sub-district facilitators (FKs)
tend to have the technical skills needed to help mediate issues, but often they do not have adequate
time or the local knowledge necessary to understand the positions of the disputants and to win
their trust. Village facilitators (FDs), on the other hand, tend to have time and local expertise/
legitimacy, but they often do not have the technical training/education necessary.
As it stands, KDP is thus not an effective mechanism for working directly on non-program
conflict. In some ways, this is a good thing—it allows it to remain a politically neutral space
wherein communities can work out their needs and priorities. Yet, at the same time, there is
scope for modifying the program to allow it to more effectively manage local conflict. On those
(relatively infrequent) occasions when non-project conflicts are addressed through the program,
they tend to be resolved successfully. This shows that there is the potential for KDP to
play a larger mediation role, in particular for development-related disputes. Improved
xii
Executive Summary
training for facilitators (in particular at the village level), and increasing the discretion of
facilitators implementing the program (in particular at the sub-district level), would improve
this aspect of the program. At the same time, it may be necessary to have other
complementary programs in place to directly address conflict.
Indirect Impacts of KDP on Local Conflict and Conflict Management

Whereas the direct impacts of KDP on conflict management are small, the program has notable
(and positive) indirect impacts on the local institutional environment in the areas in which it
operates. KDP is helping improve inter-group and state-society relations, and this is helping
making areas more immune to violent conflict, for several reasons.
First, the research found that across a range of different identity cleavages, KDP had helped
contribute to improvements in inter-group relations. Ethnic, religious and class relations in NTT
have improved since KDP was introduced, and these changes are greater in treatment than
control areas. There is also some evidence that relationships between those affiliated with
competing martial arts (silat) groups have improved in East Java. Further, improvements in the
quality of group relations grow larger over time. Villages that have had KDP for four years
show, in general, greater improvements than those that have had the program for shorter
periods. KDP provides a space for different groups to come together to collectively discuss
their needs and priorities, an opportunity that is rarely afforded them elsewhere (especially at the
inter-village level). KDP also facilitates group interaction by improving transportation networks.
Second, KDP also appears to contribute positively to the reconfiguration of citizen-state

relations at the local level. The evidence shows that KDP is successfully helping to democratize
village life. Many different social and political factors at numerous levels have contributed to
democratization in Indonesia in the recent years, but our comparative analysis of control and
treatment sites suggests that KDP is helping to drive and consolidate this process. Marginalized
groups (and, in particular, women) are far more likely to take part in KDP meetings than in other
village government meetings. Moreover, increased participation in KDP appears to be spilling
over into other domains of village life: fifty percent more villagers reported that more marginalized
groups were coming to village meetings in KDP areas than in the control sites, and three-
quarters of all villagers in those areas that had received KDP for the longest time (four years)
reported that more groups came to village meetings than in the past. Decision-making in village
meetings has also become more democratic, and this effect is greater in KDP areas than in the
control sites.
Third, there is mixed evidence that KDP has helped improve problem solving and conflict
resolution. The survey results show similar perceived improvements in control and treatment
sites; the qualitative fieldwork, however, shows clear linkages between the program and such
normative changes, with KDP (when it functions well) creating a positive precedent, in the
process helping to stimulate demand for changes in the ways in which local decision-making and
conflict resolution operate.
At present, KDP has been in place for too short a time to establish unambiguous empirical
links between its positive impacts on inter-group and state-society relations, and the
xiii
abilities of communities and the state to effectively manage conflict. Yet theory, examples
from elsewhere, and limited evidence from our research areas, points towards these factors
collectively helping to improve local conflict management capacity. The research shows
that KDP alone does not create these profound changes. Rather, when it works well, and
in environments favorable to change, KDP interacts with existing processes of social and
political transformation, acting as a catalyst that legitimizes processes already underway.
KDP’s impact, more generally, is highly contingent on implementation effectiveness (i.e.,
the extent to which KDP functions as intended) and local context (in particular, the
prevailing capacity in both the wider district and the more immediate social structures to
govern everyday village life). KDP can minimally compensate for poor local governance,
but is most effective when it complements broader reform initiatives in well-governed
districts and villages.
Pre-existing Local Capacity to Manage Conflict, Program Functionality and KDP

The strength and direction of each of the different forms of (direct and indirect) impact are
contingent upon the extent to which the program functions well, and the contexts in which it
works. The capacity existent in the local environment is crucial for determining overall project
impact, for several reasons. First, in low capacity environments, where KDP is poorly
implemented (as a result of, say, inadequate socialization of participants or weak enforcement of
program rules), KDP can exacerbate local conflict by providing a new resource over which
elites (and subordinate villagers) compete. A well-implemented program in a low capacity
environment can, however, produce positive outcomes.
Second, ‘capacity’ operates at multiple levels, and can work in multiple (sometimes opposing)
directions, depending on the form of impact being assessed. The use and success of KDP
mechanisms for addressing conflicts stemming from KDP, for example, varies considerably
depending on specific interactions between implementation effectiveness and ‘district’ and ‘village’
capacity, with district capacity often playing a primary role. Where conflict stems from KDP
malfunction, lower district capacity increases the use of KDP forums and facilitators, while lower
village capacity decreases it. Conversely, local capacity has little effect on the use of KDP
forums for in-built issues.
Third, whether KDP is working well or not is more important than the context in which it operates
in determining the level and direction of impact. Where program functionality is poor, hardly any
positive spillover effects are observed, with only marginal positive impacts on group relations
and in decision-making. In both low and high capacity areas, getting the program working
properly greatly enhances the positive impacts the program can have.
Fourth, where the program does not work as intended, positive effects are not only limited, but
there is also a chance of the program having negative impacts. Given that program functionality
is more likely to be poor in low capacity areas, attention and resources should be directed
towards improving KDP performance in areas with low capacity.
Fifth, the impacts of KDP are primarily indirect. KDP forums and facilitators are rarely
called upon to address non-KDP problems; KDP itself generates the positive impacts it
xiv
Executive Summary
does largely indirectly. Local capacity (at the village or district level) seems to have little
effect on whether or not KDP forums are used to address in-built conflict, and little effect
on the likelihood that KDP processes will successfully resolve either in-built or malfunction
conflicts when they are used. However, local capacity does appear to have an impact on
the likelihood of KDP processes (forums and/or facilitators) being used to address KDP
malfunction conflicts.
Finally, KDP does not displace existing forums for local dispute resolution; indeed, in high capacity
environments it can serve as a valuable complement, strengthening already well-functioning
institutions, while in low capacity environments it can provide a positive alternative to (or substitute
for) absent, captured, or dysfunctional forums. The marginal impacts of a well-functioning KDP
are higher in low capacity areas (because KDP forums need to take on a wider range of tasks),
though a minimal level of capacity is needed to provide a basic foundation on which to build. On
the other hand, indirect impacts are greater in high capacity areas, where KDP can facilitate, and
act as a catalyst to, ongoing processes of political and social change.
5. Implications
These results, and the mechanisms through which they materialize, suggest a number of broad
implications and specific recommendations pertinent to the design of current and future
development projects, including subsequent iterations of KDP. These relate to: (a) how we
understand the complex relationship between development (projects) and conflict; (b) the negative
impact of development projects on local conflict; (c) the positive roles that projects can play in
directly addressing conflict; (d) the ways in which projects can indirectly impact on conflict and
conflict management capacity; and (e) how we think about, and go about undertaking, evaluations
of such projects.
The Development-Conflict Nexus
Development and conflict go hand in hand. By virtue of introducing new resources into poor
communities, development programs inevitably shape local conflict dynamics, not only in areas
which have experienced high levels of violent conflict but elsewhere, too. Competition over
these resources can lead either directly to conflict, or can interact with existing tensions, thereby
causing them to escalate. Programs such as KDP that aim to reconfigure both inter-group and
state-community relations are especially likely to influence local power relations and hence conflict
dynamics; the challenge is to ensure that these conflicts are constructively addressed so that they
do not become violent but, rather, become part of a force for progressive social change.
Acknowledging the intrinsic linkages between development and conflict results in a number of implications
for how we conceive of the development process and how we prepare development projects.
(a) Conceptualizing conflict and development: the political dimensions of development

Conflict is not only something that occurs in ‘conflict-ridden’ places. It is an inevitable
feature of social life the world over, varying in intensity from fully blown international
and civil war to more local tensions. Processes of contention take place at multiple levels,
xv
from national politics down to the village; this is true even in geographic areas that are not
seen as being particularly prone to conflict. Development projects and policies operate in
such arenas, where power relations are constantly being (re)negotiated. They are a political
resource: development activities (whether they involve building schools or infrastructure,
providing micro-credit, supporting the formation of women’s groups, etc.) constitute
external injections of resources and rules systems that legitimize particular actions and
discourses and thus strengthen particular individuals, groups or ideas at the expense of
others. This is true for all development interventions, including those not specifically
targeted at influencing local forms of social or political organization or conflict. They are
thus likely to impact—positively or negatively—on conflict dynamics. This suggests a
number of basic implications. First, the political bases of the development process
generally, and development interventions in particular, must be explicitly acknowledged.
Second, potential and actual conflict dynamics must be taken into account when planning
and implementing projects; all development projects should recognize the ways in which
they overtly shape local dynamics of contention and, as a result, the ways in which they
might provoke or limit violent conflict. Third, the distributional consequences of
interventions—who benefits, who does not, and how this maps onto existing power
dynamics and cleavages—must be assessed, and accessible mechanisms provided for
addressing the tensions to which they give rise.
(b) Understanding local context: prioritizing and investing in local knowledge

Project interventions cannot be treated in isolation from the contexts (political, cultural,
institutional, and economic) in which they operate. Projects have the impacts they do because
of their interaction with phenomena external to the project. As such, what works in one context
will not necessarily work in another. It is partly for this reason that KDP has varying impacts—
at times positive, at other times less so—in different areas. It is thus necessary for development
organizations to understand in detail the contexts in which projects will operate, and to factor
this knowledge into project design, implementation and evaluation. In organizations like the
World Bank, there are important staffing and project preparation implications. More staff should
be based in country offices and, when there, they should spend more of their time in the villages
in which projects operate than in national capitals. Time also needs to be given over to
understanding the contexts in which projects will operate; this requires investing in studies and
assessments up-front.
Ensuring Projects ‘Do No Harm’
Development programs inevitably shape local conflict dynamics, not only in areas that have
experienced high levels of violent conflict but elsewhere, too. Our research found numerous
examples of cases where projects had negative impacts. In our research areas, 36 violent conflicts
related to development programming (all but one of them from non-KDP projects) took place
between 2001 and 2003. Comparative case analysis points to a number of reasons why some
development conflicts become violent while others do not, and suggests a number of
recommendations to ensure that development interventions ‘do no harm’ (or at least
minimize their potentially harmful impacts).
xvi
Executive Summary
(a) Involve communities at every stage

One of the reasons that KDP was less likely to trigger conflicts than other development
projects in our research areas was that it involves communities in a way that other projects
do not. In areas where KDP is working well, communities put forward proposals, are
involved in deciding which should go forward from the village, which should then be
funded and how the projects should be implemented. They are also involved in
accountability meetings throughout to ensure that work is proceeding as intended. We
found that this constant community involvement helped to limit tensions relating to the
program; individuals and communities were less likely to protest outcomes, because they
felt they had a genuine stake in deciding how these outcomes and decisions were reached.
(b) Ensure programs work as intended: socialization, monitoring and follow-up

All forms of development-related conflict are more likely where program processes do not work
as intended. Where program malfunctions occur, the risk of fresh conflict, or of the program
interacting with existing tensions, increases. Where people understand decision-making processes,
they are more likely to accept outcomes, even when these are contentious. Program implementers
have three main tools for ensuring that development programs work as intended. First,
socialization, the process of sharing information about the project and how it is meant to work,
must be comprehensive and ongoing. Where communities understand program rules and
objectives, they are less likely to dispute program outcomes and are more likely to hold local
elites, who may try to co-opt or bend program rules, to account. Second, monitoring the program
is extremely important. This involves not only monitoring program outcomes, but also, more
importantly, processes. Third, following-up on cases of malfeasance, such as corruption, can
send out strong signals that make it less likely that others will try to break program rules.
(c) Enhance internal complaints mechanisms and avenues of redress

Many conflicts triggered by development programs in general, not just KDP, are the result of
program malfunctions and frustrations with processes, problems that are compounded when there
are no or weak avenues for redress. Where there are clear and accessible mechanisms within
development programs to deal with these concerns early and often, the escalation of problems is
much less likely. Frustrations are also a product of the failure of putative feedback mechanisms to
report back on actions taken. All development programs and projects need clear and effective
feedback mechanisms for handling complaints if they are not to provoke conflict.
(d) Promote greater understanding of the project cycle, focusing on the points within it—
competition for funding, opportunities for ‘elite capture’—at which conflict is most likely
In programs which have a three or four year cycle and yearly phases of competition for and
allocation of funding (such as KDP), intra-village competition for prioritizing funding proposals
is most likely to stimulate tensions as the socialization process begins; this decreases in later
years as villagers have a clearer understanding of competitive processes, and as the aggregate
number of projects being put forward increases over time. Inter-village competition for allocation
of funds to villages, however, is most likely to increase in the last two years of the program,
particularly when certain villages miss out on funding repeatedly as the aggregate number
of proposals being funded increases. In other cases, some tensions arise when particular
villages have been sanctioned for previous poor performance. Given that KDP overtly
xvii
seeks to establish more inclusive forms of decision-making, it should be expected (and it

was clear from the evidence in this study) that certain entrenched elites will resist such
efforts, thereby heightening the likelihood of conflict.
(e) Where appropriate, include both formal and informal leaders in development programs
Problems were best resolved when redress involved program staff, state representatives,
and informal leaders. State and informal leaders need to be embedded in the process. In
conflict situations, actors involved should be included in ways appropriate to local customs
(insofar as that is possible).
Direct Interventions
KDP has little direct effect on aggregate levels of violence in the research areas. It is clear that
in areas that have experienced significant outbursts of violent conflict, programs like KDP must
be complemented by other initiatives aimed at preventing or managing conflict. At the same
time, there is scope to extend the role of KDP to address particular sorts of conflict, most
notably other development-related disputes.
(a) Design complementary programs/strategies to constructively address the drivers of conflict

Increasingly Community-Driven Development (CDD) projects, and other bottom-up approaches,
have been viewed as having the potential to address local conflicts and to increase local conflict
management capacity. The evidence from our study suggests that a more measured conclusion
is necessary. At times project forums and facilitators are used for conflict management; but,
more often, they are not. It is necessary to resist the temptation to believe that a single
development intervention, even (or especially) one with a strong participatory component, can
single-handedly reduce local conflict. At best, effective community development projects can
be part of a solution if they are able to function as effective complements to (and not substitutes
for) the state, other formal civic institutions, and related democratization initiatives.
(b) There are opportunities for facilitators and forums to play a larger role in some
conflicts, particularly those related to development
KDP facilitators infrequently deal with non-KDP conflicts and KDP forums are rarely used for this
purpose. This is largely because for most forms of conflict, they are not seen as the appropriate
actors/arenas, with other systems of recourse being in place. However, for conflicts related to
other development projects, and in cases related to other forms of resource distribution, there is
room for KDP to play a larger role. Most development projects do not have adequate complaints
systems or means of redress for beneficiaries who feel that processes or outcomes have been
unfair. In such cases, KDP is a suitable mechanism for directly addressing conflict.
Indirect Interventions
(a) Development programs can indirectly have a positive impact on local conflict and
conflict management capacity
Projects such as KDP can have important impacts on and interactions with processes of
social, political and economic change. Often the result of such interactions may be as
xviii
Executive Summary
important and significant as the direct impacts of the project. KDP is leading indirectly to
changes in social structures, institutions and norms that appear to be making communities
more robust to conflict in the medium- to longer-term. There are opportunities to build
upon this by thinking more strategically about how KDP (and other programs) can positively
affect these variables in conflict situations. For long-term conflict prevention and
management, equal value should be placed on the impact on these intermediate variables
as on the desired outcome variable of ‘reduced conflict’.
(b) Improving inter-group relations: the impact of KDP on social structure

Conflict is less likely to ensue, and is more likely to be addressed early when it does emerge,
where the social fabric of village life is characterized by diverse ties between different identity
(and ‘stakeholder’) groups. This hypothesis holds up at the micro-level in the context of the
Indonesian villages in our sample. On-going participation by diverse groups in civic spaces
(such as village councils, recreational gatherings, business associations, water-users groups) can
serve to nurture, expand, and reward such ties. Those designing projects—both those explicitly
aiming to address conflict, and those with other development goals—should think about ways in
which project designs create or limit incentives for inter-group participation and the development
of sustainable inter-group ties.
(c) Normative and behavioral changes

KDP has an impact on people’s participation, which is spilling over into demands to participate
in other development programs; this, in turn, helps to reform local governance more generally.
This process is occurring to a greater extent in areas where democratization processes are already
underway; in such instances, the program and other forces of social and political change reinforce
each other. Even where democratization is not occurring to the same extent, there appear to be
program effects, although these occur more slowly. Further, when well implemented, KDP can
improve not only the diversity and equity of decision-making spaces, and the civic skills of
participants, but alter the normative and cultural expectations regarding procedures for resource
allocation and conflict management. This is an important accomplishment in environments
heretofore characterized by pervasive elite capture (including corruption), norms of violent
retribution, and the systematic exclusion of marginalized groups.
Evaluation of Community Development Projects
Assessing the efficacy of complex social development projects to negotiate new procedures for
decision-making in village life requires a diverse set of methodological and analytical tools. Project
evaluations traditionally measure progress against a set number of outcome variables, specified
before the project begins. While it is important to strive for clearly defined metrics of project
success (e.g., the economic rate of return achieved by projects delivered through community
mechanisms versus those provided by external contractors), the very nature of projects such as
KDP and the key social objectives (participation, empowerment) towards which they strive
belie stringent assessment by a single research tool, or of a set of preordained outcome variables.
In this context, the orthodox (economic) criteria for what constitutes a ‘rigorous’ project
assessment need to be considerably expanded, with a focus on understanding causal process as well
as outcomes.
xix
xx
Table of Contents
EXECUTIVE SUMMARY VII

1. Introduction and Context vii
2. Research Questions viii
3. Methods viii
4. The Impact of KDP on Local Conflict and Conflict Management x
5. Implications xv
TABLE OF CONTENTS XXI
ACKNOWLEDGEMENTS XXIX
GLOSSARY XXX
CHAPTER 1. INTRODUCTION: UNDERSTANDING THE CONFLICT-

DEVELOPMENT NEXUS IN INDONESIA 1
1. Development Projects and Conflict: Part of the Problem, or Part of a Solution? 1
2. Research Questions and Paper Structure 3
3. Conflicts of Interest? 4
4. Making the Case 5
CHAPTER 2. ASSESSING KDP AND LOCAL CONFLICT: ANALYTICAL AND

METHODOLOGICAL FRAMEWORKS 7
1. The Kecamatan Development Program (KDP) and Local Conflict 7

1.1 The Kecamatan Development Program (KDP) 7
1.2 Local Conflict in Indonesia 8
2. Analytical Framework: How Might KDP Influence Local Conflict? 9

2.1 Direct Effects 9
2.2 Indirect Effects 10
2.3 The Role of Context: What Influences the Ways in Which KDP Shapes Local Conflict? 12
2.4 A Framework for Analyzing Process and Impact 12
3. Methodology and Research Locations 13

3.1 Evaluation Strategy: Data Sources Used 13
3.2 KDP and Conflict in East Java and NTT: Areas of Study 14
Provincial Selection 14
District Selection: High and Low Capacity 16
Local Level Sampling 18
xxi
CHAPTER 3. DEVELOPMENT PROJECTS AS ‘PROBLEM’: KDP AND OTHER

PROJECTS AS SOURCES OF LOCAL CONFLICT 21
1. The Prevalence and Impact of KDP-related Conflicts 21
2. Forms of KDP-related Conflict 23

2.1 Competition in KDP: A Source of Local Conflict? 24
Tensions over Competition Within and Between Villages: Changes over Time 29
2.2 When Things Go Wrong: Program Malfunction Conflict 31
Problems of Omission: Poor Socialization and Implementation 32
Problems of Commission: Intentional Deviations and Corruption-Related Disputes 36
The Importance of Socialization and Links to Conflict 38
2.3 The Link between KDP and Existing Tensions: Interaction Conflict 39
Power Battles between Elites and KDP 40
Development Resources and KDP 41
The Resistance of Elites to Democratization 42
Limiting Interaction Conflict 46
3. Why Do KDP Conflicts Not Escalate? 46

3.1 Problem Solving in KDP Forums 47
3.2 The Role of Facilitators 49
3.3 Other Complaints Mechanisms 51
The Importance of Complaints Mechanisms 53
3.4 The Role of Outside Actors in Solving KDP-related Conflicts 54
4. Comparison of KDP and Other Development-Related Conflicts 56
5. Summary 59
CHAPTER 4. DEVELOPMENT PROJECTS AS PART OF ‘THE SOLUTION’? THE IM-

PACT OF KDP ON NON-PROJECT CONFLICT 61
1. Forms of Conflict in East Java and NTT 61
2. Conflict in KDP and non-KDP Areas 64

2.1 Average Levels of Violent Conflict in KDP and non-KDP Areas 64
2.2 Changes over Time 66
CHAPTER 5. THE DIRECT IMPACTS OF KDP ON CONFLICT MANAGEMENT: THE

ROLE OF FORUMS AND FACILITATORS 69
1. The Use of KDP Forums for Managing Non-Project Conflict 71

1.1 Problems and Jurisdictions 73
1.2 Resistance from Local Elites and Facilitators 75
2. The Role of KDP Facilitators in Conflict Management 76

2.1 Legitimacy 78
xxii
Table of Contents
2.2 Willingness to Act to Seek a Solution 79

2.3 Capacity 80
Sub-district Facilitators (FK) 80
Village Facilitators (FD) 81
3. Summary 82
CHAPTER 6. INDIRECT IMPACTS OF KDP ON CONFLICT MANAGEMENT:

CHANGES IN GROUP RELATIONS, BEHAVIOR, AND NORMS 83
1. Group Relations 84
1.1 Forms of Difference in East Java and NTT 86
East Java 86
NTT 88
1.2 The Impact of KDP on Group Relations 90
Ethnic Relations 92
Religious Relations 93
Class Relations 94
Political Affiliation Relations 95
Relations between Silat Groups 96
2. Behavioral Change 96
2.1 The Dynamics of Local Level Democracy in Indonesia 98
2.2 The Impact of KDP on Local Democracy 99
Changes in Participation 100
Women and Participation 100
Spillover Effects? 103
KDP and Local Decision-Making 105
3. Changes in Norms and Perceptions 106

3.1 Changes in Problem Solving and Conflict Resolution 106
3.2 Is KDP Responsible for Changing Norms? 109
4. The Impact on Conflict Outcomes 111

4.1 Extra-Local Factors and the Incidence of Violent Conflict 111
4.2 Local Factors and the Likelihood of Conflict Incidence and Escalation 114
4.3 Strategies for Improving Conflict Management Capacity: the Demand Side of Good
Governance 117
5. Summary 118
CHAPTER 7. EXPLAINING VARIATION: THE INFLUENCE OF CONTEXT ON

PROGRAM SPILLOVERS 121
1. Explaining Variation 121
2. KDP-Triggered Conflict and Local Context 123
xxiii
2.1 The Effect of Local Capacity on KDP-Triggered Conflict 123

Levels of KDP-Triggered Conflict: The Impact of District Capacity 123
Levels of KDP-Triggered Conflict: The Impact of Village Capacity 125
Forms of KDP-Triggered Conflict and Forum and Facilitator Use 126
The Success of Forums and Facilitators in Managing KDP-Triggered Conflict 128
2.2 Program Functionality and KDP-Triggered Conflict 130
3. The Effect of Local Context on the Direct Impacts of KDP on Conflict Management Capacity 132
3.1 The Effect of Local Capacity on Direct Impacts 132
Use and Success of KDP Forums for Non-KDP Conflicts 132
The Different Impacts of District and Village Capacity 133
3.2 Program Functionality and Direct KDP Impacts 136
4. Indirect Impacts and Local Context 137

4.1 The Effect of Local Capacity on Indirect Impacts 137
Group Relations 137
Behavioral Change 138
Changes in Norms 141
4.2 Program Functionality and Indirect KDP Impacts 144
5. Summary 146
CHAPTER 8. CONCLUSIONS AND RECOMMENDATIONS 149
1. Introduction 149
2. The Development-Conflict Nexus 150
3. Ensuring Projects ‘Do No Harm’ 152
4. Direct Conflict Management Interventions 158
5. The Indirect Impacts of Development Projects on Conflict Management 161
6. Implications for Evaluating Community Development Projects 163
7. Concluding Thoughts and Future Directions 165
ANNEXES 167
Annex A: Details of Research Areas 167

1. East Java 167
2. NTT 174
Annex B: Evaluation Methodology 180

1. Qualitative Methods 181
2. Quantitative Methods 183
xxiv
Table of Contents
3. Secondary Data 185

4. Integration of Methods 186
5. Sampling Frame 187
Annex C: Analytical Tools 191

1. Assessing local capacity: use of vignettes 191
2. Classifying conflict outcomes 194
3. Patterns of Conflict in KDP and non-KDP Areas 194
Annex D: List of Conflict Pathways Case Studies 197
Annex E: Summary Statistics from Key Informant Survey 205
Annex F: Summary Statistics from KDP & CCN Dataset 206

1. Actor Combinations in Conflicts in East Java and NTT 206
2. Impacts of Conflict in East Java and NTT 207
3. Impacts of Conflict in Qualitative Research Areas 208
REFERENCES 209
BOXES
Box 2.1 : The Kecamatan Development Program (KDP) 8
Box 2.2 : East Java and NTT: Two Different Provinces 16
Box 2.3 : High and Low Capacity Districts 17
Box 3.1 : Conflict Avoidance in Madura: Dividing the Funds 26

Box 3.2 : Tensions from Competition in Slahung, Ponorogo 29
Box 3.3 : Examples of Village Level Corruption in Pamekasan, East Java 37
Box 3.4 : Power Takes Precedence: Four Million Rupiah Vanishes 38
Box 3.5 : Corruption and Revenge in Flores Timur 38
Box 3.6 : Road Surfacing Project - Desa Tattangoh, KDP-I Year 3 (2000) 41
Box 3.7 : Village Elections in Desa Tengku Leda, Manggarai 41
Box 3.8 : Abattoir Destruction and Violence in Madura 45
Box 3.9 : KDP Forums and Meetings 47
Box 3.10 : The Network of KDP Facilitators 50
Box 3.11 : Rice for the Poor ? Two Cases of Conflict over Rice Distribution in Sikka 53
Box 3.12 : Competition and KDP: Disputes in Manggarai, Flores 55
Box 3.13 : Aid and Conflict in Madura 58
Box 6.1 : The Village Head Elections in Palengaan Laok 87

Box 6.2 : The Role of Women in KDP in Flores 102
Box 8.1 : The Development-Conflict Nexus - Recommendations for KDP 152

Box 8.2 : Ensuring Projects “Do No Harm” - Recommendations for KDP 157
xxv
Box 8.3 : Direct Conflict Management Interventions - Recommendations for KDP 160
Box 8.4 : Indirect Impacts of Development Projects on KDP - Recommendations for KDP 163
Box 8.5 : Implications for Evaluating Community Development Projects-Recommendations for KDP 165
Box A1 : Political Voice in Mexico and China 191

Box A2 : Anchoring Vignettes Used to Assess Capacities for Problem Solving and Collective Action 192
FIGURES
Figure 2.1 : Forms of Impact 12
Figure 2.2 : Different Data Sources 13
Figure 3.1 : Did KDP Forums Trigger Conflict ? 22

Figure 3.2 : Forms of KDP-related Conflict 24
Figure 3.3 : The KDP Cycle 27
Figure 3.4 : Tensions Generated by Competition in KDP 30
Figure 3.5 : People who Helped Resolve KDP-Related Problems/Conflicts Outside KDP Forums 51
Figure 4.1 : Number of Violent Conflicts by Type in East Java and NTT (2001-2003) 64
Figure 4.2 : Mean # Violent Conflict Incidents per Year per Sub-District in East Java 65
and NTT (2001-2003): KDP and Non-KDP
Figure 4.3 : Mean # Violent Conflict Incidents per Sub-district per Year in East Java (2001-2003) 66
by Year of KDP
Figure 5.1 : Potential Direct Impacts of KDP on Conflict 70

Figure 5.2 : Types of non-KDP Problems/Conflict Addressed in KDP Forums by Province 72
Figure 5.3 : Were Problems/Conflicts Resolved by KDP Facilitators Outside of KDP Forum 77
Related to KDP ?
Figure 6.1 : Direct and Indirect Impacts on Conflict and Conflict Environments 84
Figure 6.2 : Changes in Ethnic Relations in NTT by KDP Years 92
Figure 6.3 : Changes in Religious Relations in NTT by KDP Years 94
Figure 6.4 : Changes in Class Relations in NTT and East Java by KDP Years 95
Figure 6.5 : Change in Number of Groups Coming to Village Meetings by KDP Years: 104
NTT & East Java
Figure 6.6 : Changes in Village-Level Decision-Making since KDP: NTT & East Java 105
Figure 6.7 : Changes in Village Decision-Making since KDP Years: NTT & East Java 105
Figure 6.8 : Has Village Problem Solving Improved Since KDP ? 107
Figure 6.9 : Changes in Village Conflict Resolution Since KDP 108
Figure 6.10 : How are Conflicts Solved in Your Area Since KDP Arrived by Years of KDP ? 109
Figure 6.11 : Changes in Village Problem Solving by Years of KDP 109
Figure 7.1 : Percentage of Respondents Reporting KDP-Triggered Conflict 124

by District Capacity
Figure 7.2 : Local Capacity and Whether KDP Forums Triggered Problems/Conflicts 126
xxvi
Table of Contents
Figure 7.3 : Local Capacity and Whether KDP Forums were Used for Program 127
Malfunction Conflict
Figure 7.4 : Local Capacity and Successful Resolution by KDP Conflict Type 129
Figure 7.5 : Types of KDP-related Conflict and Program Functionality 132
Figure 7.6 : Local Capacity and Positive Changes in Participation in non-KDP Meetings 139
(KDP vs. Control Sites)
Figure 7.7 : Local Capacity and Positive Changes in Decision-Making 140
Figure 7.8 : Local Capacity and Positive Changes in Decision-Making (KDP vs. Control Sites) 140
Figure 7.9 : Local Capacity and Positive Changes in Problem Solving 142
Figure 7.10 : Local Capacity and Positive Changes in Conflict Resolution 142
Figure 7.11 : Local Capacity and Positive Changes in Problem Solving (KDP vs. Control Sites) 143
Figure 7.12 : Local Capacity and Positive Changes in Conflict Resolution (KDP vs.Control Sites) 143
Figure 7.13 : The Impacts of KDP on Conflict Management in Different Environments 146
Figure A1 : Research Methods Used in the Study 187

Figure A2 : Overview of Site Selection 189
Figure A3 : Seriousness of Conflicts 194
TABLES
Table 3.1 : KDP Conflicts and Their Impacts (2001-2003) 23
Table 3.2 : Elite Involvement in KDP Compared to Non-KDP Development Projects 43
Table 3.3 : Change in Elite Involvement and Level of KDP-triggered Problem/Conflicts 44
Table 3.4 : Types of KDP Problems Dealt with in KDP Forums and Success Rates 48
Table 3.6 : Development Conflicts and Their Impacts 57
Table 4.1 : Conflict-related Deaths (2001-2003): KDP & CCN dataset 62

Table 4.2 : Mean # of Cases with Impact per Year per Sub-district in NTT & East Java (2001-2003) 65
Table 6.1 : People Who Go to KDP Meetings Who do not Usually Go to Village Meetings 100
Table 6.2 : The Role of Women in Village Life in Sikka 102
Table 6.3 : Type of Meeting and Women’s Attendance in Sikka, Flores 103
Table 7.1 : Local Capacity and Use of KDP Forums for non-KDP Conflicts 133
Table 7.2 : The Impacts of District and Village Capacity on Change Variables 134
Table A1 : East Java Research Sub-Districts (Population and Poverty) 168

Table A2 : East Java Research Villages 173
Table A3 : NTT Research Sub-Districts (Population and Poverty) 175
Table A4 : NTT Research Villages 179
Table A5 : Overview of Units of Analysis/Primary Data Sources 190
Table A6 : Village-level Capacity 193
Table A7 : Overview of KDP and Conflict Data 195
Table A8 : Overview of Comparative KDP Data in Different Villages 195
xxvii
Table A9 : Overview of KDP Impact Data Over Time 196

Table A10 : List of Conflict Pathway Case Studies 197
Table A11 : Key Informant Survey Respondents (treatment sites) 205
Table A12 : Key Informant Survey Respondents (control sites) 205
Table A13 : Characteristics of Key Informants 205
Table A14 : Actor Combinations in Conflicts Resulting in Death (2001-2003) 206
Table A15 : Conflict Impacts by District, East Java (2001-2003) 207
Table A16 : Conflict Impacts by District, NTT (2001-2003) 207
Table A17 : Levels of Impacts of Conflict in Research Sub-Districts (2001-2003) 208
Table A18 : Levels and Impacts of Conflict in Research Villages (2001-2003) 208
MAPS
Map 2.1 : Location of Research Provinces in Indonesia 15
Map A1 : Location of Pamekasan Research Sub-Districts 170
Map A2 : Location of Ponorogo Research Sub-Districts 173
Map A3 : Location of Manggarai Research Sub-Districts 177
Map A4 : Location of Sikka Research Sub-Districts 178
xxviii
Acknowledgements
This paper is part of a larger research project on local level conflict and participatory devel-
opment projects in Indonesia, the KDP and Community Conflict Negotiation (KDP & CCN)
study. The research was conducted by the Conflict and Community Development Team
within the Social Development Unit of the World Bank, Jakarta.
A number of people helped with the analysis presented here, and helped write sections of the
paper. Samuel Clark and Ambar Mawardi analyzed the key informant survey data used through-
out the paper. Joanne Sharpe supervised the collection and collation of the newspaper dataset.
Karrie McLaughlin, with the support of Ambar Mawardi, led the analysis in Chapter 4 looking at
the relative incidence of conflict in KDP and non-KDP areas. Samuel Clark helped write parts of
Chapter 7. Much of the qualitative data and analysis is from provincial reports on East Java
(written by Rachael Diprose) and NTT (written by Adam Satu and Patrick Barron). Suzan Piper,
Joanne Sharpe and Kristen Stokes translated much of the field material (from both Indonesian to
English, and vice versa). Michael Woolcock, Claire Smith and Patrick Barron developed the
initial hypotheses, research questions and the framework for analysis. Claire Smith co-managed
the project in the preparation and data collection stages. Patrick Barron managed or co-managed
the study from its inception to completion. Scott Guggenheim, Sri Kuntari and Michael Woolcock
provided guidance and support throughout the study.
The field researchers were Cici Novia Anggraini, Imron Rasyid, Endro Crentantoro, Saifullah
Barnawi, Luthfi Ashari, Mohammed Said, Olin Monteiro, Don dela Santo, Stanis Didakus, Peter
Manggut, Agus Mahur and Yan Ghewa. They were supervised in the field by Rachael Diprose
(East Java) and Adam Satu and Jessica Gillmore (NTT). Mohammed Said, Don dela Santo,
Elisabeth Dua Soru, Getrudis Erni, Budi Santoso, Mirnawati Ladongga, Ambar Mawardi, Susan
Mety and Hafez were involved in creating and cleaning the newspaper datasets. Others involved
at various stages of the research included Vivi Alatas, Anton Baare, Victor Bottini, Juana Brachet,
Jozefina Cutura, Leni Dharmawan, Christopher Gibson, Ela Hasanah, Alexa Hergesell, Kai Kai-
ser, Yatrin Kaniu, David Madden, Olivia Melissa, Melina Nathan, Ben Olken, Junko Onishi, Menno
Pradhan, Arie Purwanti, Taufik Rinaldi, Michael Roston, Sentot Satria, Matthew Stephens, Inge
Tan and Katherine Whiteside.
For generous financial assistance we are grateful to DfID, AusAID, the Norwegian Trust Fund
(Measuring Empowerment Study, coordinated by Ruth Alsop), and the World Bank’s Conflict
Prevention and Reconstruction Unit (led by Ian Bannon), Social Development Department, and
Development Economics Vice Presidency (especially the Research Support Budget-funded study
on assessing CDD Projects, led by Ghazala Mansuri). For helpful and detailed comments on ear-
lier drafts, we are also grateful to Junaid Ahmad, Michael Dove, Scott Guggenheim, Rick Messick,
David Mosse, Susan Wong, Yongmei Zhou, and the probing questions of participants at seminars
hosted by George Mason University, Johns Hopkins (SAIS), Harvard, MIT, Princeton, Universi-
tas Indonesia, the World Bank (Jakarta and Washington), and Yale, and at the Southeast Asia
Conflict Studies Network (SEACSN) conference in Penang, Malaysia, and the MOST-LIPI/
UNESCO conference on Conflict in the Asia-Pacific in Jakarta. Our respective families and part-
ners have endured our many long absences in the quest to bring this large study to completion; a
final and very special thanks to them for their forbearance and active support.
xxix
Glossary
Acronyms
ADR Alternative Dispute Resolution

APBDes Village budget (Anggaran Pendaptan dan Belanja Desa)
CCN Community Conflict Negotiation
CDD Community-Driven Development
CDR Community-Driven Reconstruction
CRDP Community Reintegration and Development Project (Rwanda)
FGD Focus Group Discussion
GAM Free Aceh Movement (Gerakan Aceh Merdeka); organization
formerly seeking Acehnese independence
GDS Governance and Decentralization Survey
IDPs Internally Displaced People
IDT Presidential Instruction for Underdeveloped Areas
MDGs Millennium Development Goals
NGOs Non-Governmental Organizations
NTT Nusa Tenggara Timur (Indonesian province)
OPK Special Market Operation (Rice Subsidy Program)
PAN National Mandate Party (Partai Amanat Nasional)
PKB National Awakening Party (Partai Kebangkitan Bangsa)
PODES Village Potential Statistics (Potensi Desa)
PPP United Development Party (Partai Persatuan Pembangunan)
SPADA Support for Poor and Disadvantaged Areas project (Indonesia)
Indonesian Jurisdictional Levels
Kabupaten District
Kecamatan Sub-district
Desa Village
Dusun Sub-village/hamlet
RT/RW Neighborhood/collection of neighborhoods
Institutions
BPD Village Representative Council (Badan Perwakilan Desa)

BPS Bureau of Statistics (Badan Pusat Statistik)
DPRD II District Legislative Council
Dolog Government warehouse for foodstuffs
Itwilkab District Inspectorate (Inspektorat Wilayah Kabupaten)
Kesbang National Security and Emergency Agency (Kesbang Linmas)
KPU National Electoral Commission
LKMD Village Community Resilience Council (Lembaga Ketahanan
Masyarakat Desa)
xxx
Glossary
LMD Village Development Board (Lembaga Masyarakat Desa)

Menko Polkam Coordinating Ministry of Politics and Security
Muspika Forum of district representatives from all security and government
agencies
PKK Family Welfare Organization (Pemberdayaan Kesejahteraan
Keluarga)
PMD Community Development Agency under the Ministry of Home
Affairs (Pemberdayaan Masyarakat Desa)
P2JD Village Road Upgrading Project (Proyek Peningkatan Jalan
Desa)
Government / Military Positions
Camat Sub-district Head

Bupati District Head
Kepala Desa Village Head
Klebun Village Head (Madura)
Lurah Urban Village Head
KDP Positions / Institutions
KDP Kecamatan Development Program

KDP-I First phase of KDP
KDP-II Second phase of KDP
KDP-III Third phase of KDP
FD Village Facilitator (Fasilitator Desa)
FK Sub-district Facilitator (Fasilitator Kecamatan)
KMKab District Management Consultant (Konsultan Managemen
Kabupaten)
KSP Savings and Loans groups (Kelompok Simpan Pinjam)
MAD Kecamatan Development Forum (Musyawarah Antar Desa),
also known as UDKP
MD Musyawarah Desa
Musyawarah Desa Village Meeting, also known as Musbangdes (Village
Development Forum, or Musyawarah Pembangunan Desa)
Musdes See Musbangdes
Penggalian Gaggasan Brainstorming Sessions (sub-village KDP meeting), also known
as Musbungdus (Musyawarah Pembangunan Dusun, Hamlet
Development Forum)
PjOK Kecamatan-level Development Coordinator
PTO Operational Technical Guidelines
RAB Draft budget
RMU Regional Management Unit
TPK Project Implementation Team (Tim Pelaksana Kegiatan)
TTD Village technical staff (Tim Teknis Desa)
xxxi
UDKP Sub-district Development Forum (Unit Daerah Kerja

Pembangunan), also known as MAD
UEP Productive Economic Enterprises (Unit Economi Produktif)
UPK Financial Unit at the sub-district level (Unit Pengelola Kegiatan)
Other Terms
Abangan Non-Muslims or Muslims influenced by traditional mystical beliefs

(Java)
Adat Traditional or customary (as in hukum adat, traditional law)
Bajingan Mafia-like criminal groups (Madura)
Belis Bride price
Bima System of governance (Manggarai)
Campursari Dance performance (Java)
Carok Duels over honor or for justice, usually carried out with sickles
(Madura)
Engkel-engkelan To be locked in argument (Java)
Gotong royong Mutual cooperation
Kampung Residential area or hamlet
Kedaluan Traditional sub-districts (Manggarai)
Kejawen Syncretic Islam (Java)
Kelompok Group
Kerja bakti Community service (Java)
Ketua Head (as in head of)
Kota Town, city
Kyai A Muslim cleric or teacher
Musyawarah Meeting where decisions are made based on deliberation and
consensus
Nyikep Custom of carrying weapons (Madura)
Orang miskin Poor people
Pemekaran Division of administrative areas
Preman Thugs
Priyayi Bureaucrats and gatekeepers of morality (Java)
Raskin Rice for the Poor (Beras Miskin)
Pajak Bumi dan Bangunan Land and building tax
Reformasi The reform era
Rp Indonesian currency (Rupiah)
Rumah Aspirasi House of Aspiration (proposed district level forum)
Santri Orthodox Muslims (Java)
Silat Martial arts groups
Suku Ethnic group
Tim Sukses Success teams
Tokoh Masyarakat Respected community figure/elder
xxxii
Chapter One
Introduction: Understanding the Conflict-Development Nexus in Indonesia
1. Development Projects and Conflict: Part of the Problem, or Part of a Solution?
Conflict and development go hand in hand. Programs and projects introduce new resources and
rules into communities. Disputes and competition over such resources, and the rules that dictate
their distribution, have the potential to fuel existing or trigger new conflicts. The result, when
things go wrong, is the very antithesis of sustainable development: benefits from previous projects
(schools, health clinics, etc.) are wiped out. Such conflicts can harden identity cleavages and
worsen social relations, undermining the cooperation needed for successful development and
sustainable conflict management. Conversely, programs which aim to be ‘conflict sensitive’ or to
explicitly address the causes of'Monflict2 can make communities less susceptible to destructive
conflict by shaping local contexts in ways that make violent conflicts less likely to emerge
and/or by strengthening mechanisms and local institutions to manage emergent conflicts.
Understanding how, and the conditions under which, specific programs help or hinder violence-
free development is important from both a development and peace-building perspective.3
Conf{1ct is an inevitable feature of social life the world over; the very processes and products of
“development” (rising levels of education, income, political voice, geographic mobility, press
freedom, gender equality, urbanization), and the insertions of new resources and decision-making
responsibilities into poor communities, is likely to intensify the means, motives, and opportunities
for local conflict (cf. Bates 2000; Barron, Smith and Woolcock 2004). However, even if conflict is
“normal”, violence is not. Some societies and communities are able to manage broad processes of
social change, as well as “everyday” conflicts and disputes, through largely peaceful means. In
others, small disputes escalate into larger, often communal, conflicts and/or turn violent, and
transition is wrought with destructive upheaval.
This report aims to examine in greater depth questions relating to the nexus between development
and different forms of local conflict. It does so by examining how the World Bank/Government of
Indonesia’s Kecamatan Development Program (KDP) interacts with social tensions and local
conflict, and how it affects local conflict management. By local conflict we mean disputes (violent
and non-violent) that play out at the local level: at the sub-district level and down.4
2
Uvin (2002) provides a useful summary of how thinking on the impacts of development projects on conflict has
evolved over time. He argues that “the nexus between development and peace is [now] a central focus of development
thinking and practice.”
3
There has been increasing attention paid to the links between development and conflict; Mary Anderson’s Do No
Harm (1999) is the seminal work. For an extreme example, see Uvin (1998) on the role of development actors in
Rwanda before the genocide.
4
Local conflicts can be distinguished from larger conflicts by: (i) their impacts, with most incidents resulting in fewer
casualties, and (ii) the level at which they are concentrated, with most incidents taking place at the intra- or inter-
village, rather than district or provincial level. While the impacts of specific incidents tend to be relatively minimal and
locally concentrated, cumulatively these local conflicts constitute a barrier to development and human security, and in
some cases provide the fuel for larger incidents of unrest.
1
It is important to note that KDP was not explicitly designed as a conflict reduction program. Its
goals (for all its iterations) are reducing poverty and improving local governance in Indonesia.
Other community-driven reconstruction programs, such as the Community Reintegration and
Development Project (CRDP) in Rwanda, have an explicit conflict reduction aim (Cliffe,
Guggenheim and Kostner 2003). In contrast, there is no mention of conflict in KDP’s project
preparation or appraisal documents. If KDP had been designed to explicitly address conflict,
there is no doubt that there would have been changes to its design and structure.5 Nevertheless,
KDP provides a particularly interesting focal point for examining the relationship between
development (programs) and local conflict for a number of reasons:
First, its size: at a cumulative cost of over a billion dollars, and having operated in over 28,000 villages
(40% of the total) in Indonesia6, KDP is the largest social development project in Asia and one of the
World Bank’s flagship “community-driven development” (CDD) programs. Since its inception in 1998,
the program has been at the forefront of the World Bank’s strategy and portfolio in Indonesia. Further, its
influence is spreading: KDP was one of the first projects showcased at the high-profile Shanghai
development conference in May 2004, in which donors, NGOs and multilateral agencies met to
share the lessons from “success stories”—that is, projects deemed to be having a positive impact
that could potentially be replicated and “scaled-up” in order to improve progress towards achieving
the Millennium Development Goals.7 It is thus of particular importance to empirically evaluate its
impacts, not only in terms of the extent to which it is reaching its stated goals, but also the way in
which it is interacting with the broader processes of transition and societal change, in the process
generating lessons which may be applicable elsewhere. This is particularly true given that the
evidence on the record of CDD projects still lags behind the extent to which they are being scaled-
up (Mansuri and Rao 2004).8 How can projects and programs aimed at the community level positively
support national and local processes of social and political change?
Second, its nature: as the “community-driven” banner implies, KDP is a decentralized project. By
design, significant opportunities for discretion in program implementation are afforded to those at
a local level, and the input of local communities is sought at every stage in the process (see Box 2.1
below). Policymakers and practitioners alike have placed considerable hope in the potential for
‘participatory’ development projects to improve a community’s capacity for collective action and
local conflict management (Chopra and Hohe 2004). There have been attempts to apply the design
principles of participation, transparency and local choice and accountability to post-conflict
settings, with the aim of achieving more effective and longer lasting reconstruction while making
future conflict less likely.9 Do such projects, in fact, achieve these goals?
5
Thanks to Susan Wong for this point.
6
Government of Indonesia (2004). In year 1 of KDP II (2002-2004), the program is operating in just under 22,000
villages. However, because several villages covered in KDP I did not continue into KDP II, the sum total is larger.
7
A major “global exchange” aimed at scaling-up development success stories was held in Shanghai in May, 2004.
KDP was one of the projects deemed successful and scaleable. See www.worldbank.org/wbi/reducingpoverty
8
The KDP and Conflict study is but one of a larger armory of evaluations and impacts studies that have taken, or are
currently taking, place. These include studies on KDP and corruption (Olken 2005; Woodhouse 2005), an economic
analysis of KDP infrastructure investments, a poverty impact study, an infrastructure technical evaluation, a study of
matching grants, of KDP knock-offs, and an impact evaluation of participation and local governance.
2
Third, KDP operates in Indonesia, a country in the midst of a vast, incomplete, and uneven
democratic transition, a process that has, at times, been accompanied by violence. In addition to
outbreaks of large-scale violent communal conflict in a number of locations, and ongoing
secessionist conflict in two provinces (which have been widespread and often violent), local
conflict has occurred across the country. What are the strengths and limitations of projects like
KDP in an unstable political environment, where identities, rules and access to power are being
constantly renegotiated? How can outside interventions such as KDP support progressive social
change in such a dynamic and fluid context?
2. Research Questions and Paper Structure
This paper seeks to contribute towards answers to these questions. The findings on KDP and
local conflict presented here are part of a larger research effort, namely the KDP and Community
Conflict Negotiation (KDP & CCN) study. The mixed methods study aims to provide a better
understanding of the nature and dynamics of local conflicts in Indonesia, and the roles—positive
and/or negative—of large development projects (and particularly KDP) in shaping the forms and
trajectories that local conflicts take. A number of papers—some with an operational focus, others
with a more theoretical/conceptual bent—have or will emerge from the study.10
There are two primary questions we seek to answer in this paper:
Does KDP generate fewer conflicts, or at least less serious conflicts, than other
development projects?
Does participation in KDP help villagers find more constructive solutions to local level
conflict, and, if so, does it help resolve certain types of local conflict more effectively than
others?
These questions give rise to a related set of secondary questions:
Through what mechanisms are any such positive outcomes achieved, and potentially
negative outcomes avoided or minimized? That is, how exactly do these outcomes
materialize?
What elements of context are important in determining the effects of KDP? What factors—
either internal to the program or in the local environment in which it operates—influence
the extent to which and the ways in which the program has an impact?
The rest of the report is divided into seven chapters. Chapter Two provides the background to
the study and outlines the methodological and analytical strategies used to collect and interpret
the data. Chapter Three considers whether KDP triggers conflict, whether these conflicts turn
9
Within the World Bank, these programs have been labeled as Community-Driven Reconstruction (or CDR); on this
see Cliffe, Guggenheim, and Kostner (2003). CDR programs have been implemented by the Bank in a range of
countries including Afghanistan, Rwanda, and Timor-Leste.
10
These papers are available at the study’s website, www.conflictanddevelopment.org
3
violent, and how KDP conflicts compare to those related to other development projects, in
quantity, seriousness and outcome. Chapter Four assesses the overall impact of KDP on conflict
management capacity. Using an analysis of a specially created dataset of conflicts in two
provinces, as well as comparisons between matched ‘control’ (non-KDP) and ‘treatment’ (KDP)
sub-districts, we begin to assess whether there are conflict management spillovers from KDP.
Chapters Five and Six examine the processes through which spillovers (positive or negative) are
generated, and the different mechanisms or ‘pathways’ through which such change is effected.
Chapter Five assesses the direct impacts of KDP on the management of conflict; Chapter Six
assesses indirect impacts on the ‘conflict environments’ in which the program is operating. Chapter
Seven assesses the extent to which different ‘contexts’ (internal and external to the program)
result in different effects. Chapter Eight summarizes the main findings and themes from the
analysis, draws out conclusions relating to the nexus between KDP and local conflict and conflict
management, and closes by providing recommendations for policy, project design, and future
research.
3. Conflicts of Interest?
Before proceeding to the analysis, we should draw attention at this point to our role as “inside
evaluators”. All authors and researchers involved in the study are, or have been, employed by the
World Bank. The results of the study are likely to be used as a basis for future program design,
perhaps putting additional pressure on the team to come up with results that justify an already given
course of action.11 In response, we can say that we were conscious of the inherent tensions in this
arrangement from the outset, and took all possible attempts to control at all times for (real and/or
perceived) conflicts-of-interest. External peer reviewers were involved at all stages of the research
design and data evaluation, and provided extensive comments on the final draft of the analysis.
A module in the training program for the field research team attempted to address these issues
explicitly.
It should also be noted that the conflict-of-interest issue does not necessarily go away if evaluations
are “outsourced” to an ostensibly independent third party. Given the market for consultants, it can
be argued that exactly the same pressures identified above accrue to evaluators who work for
external agencies. We present our findings with the hope that they will be interpreted and assessed
with the same constructive skepticism and engagement with which all empirical research should be
treated.12
11
One of the initial reasons the current study was launched was to help work out how the KDP model could be modified
to work better in areas with high levels of conflict. The result has been the design of a new Bank/Government of
Indonesia, the Support for Poor and Disadvantaged Areas Project (SPADA). The project became operational in early
2006. For an interesting reflection on the incentives against development organizations to evaluate their programs, see
Pritchett (2002).
12
In order to facilitate cross-checking of our findings, we will make our data available in various levels of aggrega-
tion on the project web-site: www.conflictanddevelopment.org
4
4. Making the Case
In attempting to bring empirical evidence to bear against a number of hypotheses pertaining to

local conflict and social development projects in Indonesia, it is important to be clear about the
general strategy employed in this paper. As development researchers and practitioners, our goal
all along has been to uphold stringent methodological research standards while striving to generate
useful and useable findings that speak most immediately to KDP staff and participants, and those
designing projects for moderate conflict environments in Indonesia (most notably the Support
for Poor and Disadvantaged Areas project, or SPADA), but also to a wider audience concerned
with assessing the efficacy of community-driven development (CDD) projects. These
deliberations—how best to assess the impact of CDD projects, and whether either the strong
support for, or criticism of, them is warranted by the evidence—are serious, as are the consequences
of them at a time when pressures are mounting to show how finite development resources can
best spur progress towards the attainment of international targets such as the Millennium
Development Goals (MDGs).
Our strategy has followed from a canonical (though too often ignored) research principle that the
nature of the problem should determine the choice of method(s) used, not vice versa (Mills 1959).
In this instance, both our ‘independent variable’—a large, complex community-driven
development project—and our ‘dependent variables’—local conflict and conflict management
capacity—inherently defy easy measurement; even broken down into seemingly more manageable
units, the task of trying to assess the impact of (say) KDP forums on the diversity of social ties does
not lend itself to an obvious or easy identification strategy. KDP is in essence a ‘package deal’, a
vast assemblage of people, resources, and processes endeavoring to achieve a diverse set of noble
but hard-to-define objectives. Assessing whether and how it achieves all or any of them requires a
correspondingly diverse methodological and empirical package.
To this end, we have broached this challenge by attempting to bring a vast but (we hope) coherently
integrated body of evidence to bear on it. ‘Rigorous’ is an adjective that has largely been co-
opted by certain strands of social science in recent years, and used (erroneously) to distinguish
between those deploying quantitative (‘truly rigorous’) and qualitative (‘merely anecdotal’)
approaches. Our view is that rigor is more properly understood as a function of the fit obtained
between the nature of the problem and logistical constraints, on the one hand, and the
comprehensiveness and quality of the evidence, on the other. Given the problem, the project, the
context, and the constraints, we believe we have been as truly rigorous as is possible in attempting
to answer the questions before us.
As such, the culmination of our work is not a single regression table where we show that,
controlling for a range of other factors, a narrowly-defined X can, at a certain level of statistical
significance, explain a certain (small) percentage of the variance of a narrowly-defined Y. Rather,
we have adopted an approach akin to that of contending lawyers, who present various pieces of
evidence—some ‘exhibits’ very compelling, others less so, still others merely suggestive—in the
hope that, cumulatively, they add up to a persuasive account. We are content to let readers be
the jury. Since we are not advocates for a particular view, we have sought to subject our
hypotheses to a full arsenal of methodological approaches; our data sources include everything
5
from national surveys to newspaper records to ethnographic fieldwork observations. Most

notably, we have sought to use this evidence to engage in detailed comparative case study
analysis, respecting the holistic nature of local conflict dynamics and the contexts within which
projects are carried out and participants’ lives lived. We hope this approach inspires further
explorations in rigorous mixed methods research and program evaluation.
6
Chapter Two
Assessing KDP and Local Conflict: Analytical and Methodological Frameworks
1. The Kecamatan Development Program (KDP) and Local Conflict
In what ways might the Kecamatan Development Program (KDP) interact with local conflict?
Understanding the potential dimensions of impact necessitates a brief explanation of the program
and of local conflict in Indonesia.
1.1 The Kecamatan Development Program (KDP)
KDP aims to introduce transparent, accountable and participatory development planning at the
local level. The program’s defining element entails the giving of block grants to committees at the
sub-district (kecamatan) level, largely made-up of non-governmental representatives from constituent
villages (see Box 2.1). Groups of villagers brainstorm and then prioritize ideas for small things
they would like to see funded in their village—e.g., the paving of a road, the building of a bridge
over a stream, a community center, or a savings and loan fund. Supported by input from technical
experts, such as engineers, they then submit proposals for funding to the committee at the sub-
district level.
The selection committee, comprised of villagers elected specifically to membership on it, then
evaluates the merits of the proposals according to their technical and financial feasibility, their
likely impact, and sustainability. At least one of the proposals from each village must come from a
women’s group. All deliberation processes are conducted in public, and all outcomes are then
posted on community bulletin boards, with journalists encouraged to report any abuses in local
newspapers; local NGOs are also trained and contracted to monitor any problems.
KDP thus introduces, or tries to introduce, rules-based, transparent and accountable competition
into village life. In the process it creates “winners” and “losers”—some proposals get funded,
some do not—and thus the potential for conflict. However, it also creates new spaces for public
deliberation, new avenues for participation by marginalized groups, and new opportunities for the
cultivation of civic skills: debating difference, managing meetings, keeping records. Do these new
spaces, avenues, and civic skills help villagers find constructive resolutions to project and/or non-
project related disputes? Or does the program worsen tensions and make conflicts more likely?
KDP also introduces rules relating to procurement and implementation, aimed at minimizing
corruption, while building expectations among villagers for transparency and accountability
(Woodhouse 2005). Does increased transparency make program-related conflict less likely, as
corruption is harder to get away with? Or does bringing such program ‘malfunctions’ to the surface,
and creating expectations for community oversight of other village development programs, trigger
fresh conflict? In short, can projects like KDP be part of a solution to managing local conflict, or
(like too many other development initiatives) is it part of the problem?
7
Box 2.1: The Kecamatan Development Program (KDP)
KDP consists of a rather straightforward system of decision-making and administration. Block grants of between US$60,000-
$110,000 are provided directly to kecamatan (sub-districts). These grants can be used for almost anything villagers themselves
feel is a development priority for their village.
KDP aims to encourage and institutionalize community participation in decision-making and priority-setting. A series of
forums/meetings is held at the hamlet, village and kecamatan levels. In many areas, these forums were largely defunct
prior to KDP, but were revitalized and now form an integral part of KDP. Suggestions on proposals originate at facilitated
meetings at the hamlet level (Penggalian Gaggasan). These are then taken to a village meeting (MD – Musyawarah Desa),
where the community democratically decides which two proposals are most worthy to be discussed at the kecamatan
meeting (MAD – Musyawarah Antar Desa). At the kecamatan meeting, delegations (which must consist of at least two
women and one man) present their proposals and together decide on which proposals will be funded. This forum produces
vigorous negotiations and horse-trading, as KDP, purposefully, does not fund all proposals.
The KDP process is facilitated at various stages by a network of local project facilitators who help to socialize the
program, organize the meetings, link the community with outside assistance (if necessary), and ensure project
implementation runs smoothly. In each KDP village, two FDs (Village Facilitators) are elected, who introduce the project
to villagers. At the sub-district level, two facilitators (FK) are appointed, one focusing on social issues and the other on
technical matters. That the FKs have institutional backing but are relatively independent of local power structures
means they are well-placed to trouble-shoot and facilitate problems that may arise.
Once the proposals are selected, which usually takes 6-8 months, each successful village forms an Activities
Implementation Team (TPK); the villages together form a Financial Management Team (UPK) for the entire kecamatan.
These are elected positions held by locals. The community then carries out their project, with the majority of labor and
material sourced locally. Most projects are finished within six months, bringing the total program time to approximately
one year. KDP is now in its third iteration (KDP-III). KDP-III differs from KDP-I and KDP-II in that it revisits places that
have already had the three-year KDP cycle for an additional two years. The focus is on making KDP institutions and
groups sustainable once KDP phases out.
Sources: Guggenheim (2006), KDP National Secretariat (2003)
1.2 Local Conflict in Indonesia
We should note here in more detail what we mean by local conflict. Post-1998, Indonesia has
experienced both major outbursts of large-scale conflict, as well as more local conflicts that have
fewer casualties associated with single incidents, and the effects of which tend to be concentrated at
lower levels of geographic specification.
In some ways the causes of both types of conflict are similar: increasingly, evidence is pointing
towards common patterns in the local causes and triggers of violence in ‘non-conflict’ provinces
like Lampung, and ‘high conflict’ provinces like Maluku (Barron and Madden 2004; Tajima
2004). Further, the impacts of local conflict are significant: a Government survey estimated
almost 5,000 deaths in 2002 alone, and this is undoubtedly an underestimate (Barron, Kaiser, and
Pradhan 2004). What distinguishes many of the high conflict areas in Indonesia is the interaction
of local factors with national phenomenon, including the role of Jakarta elites, the military, and
so on. The proximate causes of local conflicts, in contrast, tend to be situated at the local level
(i.e., below the district level), even if some of the underlying structures that make them possible
8
(e.g., national institutional configurations) lie at higher levels.13 The conflicts in our research areas
relate to issues such as land, local politics and other administrative issues, domestic violence, vigilante
killings, and so on. The largest case was a land conflict resulting in fifteen deaths, but smaller
conflicts resulting in fewer deaths or injuries are far more common.
Because KDP operates mainly at the sub-district level and down, we would not assume that it
would have any real impact (positive or negative) on conflicts that have their genesis at higher
levels. Thus, in both this paper and the broader study, we are seeking to measure the impact of
the program on the kinds of conflict that are observable at the levels at which the program itself
works.14 We use a wide definition of local conflict, to include relatively minor disputes, i.e.,
serious disagreements, and animosity between individuals/groups. We do this in large part because
it is such disputes that can sometimes escalate into larger forms of inter-communal violence.
Comparative analysis of violent and non-violent disputes can help us understand why some cases
turned violent, while other similar cases did not.
2. Analytical Framework: How Might KDP Influence Local Conflict?
Basic analysis of the KDP mechanisms presents a number of hypotheses about how the program
could shape the nature of local conflict in the areas where it functions.15 KDP may have impacts
upon local conflict and conflict management capacity in a number of ways, direct and indirect;
these effects, in turn, may be negative or positive.
2.1 Direct Effects
Direct impacts may be observable along two dimensions. First, KDP introduces decision-
making forums at the sub-village, village, and sub-district levels. In these forums, villagers and
village representatives meet to prioritize and then vote on which proposals should be funded
(see Box 2.1, above). These forums could hypothetically have either positive or negative impacts
on local conflict and conflict management capacity. Prior research has found that in some cases
KDP forums have been used to address conflicts that are not related to KDP (Government of
Indonesia 2002). Given the extent to which the legitimacy and authority of traditional forums
were eroded during the Soeharto period, we hypothesized at the project’s outset that KDP forums
13
Indonesia has a five-tiered system of governmental administration. Below the national level there are around 40
provinces. Below this are over 400 kabupaten (districts) to which many powers have been devolved post-decentraliza-
tion. On Java, an average district has around one million residents (our East Javanese districts averaged a population
of around 800,000). In eastern Indonesia, districts are smaller: the districts in Flores, for example, have between
200,000 and 650,000. Below this is the largely administrative kecamatan (sub-district), each of which has around 20-
50 villages. On Java, these can have 50-75,000 people; off-Java, 10-15,000 is the norm. The final formal level of
government is the desa (village). While the names of the bottom three levels differ for urban areas, the system of
government is largely the same (see Guggenheim 2006).
14
As such, we are also testing the limits of local factors in determining conflict levels and outcomes—e.g. the extent to
which local capacity and program benefits is limited by overriding external factors.
15
The hypotheses presented here were developed before the fieldwork began, and then modified inductively as data was
collected. For a list of the original hypotheses, see Barron, Diprose, Madden, Smith and Woolcock (2004, pp. 14-18).
9
may create a space wherein non-KDP related problems could be addressed and (hopefully)
solved.16
Conversely, the introduction of such forums could trigger destructive conflict. The KDP model
explicitly introduces competition, and thus contestation, over resources into the development
planning process. Poor villagers have all given valuable time to preparing their proposals, but
there is never enough money to fund all projects. Potentially, this could lead to conflicts in the
KDP forums, which, if not handled adequately, could become violent.
Second, KDP introduces facilitators at the village and kecamatan (sub-district) level. These
individuals are tasked with socializing the program, helping villagers identify and prioritize their
needs, and ensuring the project process (from the formation of proposals, to decisions on their
funding and implementation) runs smoothly. They also play an important role in monitoring the
program once implementation is underway. If these project facilitators are trusted and viewed as
impartial, they may also play a role in mediating non-project-related conflicts. There is
evidence—for example, from Lampung province—that KDP facilitators played an important role
in helping to calm tensions between the migrant Javanese population and local Lampungese after
a conflict involving the burning of Jepara village (Government of Indonesia 2002). Facilitators
could potentially play such a positive role as trusted and respected ‘insiders’ or impartial
‘outsiders’.17
These facilitators could also, however, play a negative role. Where they fail to fulfill their role as
program monitors, to resolve issues arising within their jurisdiction, to report or take action
against corruption, or where they themselves steal money from the project, they may trigger new
conflicts or allow existing ones to escalate.
2.2 Indirect Effects
Development projects shape the contexts—the social, political and economic structures—in
which they operate.18 The resources and rules that programs introduce, and the incentives these
produce, help shape the structural and relational contexts in which conflict becomes more or less
likely to arise and/or to escalate. There are a number of different ways in which KDP may shape
such conflict contexts—i.e., the local societal structures that make conflict more or less likely.19
First, the program may influence the relationships between different groups. In our villages,
identity cleavages exist along a number of dimensions, with ethnicity, religion, class, and
political affiliation being the most prominent, as well as other prescribed identities. Involvement
16
In some places, KDP uses preexisting forums. However, even here, in many cases the legitimacy and use of these
forums had eroded during the New Order era, where there were few opportunities for genuine participation in local
decision-making by villagers.
17
Village facilitators (FD) are selected by vote within the villages they represent; almost all live within the villages they
represent and thus inevitably play the role of insiders. In contrast, kecamatan facilitators (FK) are outsiders, in that
they are appointed by the project bureaucracy, and almost always are placed in areas that they are not from.
18
They are also obviously shaped themselves by such contexts; see Mosse (2005) and Chapter 7 of this report. For a
good summary of the public management literature on context effects, see Fukuyama (2004).
19
At present our model does not take into account forces exogenous to the areas in which we are assessing KDP. In
future analysis, we will consider the limitations of KDP in being robust to outside shocks.
10
at various stages of the program may improve the relationship of groups, both through the
demystification of “the other” and through the promotion of forms of collective action that
operate across groups. Conversely, the program may reduce social cohesiveness and/or trust
between groups, particularly if the groups who make proposals tend to be formed by people with
similar attributes (for example, ethnicity).
Second, the program may lead to behavioral changes and, in doing so, may reshape intra-group
and state-community relations. An explicit aim of KDP is to build the participation of villagers in
political and civic life, an important dimension of empowerment (Gibson and Woolcock 2005).
Attendance and participation in KDP meetings may help shape norms in ways that encourage other
forms of local level participation, for example in village government meetings. This may have a
positive impact on building the democratic decision-making skills of villagers, and this may spill over
into an improved ability to manage conflict.20 On the other hand, the very processes of social and
political empowerment that these processes involve may be met with resistance from elites. In
addition, raising people’s aspirations and rights-consciousness (a key element of democratization),
without making remedies available to right perceived wrongs, may make increased tensions more
likely.21 KDP may also change access to decision-making through legitimizing informal leaders
and creating better interaction between them and the state. This, in allowing for the incorporation
of local skills and expertise, may help legitimize informal leaders and create more and better
interaction between them and the state.22 Conversely, in doing so, the program may undermine
the authority of formal actors, hence weakening conflict management capacity.
Third, the program may lead to changes in norms. Violence is not only a symptom of conflict
but can also be a response to it. Where norms exist that legitimize violence as a course of action
and redress, conflict can easily escalate, thus fuelling cycles of violence.23 KDP emphasizes a
collective and inclusive process of decision-making and problem solving. Does the program
help people to understand how to solve problems in non-destructive ways, helping to build an
environment where collective and peaceful problem solving is the norm? Alternatively, the
program could also result in negative changes in local norms. If the program repeatedly triggers
conflicts, this may compound existing norms of retribution (i.e., of resorting to violence to
solve problems).24
20
This is the causal chain outlined in the UN Human Development Report (2002) and is implicit in the work of Varshney
(2002). We are testing this at the micro-level.
21
Whether such change is negative in the long run is another matter.
22
As Scott (1998) and others have shown, local knowledge (‘metis’) allows villagers to solve the problems which they
face. The New Order state, in its attempts to standardize structures of local government, eroded the role of traditional and
community leaders (tokoh masyarakat). KDP, in creating a formal role for such leaders, may allow for the incorporation
of local knowledge into program and non-program problem solving, as well as help create the synergy between civil
society and the state necessary for effective development generally (Woolcock 1998) and conflict management in particu-
lar.
23
Stewart (2005) finds that violence begets violence, and that it is one of the most salient predictors of further violence and
conflict escalation. This of course makes de-escalating conflicts more difficult.
24
The research found numerous examples of areas where violence is the norm for solving certain kinds of problems.
Research in other parts of Indonesia has demonstrated the pervasiveness and impacts of vigilante justice killings (see
Welsh 2003; Colombjin 2002 and 2005; Abidin 2005).
11
2.3 The Role of Context: What Influences the Ways in Which KDP Shapes Local Conflict?
The degree to which these effects occur is likely to be a function, in part, of the existing context
in which the program is operating, and of the ways in which the program is functioning. The
impacts of KDP on conflict, and the extent to which these various processes take place, are
likely to depend on context-specific factors.25 These variable factors will be both endogenous
and exogenous to the program. Endogenous factors will include the performance of program
facilitators and staff, and elite involvement in the program.26 Collectively, these factors
determine ‘program functionality’, i.e., the degree to which the program is functioning as
intended. Exogenous factors will include those related to the general pre-existing ‘capacity’ in
the area (i.e., the quality of local governance)27 as well as those that originate outside the local
area (e.g., interventions from national politicians, population flows, etc.). We call such factors
‘context capacity’. Together, we posit that these context factors will help determine the extent to
which KDP has positive and/or negative impacts on conflict and conflict management capacity.
2.4 A Framework for Analyzing Process and Impact
As such, we identify five different types of potential impact, each of which constitutes a causal
process through which KDP may have an effect on conflict and/or conflict management capacity.
The effect, if it is present, could be positive or negative. In some cases, it may be both. Chapters
Five and Six of the paper examine in depth the extent to which the hypotheses regarding the different
mechanisms of potential impact hold up. Chapter Seven considers variation in impacts—in type
and direction—through consideration of how different elements of context, external and internal to
the program, matter in determining the type and strength of impact. Our framework for analysis is
summarized in Figure 2.1.
Figure 2.1: Forms of Impact

Context Capacity
Low High
Type of Impact Program Functionality Program Functionality
Low High Low High
Direct
Forums (places)
Facilitators (people)
Group Relations
Indirect
Behavioral
Normative
25
This is particularly true for transaction-intensive programs such as KDP, where local actors have a certain degree of
autonomy in decision-making. See Pritchett and Woolcock (2004) and Whiteside, Woolcock, and Briggs (2005).
26
While we treat elite involvement as being endogenous to the program here, in reality it will in part be a function of
exogenous factors (e.g., whether a community is accustomed to having high levels of elite involvement in decision-
making) as well as the ability of the program to manage/change this.
27
For a fuller discussion of how we understand and define ‘capacity’, see the discussion below. Our approach was
influenced by that used during a previous World Bank study, the Local Level Institutions study (see Evers 2000).
12
3. Methodology and Research Locations
3.1 Evaluation Strategy: Data Sources Used28
Assessing the efficacy of social development projects is difficult because a defining feature of
many such projects is the non-standardized ways in which they seek to adapt to idiosyncratic local
circumstances and, in the process, generate outcomes (such as enhanced ‘participation’ and
‘inclusion’) that do not have an established or clear metric (Whiteside, Woolcock, and Briggs 2005).
As such, it is extremely difficult—analytically and empirically—to isolate the effects of a program
from other factors in the community and beyond, such as governance and local power structures,
and the selection mechanisms shaping both program placement (i.e., the political economy of where
the program is, and is not, located) and the participation of villagers (i.e., the choices and/or
circumstances leading some individuals, but not others, to be involved).
For this reason, we developed a methodological strategy that employs a number of different data
sources and types of research (see Figure 2.2). Each source illuminates part of the bigger picture:
large sample quantitative surveys (PODES and GDS) at one extreme; 68 ethnographic case studies
collected by teams living in villages for long periods at the other. The overall strategy driving the
analysis is that of comparative case analysis (George and Bennett 2005). Together, comparative
analysis of the data sources, combined with a careful sampling strategy that seeks to control for
exogenous sources of ‘difference’ (see below), allows us to reach stronger conclusions about impact
than would be possible if one data source was used alone.
Figure 2.2: Different Data Sources

Quantitative (breadth)
Decentralization survey (GDS)

Village Potential Survey (PODES)
Qualitative (depth)
Newspaper analysis
Key informant survey
Case studies
A team of twelve researchers and supervisors conducted nine months of qualitative fieldwork
in 41 villages. They developed 68 case studies of conflict pathways,29 which explored the
evolution of specific conflicts. Some cases were violent, others not. The primary reason for
28
See Annex B for a fuller methodology.
29
See Annex D for a summary table of the cases and their characteristics.
13
following non-violent cases is to allow for comparative analysis to determine why some cases
turned violent, while other similar cases did not. They covered a wide range of disputes including
land and natural resource conflicts (which range from large ethnic conflicts to private conflicts over
inheritance), cases of vigilante justice (against thieves, witchdoctors, etc.), gang fights, political
disputes (e.g., over local elections and administrative boundaries), conflict over development
resources, and domestic and sexual violence. Throughout the paper, these case studies are used to
illustrate particular dynamics of conflict, of conflict management, and of the functioning of KDP.
The researchers also collected rich ethnographic material on fourteen topic areas—ranging from
how local governments function, to local socio-economic conditions, to the role of traditional and
religious leaders, and so on—to allow for cross-village comparison. In all, over 800 interviews and
100 focus group discussions were conducted. A key informant survey was conducted in the research
villages to gather comparable responses to perception questions relating to KDP, its effect on conflict,
and processes of social change. The full survey was conducted in areas that had had KDP areas; a
shorter version was implemented in control sites.
In order to assess patterns and forms of conflict, and variations between areas, we used local
newspapers to create a dataset of all reported conflicts in the research areas, and surrounding districts,
between 2001 and 2003 (the KDP & Community Conflict Negotiation dataset). This allowed us to
broadly map conflict in our research areas, to estimate aggregate levels of violence, to see the
characteristics of the incidents (conflict type, actors involved, impacts, etc.) and to help us identify
how representative our qualitative case studies were.
Two other larger-N quantitative surveys were analyzed as part of the study: the Government’s
Potensi Desa (PODES) survey, which provides information on conflict for all 69,000+ villages in
Indonesia; and the World Bank’s Governance and Decentralization Survey (GDS). Background
papers gathering together secondary data and summarizing the relevant literature were written on a
number of issues relating to KDP, local conflict, and the areas where the study took place. Together,
these data sources provide the basis for a comparative framework; utilizing the different data sources
can help us control for differences in conflict outcomes, conflict mechanisms (e.g., conflict resolution
attempts, common escalation patterns30), and contexts. In different sections of the paper we draw
on different sources, quantitative and qualitative; at different points we also use different units for
comparative analysis: sub-district, village, and individual conflict case.
3.2 KDP and Conflict in East Java and NTT: Areas of Study
Provincial Selection
We conducted the qualitative and key informant survey research in 40 villages in sixteen sub-
districts in four districts in two Indonesian provinces. Selecting a relatively small number of research
locations allowed us to examine our research areas in depth. Our two research provinces were East
Java and Nusa Tenggara Timur (NTT); see Map 2.1 (below).
We chose East Java and NTT for a number of reasons. First, while neither province has
experienced unrest to the extent of, say, Aceh, Central Sulawesi or Maluku, both areas do have
30
See McAdam, Tarrow, and Tilly (2001) for methodological and analytical insights on breaking down larger events
into smaller ‘episodes’ for analysis.
14
significant levels of (often violent) conflict. Conflicts over land and natural resources, development
money, fights between gangs and martial arts groups, domestic violence, and battles over political
authority, are all common. Most academic and policy attention has been focused on areas of Indonesia
that have experienced conflict ‘big bangs’, yet previous research shows the extent to which conflict
is rife across Indonesia, and has posited that there are links between local conflict and larger outbreaks
of violence.31 Examining conflict and conflict management in areas which have experienced low-
intensity but still pervasive conflict may help us to understand how and why conflict escalated as it
did in higher conflict areas, as well as the links between the forms of conflict across Indonesia and
those in the conflict ‘hot spots’.
Map 2.1: Location of Research Provinces in Indonesia
Second, we deliberately decided to exclude the highest conflict provinces based on the assumption that
development projects are most likely to have a measurable effect (positive or negative) at the level
at which they are operational. KDP operates at the sub-district level and below: if the project does
produce any positive externalities that help communities constructively manage conflict, it is thus
most likely to be in aiding the management of conflicts that exist at the sub-district level and below.
Given the nature of conflict in many high-conflict provinces, where cleavages exist at the district or
provincial level, the selection of such places would have biased our research against observing any
project impact. Further, in areas of high conflict, where violence levels are affected significantly
by external actors and exogenous factors (e.g., military action), it would be much harder to separate
out the potential impacts of a local level project from other causal variables in the research site.
These reasons suggested picking ‘medium-level’ conflict sites for the research. This leaves
much of Indonesia as potential research areas. The Government’s PODES survey shows violent conflict
present in every province in the nation (Barron, Kaiser, and Pradhan 2004). Given that conflict exists
31
See Barron, Kaiser, and Pradhan (2004), Barron and Madden (2004), Tajima (2004), and Smith (2005).
15
everywhere, we decided to pick the provinces based on other measures of diversity, with the
assumption that we would find interesting conflicts to study wherever we went. The selected
provinces vary in terms of population size and density (high in East Java, low in NTT), ethnic
homogeneity (homogenous in East Java, heterogeneous in NTT), dominant religious group (Muslim
and Catholic, respectively), and provincial development (East Java is relatively rich, NTT extremely
poor); see Box 2.2. The rationale for this was that if we find similar patterns in very diverse
contexts, it is more likely that these findings will hold true across other locations.
Box 2.2: East Java and NTT: Two Very Different Provinces32
East Java is the most densely populated province in Indonesia, with a population of just over 35 million people (BPS 2002a). It is
one of the most developed provinces in terms of infrastructure, public services, and employment. The capital of the province,
Surabaya, has a population of 2.3 million and is the second largest city in Indonesia behind the national capital, Jakarta.
Often considered one of the heartlands of Islam in Indonesia, the majority of the population in East Java identify themselves
as Muslims; many national Islamic parties—such as the National Awakening Party (PKB) and the United Development Party
(PPP)—have strong bases in the province. While almost homogenous religiously and ethnically (Javanese on the main
land, Madurese on the island of Madura), there is diversity of prescriptive identity groups. These include martial arts groups,
Islamic boarding schools, affiliations to different Muslim clerics and prayer groups, geographical differentiation within and
across village boundaries, family and kinship groups, crime groups, political affiliation with parties and local elites, class,
occupational affiliations, and the like. Many of the identity groups overlap, creating complex human networks and cleavages
which emerge and dissipate in conflict situations (see Diprose forthcoming).
Common conflicts in East Java include petty crime, vigilantism, land conflict (particularly with regard to inheritance claims),
“social crimes”, and disputes over elections. Generally, the community relies on community leaders, such as the Village
Head, Muslim clerics and respected elders to facilitate the resolution of disputes. Where significant resources are at stake,
state institutions and political actors at higher levels (such as the Bupati – District Head) often become involved.
The island of Flores—part of Nusa Tenggara Timur (NTT) province in eastern Indonesia—is different in almost every respect.
Most of the 1.6 million islanders are Christian (primarily Catholic, but also Protestant) and subsistence farmers (BPS 2002b).
Flores is poor, isolated, considered politically insignificant, and receives little development and infrastructural support from Jakarta.
Ethnic variety in NTT, and likewise in Flores, is extraordinarily rich. One survey estimated there to be 61 ethno-linguistic groups
in the province alone, the result not only of migration over the centuries but of the continuing strength of clan and kinship
systems. Flores accounts for 28 of these groups, although some put the number of distinct languages on the island at closer to
50 (Grimes, Therik, Grimes, and Jacob 1997). Given the linguistic variation, and the corresponding absence of an overwhelmingly
dominant ethnic group (unlike in Java), Indonesian is the lingua-franca for communication. Ascriptive ethno-linguistic and clan
identities (often overlaying geographic cleavages) form the primary bases for differentiation and resource allocation.
In Flores, conflict commonly takes place over land and other natural resources, administrative procedures and allocations,
gender/sexual issues, and also takes the form of vigilante responses. Given villages’ isolation and the continuing strength
of traditional lifestyles, adat (traditional/customary) norms and institutions continue to play a strong role in the resolution of
these conflicts, and day-to-day affairs more generally, as does the Church (see Satu and Barron forthcoming).
District Selection: High and Low Capacity

Given the extent to which we relied on in-depth fieldwork, with long periods of time spent in
each location, it was clearly unfeasible to conduct research in all districts in our selected
provinces. 33 We hypothesized that the impact of KDP would be dependent in part on the
32
Boxes 2.2 and 2.3 are summarized from Diprose (forthcoming) and Satu and Barron (forthcoming).
33
East Java alone has thirty-two districts.
16
external environment in which it works and, as such, picked districts that varied based on the
local environment. One key element of the environment is local capacity, defined as the ability
of communities and/or the state to collectively solve or manage conflicts when they arise. In
‘high capacity’ areas, emergent problems are usually handled early and effectively (by formal
actors, informal actors, or a combination of both) so that they do not escalate and/or become
violent. In ‘low capacity’ areas, fragile local level institutions mean that conflicts tend to emerge
more easily, and escalate and/or become violent (see Box 2.3, below).34
Post-decentralization, Indonesia’s districts have become extremely important (and to some extent
autonomous) politically and economically. As such, the district seemed like a sensible place to
start in terms of ensuring diversity in capacity levels in the sample. We chose high and low capacity
districts in order to see how KDP works and interacts with conflicts in a range of environments: our
‘high capacity’ districts were Ponorogo (East Java) and Sikka (NTT); our ‘low capacity’ districts
were Pamekasan (East Java) and Manggarai (NTT). Districts were chosen based on interviews
with a range of stakeholders at the national and provincial levels.
The newspaper mapping was conducted in the same four districts where the qualitative fieldwork
and key informant survey were implemented. However, in addition, we also included
surrounding districts (all in all, seven in each province) in order to determine variation in conflict
levels at the district level, and to see how representative our districts were compared to their
neighbors.35
Box 2.3: High and Low Capacity Districts
Pamekasan district, on the island of Madura, was our ‘low capacity’ choice for East Java. The Madurese have a reputation
among Indonesians as being violent.36 Legends trace violence back to 929 AD when the island was founded, and there is
a long history of nyikep (the custom of carrying weapons). Following the rebellion of the Madurese and their separation from
the Mataram kingdom, a violent culture developed (De Jonge 1998). When repressed, poor citizens no longer saw the
institutions in the kingdom as legitimate centers of power, crime rates began to rise, and as a result the community began to
develop means of problem solving which did not involve the government (Wiyata 2002). The most well known form of this
is carok, the custom of defending one’s honor and seeking justice through a sickle duel. In Pamekasan, and Madura more
broadly, governance at the district, sub-district and village levels is poor, and less transparent than in other districts in East
Java. Cultural norms that emphasize violent responses are in some cases sanctioned by the state and other informal actors
such as the Kyai (Muslim clerics). As a result, violence is common and often escalates.
In comparison, Ponorogo district, which borders Central Java province, has a relatively high capacity for managing conflicts.
Culturally, it is very different. The dominant ethnic group is Javanese rather than Madurese. The social structure involves
three broad groups: the priyayi (associated with the bureaucrats and gatekeepers of Javanese morality), the abangan (non-
Muslims or, more commonly, those Muslims whose faith retains many elements of pre-Islamic custom), and the santri
(devout Muslims).37 In general, governance is reasonable compared to other districts in East Java. The district’s capital has
34
The assumption that local capacity matters in determining development and conflict outcomes may seem an obvious
one, especially given the extent to which scholars and practitioners have increasingly emphasized the importance of
governance, often at the sub-national level, in recent years. However, there is still a tendency in much research on
conflict to focus solely on the degree and types of underlying problems as determining conflict levels and outcomes,
rather than the endogenous ability of communities to deal with it. Chapter 7 provides details on interaction effects
between KDP and the ‘capacity’ of the prevailing institutional environment.
35
See Barron and Sharpe (2005) for a fuller explanation of the newspaper methodology.
36
These passages are summarized from Diprose (forthcoming).
37
These types were originally set out by Clifford Geertz (1976).
17
been awarded a rare Adiphura trophy from the ex-President herself in recognition of being a “model town”: clean, well run,
prosperous. Nevertheless, there are significant social problems. Aside from poverty (Ponorogo is the third poorest district
in East Java), there are tensions between rival martial arts (silat) groups, which have their genesis in the resistance to the
Dutch. However, these tensions are better managed in Ponorogo than in neighboring districts such as Madiun. Violence,
while not uncommon, tends not to escalate.
Manggarai is our low capacity district in NTT.38 Located at the western end of Flores island, Manggarai is the largest district
in NTT. Manggarai is the only district in Flores that was influenced and colonized by the Goa kingdom of South Sulawesi as
well as the Bima kingdom of Sumbawa island (Verheijen 2001). As a result of its history, Manggarai adopted the Bima
system of governance and was divided into 36 kedaluan (traditional sub-districts), which were often at war with each other.
These kedaluan became a basis for group identity that persists until today. The fact that modern administrative kecamatan
(sub-district) boundaries were built upon the historical borders of the kedaluan has only served to reinforce them as basis of
identity today. Manggarai is dominated by land conflicts over communally-held and orally-titled land. According to Lawang
(1996), a process of “demanggaraisation” has taken place over the last few decades, resulting in crises of public leadership
(among both traditional and government leaders). This problem has been compounded by widespread corruption. As a
result, both land and other forms of conflict often escalate, in cases resulting in significant impacts (Prior 2003; 2004).
Sikka district, at the opposite end of Flores, has comparatively high capacity. Ethnic identity is a less prominent feature of
social and political organization, although at the extended family/kinship level affiliation remains strong. A stronger Portuguese
influence remains, and can still be seen in the architecture of churches in rural towns. While traditionally the Sikka ethnic
group was used by the Dutch administration to run day-to day affairs, resulting in a privileged economic position compared
to other ethnic groups, today the dominance of the Sikka ethnic group has declined as education levels, and access to
political positions, have increased for other ethnic groups. Compared to most other districts in Flores, formal governance is
strong and transparent. Local people tend to be vocal about government affairs. During the New Order period, Sikka was
seen as a strong base for the opposition Nationalist movement. The presence of the Ledalero seminary, and the inclusion
of Florenese in government and the military, allowed for increased political freedoms in the district compared to other areas.
The police chief, recently posted to Lampung, was a noted reformer.
Local Level Sampling

Fieldwork was conducted in a total of sixteen sub-districts.39 Within each district, four sub-
districts (kecamatan) were selected. Three were treatment sites (i.e., they had received KDP for
at least three years) and one was a control site (i.e., it had not received KDP). For the first phase
of the study we selected locations that had KDP for at least three years, and matched them with
demographically and socio-economically ‘identical’ non-KDP control locations. We later
expanded the sample to include a larger range of KDP locations which had received KDP for
varying lengths of time.40 A full key informant survey was implemented in the twelve treatment
sub-districts; a short survey focusing on processes of social change was implemented in the
control locations.41 Villages were selected based on the location of “interesting” conflict cases
38
These passages are summarized from Satu and Barron (forthcoming).
39
See Annex A for a full list and description of research locations.
40
Sub-district selection was done using propensity score matching techniques and in-field interviews, in order to match on
both ‘observable’ (variables that could serve as proxies for the economic level of the kecamatan, such as population size,
access to markets, hospitals, department stores, health and education resources, main source of income, perception of poverty
levels) and ‘unobservable’ (social cohesion, motivation, leadership quality) factors. This process generated treatment and
control sites that were as statistically (and qualitatively) ‘identical’ as possible; the other two treatment locations in each
district were picked to maximize diversity (in local conditions and KDP performance).
41
In three of the four ‘control’ locations, KDP began to operate in the year between the qualitative research finishing and the survey
implementation; thus the areas were no longer pure control locations. As such, for some of the comparative analysis in later
chapters, we disaggregate analysis by the number of years an area had had KDP, rather than simply comparing treatment and control
locations.
18
identified in the qualitative research, using detailed criteria aimed at making it easier to control
for non-program effects.42
For the qualitative fieldwork, sampling of informants was driven by two aims. First, we wanted
to talk with a wide cross-section of the population within each district, sub-district and village.
Special emphasis was placed on talking to non-authority figures and women. Second, the
researchers used “snowball sampling” in order to find interesting respondents who would be able
to provide insights into the cases being followed and, more generally, issues in the village. Focus
group discussions were conducted with particular populations groups: poor women; educated/
elite women; young men; and poor men. For the key informant survey, eight respondents were
interviewed within each village, with an additional three at the kecamatan level. Informants
were selected based on their knowledge of KDP.43
42
More details on local level sampling (including maps of the research districts) are contained in Annexes A and B.
43
See Annex B for a full list of informant sampling procedures.
19
20
Chapter Three
Development Projects as ‘Problem’: KDP and

other Projects as Sources of Local Conflict
Development projects introduce new resources into communities. In the absence of effective
rules and procedures for managing these resources, these projects may trigger new, or feed
existing, conflicts. KDP also aims to empower marginalized individuals and groups within
villages, thus changing local power balances and hierarchical structures. This in turn may cause
conflicts as elites, and their support bases, resist the process. This chapter evaluates the degree
and nature of conflicts triggered or caused by KDP. It seeks to answer the following questions:
Does KDP trigger or cause new conflict, or worsen existing conflicts?

What forms do these conflicts take, and what are their impacts?
How does KDP compare with other development projects—both governmental and non-
governmental—in terms of its effect on local conflict?
We start by outlining the prevalence and impacts of KDP-related disputes and conflicts. We find
that small-scale (‘micro’) conflicts related to KDP are common. However, these rarely escalate
into broader conflicts or have significant negative impacts. We then examine the different forms of
project-related conflict: in-built, program malfunction, and interaction conflict. The latter two are
more serious than the first. Section three examines why KDP disputes and conflicts tend not to
escalate, and argues that KDP’s very design features—its forums, facilitators and complaints
procedures—mean that any tensions tend to addressed before they grow larger. Only where these
forums, facilitators and procedures do not work effectively, and in environments where prevailing
tensions and political competition are already high, is there a significant risk of conflict escalation.
Finally, we compare the presence of KDP-related conflicts with those that stem from other
development projects, governmental and non-governmental. We find that in general KDP tends to
produce less violent conflicts than other development programs (even though it is, in many cases,
much larger), and that this is a function of these other development programs tending to have both
weak conflict management mechanisms and a lack of transparency, accountability and opportunities
for community participation in decision-making.
1. The Prevalence and Impact of KDP-related Conflicts
Micro-conflicts related to KDP are common. The program does lead to tensions between
villages, between groups within villages, and within different groups. However, these tensions
very rarely escalate into larger conflicts, and they almost never become violent. The key
informant survey asked whether KDP had caused “community problems/conflicts”44 in the
respondent’s village/sub-district. Nearly half of survey respondents indicated that
44
Throughout the study, respondents were asked about “community problems/conflicts.” We used both terms to ensure
that informants defined conflict widely to include small-scale disputes and non-violent clashes. In the paper we use the
terms problems and conflicts inter-changeably. For the full definitions used in the study, see Barron, Diprose, Madden,
Smith, and Woolcock (2004, pp. 11-14).
21
problems/conflict occurred as a result of village-level KDP meetings: 45.5% of respondents

reported problems/conflicts in East Java and 19.4% in Flores.45 In sub-district forums, significantly
higher levels of problems/conflicts were reported: 85.0% reported problems/conflicts in East
Java; 57.3% did so in Flores (Table 3.1).46
Rates of reported conflict are higher at the sub-district, for a number of reasons. First, competition
is generally more pronounced between villages than within them (see Section 2.1, this chapter,
below). There is more pressure on village representatives to deliver resources and services than
there is for groups within villages to have their proposals forwarded on to the sub-district level.
Second, inter-village meetings are more likely to be between equals than village ones. Inter-village
meetings are held with representatives of similar status from each village; power imbalances are
thus less likely. Within villages, certain groups are more likely to dominate decision-making. Where
power differentials mean there are major imbalances in the “terms of recognition” of different
groups, ‘losers’ (i.e., those who do not get proposals funded) are less likely to engage in responses
that entail violent conflict (Gibson and Woolcock 2005).47
Figure 3.1: Did KDP Forums Trigger Conflict?
Source: Key informant survey (treatment sites only)
“Don’t judge success by whether only a few problems emerged, but how many problems came up and how many were
successfully resolved. If a kecamatan says that they came across three problems but they were only able to resolve
one of them, and another kecamatan had ten problems but they resolved six … the kecamatan able to resolve six
problems was more successful…”
Sub-district Head, Jenangan, Jenangan, Ponorogo (1305)48
45
The N for East Java = 132; for NTT = 134. We explore reasons for variation between areas in greater depth in
Chapter 6.
46
The N for East Java = 80; for NTT = 82.
47
On this, see also Fung and Wright (2001).
48
For each informant, we note in sequence the village in which they live, the sub-district and the district. Where
informant names are used, these have been changed to ensure confidentiality. Numbers in parentheses relate to
the codes for the interviews conducted. Those not coded were informal interviews.
22
Development Projects as ‘Problem’: KDP and other Projects as Sources of Local Conflict
However, despite the high levels of conflict/problems reported by informants, very few of these
conflicts escalate into violence. No survey informants in East Java reported violence in village-
level meetings; in NTT, only one informant noted violent conflict.49 At the sub-district level,
one respondent (out of 68) reported violence in East Java and two (out of 47) did so in NTT.50
Even so, these incidents of violence were not serious; they were not reported in local
newspapers.51
Indeed, the findings are backed up by both the qualitative fieldwork and the newspaper survey
of conflict and violence in East Java and NTT. The field team in East Java found eleven cases of
KDP-related conflict in the six KDP sub-districts where the research was conducted. In all
eleven cases, there were no deaths or injuries; the only violent impact was the damaging of a
water pump in one location. In NTT, the team found five conflicts with no physical impacts.
Across the fourteen districts covered in the newspaper survey, forty-two KDP-related conflicts
were picked up between 2001 and 2003. Of these, only one became violent, resulting in two
injuries (Table 3.1).52 As we will discuss later, these rates are extremely low when compared
with other development programs, policies, and projects, with 37 violent development conflicts,
not relating to KDP, taking place in the research areas (see Section 4, this chapter, below).
Table 3.1: KDP Conflicts and Their Impacts (2001-2003)

KDP conflicts # Conflicts # Violent Conflicts % Violent # Killed # Injured # Properties
Damaged
Total 42 1 2% 0 2 0
East Java 10 0 0% 0 0 0
NTT 32 1 3% 0 2 0
Source: KDP & CCN newspaper dataset
Why are micro-conflicts within KDP so common? Why do they so rarely escalate into broader
problems? And why do they almost never become violent? Answering these questions requires
consideration both of the types of conflicts and problems caused by KDP, and of the ways in
which they are resolved or prevented from escalating.
2. Forms of KDP-related Conflict
KDP can cause or heighten conflict in direct or indirect ways. This is demonstrated schematically
in Figure 3.2.53 The program directly leads to conflict where a particular element or procedure
of the program results in tensions. This includes situations where individuals or groups feel
dissatisfied with either the processes or results of the program—e.g., with how decisions on
49
In East Java, 95% said there was no violence, 5% did not know. In NTT, 92.3% said there was no violence, 3.8%
reported violence, and 3.8% did not know. Sample sizes: In East Java, n = 60; in NTT, n = 26.
50
In East Java, 91.2% said there was no violence, 1.5% reported violence, and 7.4% did not know. In NTT, 91.5% said
there was no violence, 4.3% reported violence, and 4.3% said they did not know. Sample sizes: East Java (n = 68); NTT
(n = 47).
51
Interviews with editors and journalists in both provinces determined that most incidents of serious violence would be
reported in local newspapers (Barron and Sharpe 2005).
52
The one violent KDP case related to corruption (see Box 3.5 below).
53
The conflict may be violent or non-violent. The forces that help ‘cause’ a conflict may not be the same as those that
make it escalate/turn violent. In the analysis of each of the KDP conflict types (below), we explore these issues in more
depth. 23
funding were made, with the role a particular facilitator played, and so on. In such cases, KDP
can trigger conflict. There are two basic types of conflict directly related to KDP: in-built
conflict and program malfunction conflict. The first relates to the conflictual processes (of
deliberation, advocacy, and debate) purposefully designed into the KDP model. The second
concerns what happens when the program does not function as intended, for example when
there is corruption or when implementation is poor.
Other times the impact is indirect. In such cases, the program interacts with existing tensions
and power structures, indirectly leading to new or escalated conflict. Here the introduction of
the program was not a cause of the conflict per se. Rather the program, in creating a structural
opportunity for the realignment of power relations, resources, interests, and values, contributed
to conflict.
Figure 3.2: Forms of KDP-related Conflict
We can thus identify three different forms of KDP-related conflict:
1. In-Built Conflict: Conflict directly related to competition in the KDP project cycle
2. Program Malfunction Conflict: Conflict that is a direct consequence of KDP not working
as intended
3. Interaction Conflict: Conflict indirectly caused by KDP, where the program interacts with
existing tensions and/or power structures
As we discuss below, in many cases these forms of KDP-related conflict overlap. However,
separating out the three provides a useful framework for thinking about conflict caused by KDP,
and KDP’s interaction with existing local tensions, because, as we explore in more detail below,
each has different (more or less serious) impacts.
2.1 Competition in KDP: A Source of Local Conflict?

“At the village level, for example, there were some feelings of jealousy which emerged between dusun
[hamlets] because some of them had their proposals accepted while others did not. Those which were
unsuccessful asked ‘why did they receive savings and loans funds from KDP and we did not?’”
Head of KDP Financial Management Unit, Wae Belang, Ruteng, Manggarai (M.1027)
24
“The competition principle is actually very good; it encourages people to think.”

Midwife, Kemiri, Jenangan, Ponorogo (1312)
“When you compare them, I think that KDP is better than P3DT [a government development project].
The competitive system can reduce the possibility of ngamlopi [providing money in envelopes] to officials
who carry our monitoring visits to the field … P3DT always used to provide envelopes to officials who
came to the site, but there is none in this KDP, the community understands the process.”
Sub-district Head, Slahung, Ponorogo (1323)
KDP’s model of decision-making is based upon the principle of open competition. Such
competition is a potential source of conflict. Groups within villages put forward proposals for
things they want funded.54 In the village, these groups need to come together to decide which
ideas should be taken to the village-level meeting, and then which proposals should go forward
to the sub-district level for consideration. In the sub-district forum, different villages need to
collectively decide which proposals deserve funding. Project resources are not enough to fund
all proposals. In essence then, villages and groups battle it out amongst themselves for access to
resources. Groups use the traditional tools of coalition building (the powers of rhetoric, appeals
to reason, bargaining and deal-making) to win support for their proposals.
A schematic overview of the project cycle (see Figure 3.3) shows a number of points where such
competition is explicit, and hence where tensions can and do emerge:
Brainstorming meetings at the dusun (hamlet) level.
Decisions on which proposals should go forward to the kecamatan-level forum. Only

two proposals are allowed from each village, one of which must come from a woman’s
group. Given that proposals are normally from geographically concentrated groups (e.g.,
villagers from one hamlet), village-level decision-making may cause tensions between
groups in the village, particularly at the second village-level meeting (MD II).
Decisions on which proposals should be funded in the kecamatan forum. Given that not
every proposal can normally be funded, inter-village conflict may arise over which villages’
proposals are funded, particularly at the third inter-village meeting (MAD III).
Conflict may arise over deciding who will implement the projects. This may cause tensions
at the third village meeting (MD III), where village implementation teams are formed, and
during the process of implementation.
54
Roads and other small-scale infrastructure make up the majority of funded proposals. For the first year of KDP-
II, for example, KDP resources went to the following main subproject activities: roads (48.1%), savings and loans
(10.2%), clean water (9.7%), education (6.4%), irrigation (6.2%), bridges (5.0%), other economic activities (4.4%),
other infrastructure (4.0%), sanitation (2.2%), health (2.0%), markets (1.4%), and electricity (0.4%) (Government
of Indonesia 2004).
25
These types of tensions are deliberately designed into KDP. Indeed, KDP explicitly aims to
create non-violent conflict through stimulating competition at these stages. Such ‘managed
conflict’, it is hoped, can lead to better proposals, can increase incentives for participation, and
can lead to an exchange whereby different villagers (and villages) learn from each other as to
how to design and present effective projects that are likely to win funding, and, concomitantly,
learn to deal with disappointment. Further, it is hoped that the competitive process—conducted
as it is within a rules-based system that is simple, transparent, and accountable—will help villagers
acquire organizational and civic skills that carry over to the successful management of local
conflicts.55
Yet there are also risks with this strategy. In areas where tensions were already high, some
informants argued that stimulating competition between villages led to more conflict. In Pamekasan
district in Madura, East Java, village leaders and local facilitators had pursued a deliberate strategy
of avoiding competition in the allocation of funds, although some informants argued that this
was as much a means of maintaining their own power base as a strategy for reducing tensions in
the community (Box 3.1).
Box 3.1: Conflict Avoidance in Madura: Dividing the Funds
In a number of sub-districts in Pamekasan, East Java, efforts were made to limit competition between villages in the
KDP process. For Year One of KDP-I in Pademawu sub-district, it was decided at the sub-district level that only five
villages (of the twenty-two in the sub-district) would be eligible for funding in that year.56 Funds were then divided
almost exactly equally between the five villages. In Years Two and Three, again, the funds were divided equally
between all eligible villages. Competition was absent (see Probo 2003a). A member of the Village Council (BPD) in one of
the villages that received money in Year Two (again without competition) describes the motivations behind splitting
the money equally:
“The impacts can be great if [funds] aren’t divided equally between all the villages, for instance if Village A gets [funds] while
Village B alongside it doesn’t. This can turn into a ‘boomerang’ for the Klebun (Village Head) because he will be seen as
incompetent in attracting the program.”
Desa Padellegan, BPD member, Pademawu, Pamekasan (1217).
The limitation on competition was less about limiting community conflict, than about Village Heads cementing their own
power. Village Heads see the potential for attracting development programs to the village as a means of consolidating
their power and status; they thus spare no effort to ensure they receive a project. The easiest way to do this is to
agree with other Village Heads to split the money.
55
We test this hypothesis in Chapter 6 below.
56
This was not in line with KDP-I regulations. For year one of KDP-I, program rules stipulated that if a sub-district
(kecamatan) had more than ten villages, half should be eligible for the program, with the remaining fifty percent
eligible in year two, and all villages eligible in year three. Thus eleven villages should have participated in year one.
The limit on the number of eligible villages in the first two years was not instituted however, to limit competition.
Rather, the number was limited in order to ensure all new villages could receive adequate facilitation (see Wong 2003).
26
Figure 3.3: Potential Flash Points in the KDP Cycle
KDP CYCLE
HOT SPOT
Program Preparation Group/sub-village
meetings
Workshop FK Selection FD Training

Prov/District and Training Special Womens Meeting
(Womens Proposal)
Inter-village Village Meeting MD II (Determining Village

Meeting (MAD) I (MD) I proposals)
Proposal writing
Proposal with/without Design
Maintenance of Infrastructure Verification and Budget
HOT SPOT
and Loan Repayment
Preparation for MAD II (Ranking Villages Preparation

MD Handover to and finalization
Implementation Projects)
Village Recruitment of of Design and
Workers Budget.
MAD III (selection and Proposals
prioritized.
Supervision of Dana Revisions of determintation of proposals to
Implementation of funds and activities be funded, complete with
activities, visits for implementation, design and budget)
between villages and disbursement HOT SPOT
MD III (reporting on outcome

HOT SPOT HOT SPOT of MAD III and discussion to
MD Accountability
prepare for Implementation)
(min2x)
HOT SPOT
27
The research in all four districts in the two provinces found that the competition inherent in the
KDP process did sometimes cause tensions. In East Java, 61.4% of survey respondents reported
that competition over proposals in village level meetings had led to tensions or conflict; 19.4% of
respondents in NTT reported the same.57 The qualitative fieldwork backed up this finding, with a
number of informants reporting cases of tensions related to KDP competition:
“It is normal for people, who feel disappointed because they didn’t get a loan, to make threats, ordering the people who did
receive loans to pay their money back immediately. They will say: ‘seize their processions’. If they don’t repay their loans we
surely won’t get any more loans. Someone even suggested that they be sent to the police or to jail.”
Respected Community Figure, Nebe, Talibura, Sikka (S.1114)
“Feelings of dissatisfaction with the KDP mechanism can be seen on a micro-scale with the onset of conflict in the KDP
forums. The community likes to make a lot of noise in the KDP forums. At the kecamatan level, they speak loudly and
assertively because they feel dissatisfied with the competition process. In fact, there have even been people who have
walked out because they didn’t agree with the decisions made at the MAD (Inter-Village Forum).”
Sub-District Head, Ruteng, Manggarai (M.1027)
“In the Inter-village Meeting in Year Three of KDP, the [atmosphere] was tense. [Q: In what way was the atmosphere tense?]
Yes … tense … well, each hamlet put forward their proposals and each hoped that their proposal would be prioritized. This
was the cause of the tension because the community knows that the KDP process begins at the bottom [it is bottom up]. So
each hamlet hoped to win.”
Villager, Sumedangan, Pademawu, Pamekasan (1221)
In one village in Slahung in Ponorogo district, East Java, competition fuelled tensions between
different sub-villages (Box 3.2).
Yet in general, while competition (either within or between villages) led to disappointment for
groups who were not successful in getting funding, rarely did this lead to prolonged tensions. In
the case in Slahung, the research team felt it unlikely that the conflict would reemerge in a destructive
form. Indeed, as we will explore in more depth, in no cases did tensions related to competition
escalate into more serious conflict. Competition means that meeting participants argue (engkel-
engkelan), but in most cases, with only a few exceptions, tensions within KDP forums (at every
level) do not spill over into everyday life.58 We did not find any cases in the research where tensions
relating to competition per se fuelled anything more than small-scale disputes. Indeed, as we
explore in Chapter 7, it was lack of competition that was more likely to cause tensions.
57
Sample sizes: East Java (n=132); NTT (n=134).
58
In Jenangan sub-district in Ponorogo, a number of informants made this claim: Head of the Savings and Loans
Group, Jenangan village (1304); Teacher and Former Member of the LKMD, Kemiri village (1310); Treasurer,
KSP/Kelompok Simpan Pinjam, Panjeng village (1314); Village Head, Panjeng village (1316); Krajan Hamlet
Head Assistant, Panjeng village (1319); woman community figure, Panjeng village (1320). The only case we
found where tensions did spillover related to tensions surrounding the building of a kiosk in Badegan village,
Ponorogo. See the case study “Caught Between a Rock and a Hard Place: The Dilemma for KDP Facilitators in
the Kucur Tourism Market Case”.
28
Box 3.2: Tensions from Competition in Slahung, Ponorogo
Competition over funding led to tensions in Wates village, Slahung sub-district, Ponorogo. Two hamlets (dusun) proposed
the paving of roads in the village. When the proposal from Bedog hamlet lost out, tensions emerged.
“The Musbangdes (village KDP meeting) led to tensions between Joso and Bedog hamlets … The tensions were due to
both of them submitting road paving proposals … The Musbangdes participants were engkel-engkelan (locked in argument)
so the meeting became tense…”
Male Religious Community Leader, Wates, Slahung, Ponorogo (1326)
Bedog sub-village had already submitted the proposal twice (in KDP-I, Years Two and Three), but had never received
funding because the money needed was relatively large. Determined to win funding this year, the community had spent two
to three months of kerja bakti (community service) to widen the road, in the hope that this would make funding more likely.
However, the problems began when the verification team said the road was still not wide enough. Further, the Village Head
(Kepala Desa) was worried there would be problems because the road to be paved passed though state forestry land, and
there was no permit to allow this.
“The Bedog community was not only somewhat disappointed but felt very disappointed with the competition results. How
else would they feel? The community had [carried out] gotong royong (self-help, mutual cooperation) ngepras (cutting the
hillsides to widen the road) for three months but they weren’t funded.”
Male Teacher, Wates, Slahung, Ponorogo (1331)
Instead, a rival proposal for a road to be paved in Dusun Joso won out. The Village Head was from Joso, and villages
started to suspect him of bias. It did not help that the FD (the village level KDP facilitator) was the Village Head’s younger
sibling. Residents from Bedog were extremely upset and felt that it was a continuation of an ongoing trend where they were
neglected in the development of Wates. Disappointment and suspicion fed into other problems in the village, in particular
the polarization of the village around two different candidates in the 1999 village head elections. While no explicit conflict
emerged, relations in the village have worsened.
In the same sub-district, competition at the inter-village also led to tensions. Certain villages who did not have their proposals
funded accused the sub-district facilitator (FK) of receiving bribes, although no evidence was produced. Some of the
villages that lost refused to pay the Pajak Bumi dan Bangunan (land and building tax) to the kecamatan authorities.
Tensions over Competition Within and Between Villages: Changes over Time
In the short-run, tensions related to competition are more likely to take place within villages.
However, as the amount of time KDP has been in a sub-district increased, these tensions tend to
ebb, while, conversely, the potential for inter-village tensions related to competition increases
(see Figure 3.4).
In the villages where fieldwork was conducted, intra-village tensions related to competition
increased gradually in the first couple of years of the program, but then started to tail off as the
program stayed for a longer period of time. In the first couple of years, tensions are often
relatively high. This is because in many cases the principles KDP embodies are different than
those normally applied in village-level development-related decision-making. While forms of
consensual decision-making such as musyawarah have a long history in many villages, these
principles were rarely applied to decision-making over village budgets. Within villages, top-
down decision-making over development resources was the norm. KDP brings with it decision-
making processes that encourage participation across the village and a meritocratic basis for
distributing resources.
29
Figure 3.4: Tensions Generated by Competition in KDP
Tensions within villages over competition were less likely in Year One of KDP than in Year Two.
This is the case for two reasons. First, KDP is a process rather than an event; in areas
where democratization was less advanced, it would often take more than one year for villagers
to realize that they could truly put forward proposals of their choosing. In Year Two, with better
understanding of how KDP principles worked, villagers undertook the competition process with
vigor, compared to the first year when village elites were more able to dominate the process.
Hence, tensions within villages tended to be higher in Year Two.59 Yet these tensions did not
continue to escalate. As program processes were better understood, villagers also, in general,
understood that the primary basis of competition was need, rather than equal distribution of
money across hamlets over time. This concept made intuitive sense to most of the villagers we
spoke with. As such, as understanding of KDP increased, and the decision-making cycle was
repeated, tensions within villages declined (see the blue line in Figure 3.4).
“The tensions which emerged in the village meetings in KDP Year 1 were not like those in the third year [KDP Year
3]…. In the first year, it was more certain which proposals would get funding, while there were proposals from
different hamlets, the community did not feel there was competition … Later, when each proposal was discussed in
the forum, it appeared that several individuals forced the proposals from their hamlet to be put forward [to the inter-
village meeting]. However, after the session, there were only two proposals left [to decide between]. These two
proposals caused tensions. Each supporter wanted their proposal to be prioritized. This was to such an extent that
there were exchanges of foul words. Then there was agreement from the forum to determine the priority by taking a
vote, and in the end all the attendees agreed … Outside of the forum there were still tensions, but, yeah, these were
limited to complaints and grumblings … But the village officials [and the community leaders] and I immediately
approached them. We went directly to their houses.”
Villager, Sumedangan, Pademawu, Pamekasan (1221)
In contrast, inter-village tensions related to KDP were low in Year One and tended to increase over
time, with tensions highest in Year Three. In Year One, tensions were relatively low, in large
part because true competition, in many areas, was absent. In both Manggarai and Pamekasan
59
In villages where participatory mechanisms were more foreign, or where socialization was poor, tensions
continued to rise.
30
districts, which villages would get projects was pre-decided (illegally) by village heads and the
program’s sub-district facilitator (see Box 3.1 above). Further, even when rules were kept to,
program rules dictated that some villages would be eligible in Year One and others in Year Two.60
In Year Two, tensions were higher as villages, and their representatives, gained a better understanding
of program rules. However, in contrast to intra-village tensions, tensions continued to rise in later
years of the program. Tensions were significant in later years in the inter-village forums when the
same villages or specific proposals repeatedly missed out on funding. This process can take place
within villages, as demonstrated in the Slahung case, above. However, it is more likely to take
place between villages, in large part because the same ethos of solidarity and support does not, in
general, exist across villages as it does across hamlets within the same village. A perception emerged
in some circles that certain villages (and in some cases certain hamlets) were constantly being
prioritized over others. In the inter-village forums, in the early years, there was always the hope
that if a proposal was not funded in a particular year, it could be forwarded again in the following
year. This became less likely as the program came to a close and the same villages (for varied
technical, needs-based, and even political reasons) missed out (shown in the red line in Figure 3.4).
It is an open question whether tensions between villages would continue to increase if the
program stayed in sub-districts for more than three years. Our sample of four-year sub-districts
is too low (two) to draw general conclusions, especially given that these sub-districts were
“reward” ones, with an extra year of the program given for good performance in the past.61 In
principle, with more rounds of KDP, villages should understand even better the basis of
competition; it is also more likely that each village will receive funding at some point. These
processes would work towards limiting inter-village tensions relating to competition within
the program. However, on the other hand, if certain villages manage to “capture” the process—
something that did occur in some of our research sub-districts—it is likely that tensions would
continue to grow.
This backs up the basic point made above: competition related to KDP does not in itself cause
significant problems. Villagers, individually and collectively, generally learn the program rules
and, over time, tend to accept them, even like them. KDP forums and/or facilitators are almost
always able to address directly (and effectively) any tensions relating to individuals or villages
not having their proposals funded. However, where the program malfunctions, or where large
power differences and processes of explicit political contestation over development resources
remain, competition does in certain instances spark larger disputes.
2.2 When Things Go Wrong: Program Malfunction Conflict
Where the program functions as it is meant to (i.e., where socialization, decision-making,

implementation, etc. goes as planned), competition does not result in serious conflict. As we
discuss in greater depth below, the in-built conflict management mechanisms work well in limiting
conflict related to competition. However, in cases where KDP does not operate as it is meant to,
60
See footnote 56 above.
61
The decision on which kecamatan should be rewarded with a fourth year of KDP is made by the Regional
Management Unit (RMU) in consultation with the district level consultants. Across Indonesia, 48 kecamatan
received a fourth year of KDP-I.
31
the potential is there for the program to lead to larger conflicts. That said, we found only one
such case that turned violent.
Program malfunction can stem from problems of omission or commission.62 The first concerns
instances where the program did not function properly because of inadequate socialization, or
transparency of the rules and processes of the program. This can either be intentional, as when
rumors or misinformation are spread, or unintentional. Either way, when individuals or groups did
not properly understand how the program works, and why certain decisions were made, the likelihood
of conflict increased. Second, in other cases program malfunction was a product of deliberate
malfeasance, most commonly corruption—by program staff, local government, or actors otherwise
involved in the KDP process. In these cases, too, the program was more likely to lead to conflict.
In some instances, deviations were not picked up at the time, stimulating grievances that later
escalated; in others, intentional program deviations were picked up at the time, with the responses
(of communities or program staff) to these acts often triggering conflict.
Problems of Omission: Poor Socialization and Implementation

The evidence from the field demonstrates that poor socialization and implementation can lead to
disputes over the allocation of resources or the implementation of the program. In such cases, no
deliberate act of wrongdoing occurs. However, poor performance from facilitators or program
staff—what we deem “problems of omission”—results not only in sub-standard program
performance, but also negative spillovers in terms of increased local tensions. In many cases, poor
socialization allows for rumors and the spread of misinformation, even when the program is being
implemented by the book. These are typically minor disputes, but in some instances they escalated
into more widespread conflicts. In such cases, if there had been better socialization, the problems
would most likely have been solved early on. The cases demonstrate that following the rules is not
enough for KDP facilitators. Villagers often have a lack of experience in democratic decision-
making, and entrenched interests often work against the forces of transparency and participation
that KDP promotes. It is thus necessary for facilitators to work intensely and creatively to ensure
that target communities understand properly KDP’s aims and procedures.
Take, for example, a case from Madura, East Java. The village facilitator thought he had been
fired by the Village Head (referred to in Madura as the Klebun). His term had actually come to
an end in the KDP cycle, but lack of information regarding program rules meant that this was not
clear for the facilitator. Consequently, the village facilitator shifted his allegiance to the Klebun’s
opponent, and this played out in later conflicts in the village (see discussions below of the intersection
between KDP and local power battles).
62
A sub-set of this type of tension could result from (what might be called) ‘rational resistance’ on the part of
participants, which could take two forms: (a) villagers fully understand KDP’s rules and procedures, and appear to
comply with the outcomes of its competitive bidding mechanisms and inclusion requirements, but nonetheless use
the occasions and resources provided by KDP to express, actively or passively (a la ‘weapons of the weak’ (Scott
1985)), their more general dissatisfaction with external development interventions and the changes/requirements it
makes of them; (b) villagers may again fully understand project rules, but not follow them because to do so would
knowingly lead them into direct conflict with powerful local elites. (Thanks to David Mosse for reminding us of this
point.) This is an issue we will address in more detail in subsequent work.
32
“I was only FD for five months… I was terminated by the Klebun without clear reason. Without saying thank you… I was
pushed aside because the Klebun was afraid I would straighten out the road [criticize the Klebun’s flaws]. I was considered
to know a lot about the matter of the assistance, so I had to be pushed aside. Perhaps the Klebun was worried that later
I would make trouble over the matter of the way the money was used.”
Former Village Facilitator (FD), Panagguan, Proppo, Pamekasan (750)
In the example above, there was no deviation in program rules, but poor socialization led to
confusion, misperceptions, and misinformation that later fueled other village conflicts. A similar
process occurred in another village in Pamekasan:
“… Rumors appeared saying that the Lawangan Daya savings and loans funds had been used for the personal interests
of the TPK (Village Project Management Team), Lurah (Village Head) and LKMD (Village Community Resilience Council)
organizers. In fact, people said that the program money was used to build the LKMD Chair’s house, where as I know
myself that Pak Jamat [the LKMD Chair] built the house from the proceeds of selling his own car ... These are just the
wrong views of people who are envious of the success of Lawangan Daya in implementing the saving-borrowing
enterprise… The program money has now reached Rp. 97 million from Rp. 60 million …”
TPK Chair, Lawangan Daya, Pademawu, Pamekasan (1233)
Indeed, poor explanation of program rules was a common cause of tension. A problem arose in
Satar Punda village, in Manggarai, Flores, when the village facilitator’s explanation of KDP
mechanisms differed to those applied in the field. At first the facilitator made a list of about fifty
households in Pering sub-village from two different groups and explained that both would
receive funding. However, only 22 households from the kelompok Bawang Merah (Red Garlic
group)63 received funding, while none of the members of kelompok Mangga Udang (Mango
Shrimp group) received money. This, again, resulted in tensions within the village.
Disputes relating to poor implementation emerged not only during the competition process, where
the funds are allocated, but also surfaced during the implementation phase in the field. Often
these were a result of poor socialization in the initial project cycle. Four examples below from
the different research districts demonstrate how these conflicts emerged because of poor
socialization and implementation, and also demonstrate the potential broader consequences.
Water Pipes in Sikka

The first case pertains to a clean water project in Magepanda village, Sikka. The work which
took place in the field was not consistent with that planned for in the RAB (draft budget). As a
result, not all of the pipes were laid and tensions within the village elite mounted:
“There was a tug-of-war going on between the government and the project manager. This was because they should
have used steel pipes, but they used PVC pipes instead. Apparently, there weren’t enough pipes. They have started
to ask for contributions from the community, but the community is still asking for the report from the project manager,
the Head of the LKMD. The villagers are willing to contribute, but first the report must be clear.”
Former Village Head and Current Member of the BPD, Magepanda, Nita, Sikka (S.629)
In addition, there was a large discrepancy between the length of the pipes in the draft budget and
the length of the pipes that had already been laid in the field. According to the approved funding
proposal, 3.5 km of pipes had to be laid to complete the project. However, in reality 5.8 km of
were needed to cover the distance. The problems were in part due to poor technical planning.
63
This name derives from the fact that the savings and loans scheme was used to grow red onions. Another group was
called Bawang Putih (white garlic). Villager, Satar Punda, Lambaleda, Manggarai (M.514:2)
33
Another contributing factor, however, was that the initial proposal was not particularly clear. As
such, the project manager, the Village Head, the head of the BPD (Village Representative Board),
and the head of the TPK (Village Project Management Team), all interpreted the proposal in
different ways. This led to tensions, as each accused the others of corruption or deception:
“The clean water project in Magepanda has yet to be finished. There are still parties who are blaming each other and
trying to obstruct the construction process. The KDP implementers in the village have already been selected, but
they have no function.”
Former Village Head, Magepanda, Nita, Sikka (S.650)
Road Construction in Manggarai

A second case concerns a dispute over a 2.5km stretch of the Bea Mese–Lando road. Tensions
erupted when the proposal from Lando village for KDP to re-surface the road was accepted. The
road re-surfacing began as planned with the first 150m of road completed. However, 500m of
the remaining road to be re-surfaced lay within Bea Mese village’s administrative jurisdiction.
The community there, spurred on by their leaders, started to sabotage the stretch already completed,
pushing the stones to the side of the road. They complained that they should be involved in the
laying of the stones for the road in their area.64 Meanwhile, the villagers from Lando village did
not agree because they had already followed the correct proposal submission procedures, including
the procedures at the meetings at the sub-district level. Why had the villagers from Bea Mese not
objected to the project at the inter-village meeting?
Again, the primary reason for the problem was poor socialization of program rules, this time at
the sub-district level. With tensions mounting, conflict was only avoided after the intervention
of the Camat (Sub-district Head), the Head of the UPK (KDP Financial Management Unit), and
the police. The Camat invited the village heads, the Head of the LKMD, respected community
figures, and the KDP staff from the two villages to a meeting to discuss the problem. After
discussion, they agreed that the Bea Mese community would be given Rp. 12 million to resurface
the 500m stretch of road located within the administrative boundaries of Desa Bea Mese.65
Bridge Building in Ponorogo

The third case in Slahung, Ponorogo, differs, in that the dispute was the result of effective
socialization processes that ensured that villagers understood when a facilitator had deviated
from program rules. While on the one hand this triggered tensions, on the other the resulting
dispute demonstrated accountability and adherence to program rules. Slahung village received
funds to build a new bridge connecting two hamlets after a proper competitive process. The
project became controversial, however, when the District Consultant managing the program
viewed the site and decided that the proposed design should be slightly altered. While the
decision was made on technical grounds, the design was changed and a foundation was quickly
built without consultation with the community, which had already had their project and its design
approved. Moreover, the changed design meant that a small section of the river bank cultivated
by a local farmer would be lost to the bridge and the road leading up to it. This farmer’s son
was offended that his family was not consulted, and that their permission was not sought for access
to or use of the land; the community was also angered by the lack of consultation:
64
Female Teacher, Lando, Cibal, Manggarai (M.1012).
65
Villager, Lando, Cibal, Manggarai (M.1013); Village Head, Lando, Cibal, Manggarai (M.1014); Village Head,
Lando, Cibal, Manggarai (M.1015).
34
“The conflict over the bridge arose due to a shift in the bridge location, so while the proposal was designed so that the bridge
went straight ahead, it was shifted to the south a little… the changes were not a result of consultation with the community…
The community then started questioning why the bridge was shifted, thinking that the result would not be good, let alone the
fact that there was land owned by one citizen that would be negatively affected…”
Ex-TPK Chair, Slahung, Slahung, Ponorogo (1342)
The newly formed BPD (Village Representative Council) initially became involved by advocating
for the land user, supporting his appeals to the DPRD (provincial parliament), and by attracting
media attention to the case. The BPD also assisted in resolving community tension over the design
change by facilitating a meeting where KDP implementers spoke about the reasons for the design
change and explained elements of the KDP procedure, while the community was able to air their
grievances. The issue was put to a vote and it was decided that the planned bridge would revert to
its original design. Members of the district parliament and executive, the security forces, KDP
facilitators, and the media all attended the meeting. While the community was satisfied, the labor
and materials that had gone into the construction of the foundation were wasted and villagers were
required to make up the shortfall themselves.
Micro-credit in Ponorogo
In a final example, poor socialization and implementation of KDP led to an ongoing conflict,
failed program outcomes, and wasted resources. Members of a KDP savings and loans group in
Panjeng village, Ponorogo, had borrowed money from the program for a cattle fattening initiative
and failed to pay back the funds. However, deviations from program rules, poor socialization, and
poor verification of the proposals put forward by the groups occurred. Most of the verification
team did not live in the sub-district and had no local knowledge of the relative success, or lack
thereof, of local cattle fattening initiatives and market trends. Hence, the money was not always
used for its intended purpose:
“… For example when locals wanted to borrow money for cattle-fattening initiatives. The sub-district official would first
check the site [to verify] whether the borrower really had cattle and had a cattle stall. When the site was checked there
was indeed a stall and cattle. Perhaps the official considered that the inhabitants had experience in raising cattle, so the
funds flowed. But in fact the cattle were not his, and nor was the stall. They were just his neighbor’s cattle that he claimed
to own, and the cattle stall too belonged to someone else. … The sub-district official came to the village, the term is to
check the reality in the field. In fact he was tricked by the community [laughing].”
Chair II of Farmers’ Group Margo Makmur, Panjeng, Jenangan, Ponorogo (1322)
“… There was always someone who owed money … Sometimes they didn’t understand that the money really had to be
returned, that it hadn’t just been given [to them]. …. The worse debts at the time were those of that cattle farmers’
group… Actually profits were made from the first sale. Then when the second cattle auction came round, what do you
know, the prices had dropped. They were only enough to cover the buying price and costs of cattle feed, but the costs
of rearing the livestock, well, were not covered. After that the prices dropped even further and the farmers couldn’t
return their loans…”
Treasurer, KSP (Saving and Loan Group), Panjeng, Jenangan, Ponorogo (1314)
“Actually the three monthly repayments would not have been that burdensome if they were just interest repayments, but it
was really difficult because the principal also had to be repaid. … In my experience every three months the cattle had
to be sold because there was no money to pay off the principal. There was no guarantee that within that three month
period we’d make a profit because cattle prices weren’t stable. In fact after a while my capital was used up because
I hadn’t made any profit over three months… What I did was to buy cattle, sell them after three months to make the
interest and loan principal repayments, and use the remaining money to buy more cattle. That’s how I used up my
capital through this method.”
35
As members of the savings and loans groups stopped making repayments, criticisms also emerged
that the repayment system being implemented in the Panjeng village was not in accordance with
program rules and was costing the group members more in interest repayments. Villagers made
comparisons with neighboring villages’ interests systems and began to distrust program facilitators,
particularly in the Financial Management Unit (UPK), and instead asked the male FD (Village
Facilitator) to intervene. Various parties tried to get the group members to make their repayments,
but they simply had no money to do so. To this day, tensions still exist between program beneficiaries
and five group members who have still not paid back their loans.
In this case, the drop in cattle prices was not the cause of the conflict; it merely triggered the
dispute over failed repayments. Poor socialization and implementation of the program stimulated
grievances which then surfaced when the match was thrown into the kindling.
Problems of Commission: Intentional Deviations and Corruption-Related Disputes

The cases above show the consequences of unintended poor socialization and/or implementation.
In these cases, facilitators or other program staff did not deliberately set out to subvert the program.
However, in other cases program malfunction occurs as a result of intentional deviations from
program procedures. These cases are more serious than those above, none of which resulted in
violence or escalated beyond the level of a small dispute (with the partial exception of the bridge
case). In particular, cases related to corruption, which are by no means rare, can sow seeds that in
the right conditions grow into fully-formed conflicts. This can have negative outcomes if perpetrators
are caught out but there is no means for recourse. They can, however, also have positive outcomes
in terms of increasing demands for accountability and transparency, thus increasing the likelihood
that the needs of beneficiaries will be met.
Corruption and KDP

In both provinces, numerous cases of corruption of KDP funds led to tensions that were significantly
larger than those relating to competition within the program itself. Often corruption works hand in
hand with elite capture of program processes (see Box 3.3).
Such cases were by no means rare in the villages we visited, particularly in Manggarai and
Pamekasan, our two ‘low capacity’ districts.66 In Lambaleda sub-district, Manggarai, KDP was
suspended after large-scale corruption by the sub-district facilitator (FK). It was discovered that
10% of the budget (Rp. 59,897,200) had not been distributed. The FK claimed he would
distribute the funds at a later date, but this never eventuated and the funds were never seen again.
In total, he is believed to have corrupted approximately Rp. 80 million in KDP funds. However,
66
Indeed, corruption is a common problem within KDP, as it is with other development programs. A study of corrup-
tion in KDP road projects, which looked at the effect of various “interventions” such as increased probability of audit-
ing, found that in control locations—that is, locations without any extra audits—29% of funds were lost to corruption
(Olken 2005). These figures are high, although it should be noted that this was 29% of micro-project funds, rather than
of the project’s budget. However, this total is still probably significantly lower than for other projects without partici-
patory mechanisms (Guggenheim 2006). For example, Mallaby (2004) notes that a similar level (approximately 30%)
of World Bank loans in Indonesia, most of which were of a top-down nature, were corrupted in the years leading up to
the financial crisis. Indeed, the fact that corruption rises to the surface in KDP can be seen as a positive sign; it means
that redress can be pursued and this may have deterrent effects for the future.
36
while he and the program have been suspended, he has not yet been punished in the courts and the
funds have not been returned. 67
Box 3.3: Examples of Village Level Corruption in Pamekasan, East Java
“The proposal was determined by the Village Head in the Musbangdes forum [village-level KDP meeting]. Beforehand
each hamlet had been asked to submit their proposal but we spread the word that the proposals would not necessarily
be agreed to. And, if they weren’t agreed to, the hamlets shouldn’t be disappointed … There was a fierce debate in the
forum, and then subsequently each hamlet submitted their final decision to the Village Head. Procedurally there is no
difference in decision-making between KDP and what is usually done by village communities, all the processes are
decided through deliberations although the final decision rests in the hands of the Village Head ... What’s confusing was
that these people asked for dana siluman [a term when the destination of outgoing funds is unclear, in this context
meaning that the funds are being used as illegal fees for the kecamatan staff] of 3-5% of the total assistance received
by the villages, as was the case for all the villages in Pademawu, according to instructions from the District Government.
But when I wrote down the allocation of money that I give to the kecamatan people as a fee for District Government staff
who came to the village, they were angry and asked me to change it, to charge it to the purchase of materials with a price
slightly higher that the actual price…”
Former Village Facilitator (FD) Year Two, Padegelan, Pademawu, Pamekasan (1214)
“… Don’t be surprised if the Village Head gets involved in minor acts of corruption of development program money which
enters the village. Frankly, I’ve also been involved because, if you’re not like that, how is the Village Head meant to get
enough money to pay for all their activities? In my opinion, minor acts of corruption are natural—some of the money is
used to help the community anyway… The officials at the top are involved in corruption, let alone those at the bottom.”
Village Head, Panagguan, Proppo, Pamekasan (756)
The Village Head from Panagguan argued that corruption is “natural” and that it often indirectly
benefits the community (Box 3.3). Neither point is valid. Indeed, we had many more informants
who were outraged by corruption than who thought it helped the poor (orang miskin). For our
purposes here, however, it is most disturbing to note how such instances of corruption can lead
to conflicts. In some cases, where corruptors are exposed, the long-term effects may be positive;
in others, corruption emerges only later, resulting in grievances that can spillover into broader
conflicts (Box 3.4).
67
Chair of LKMD, Golo Mangung, Lambaleda, Manggarai (M.545).
37
Box 3.4: Power Takes Precedence: Four Million Rupiah Vanishes
KDP arrived in Satar Punda village in Manggarai in 1998. Local facilitators were elected and trained, and a Village
Project Management Team (TPK) was formed. Rafael Hommo was selected as its head, Alexius Djomnag (the village
secretary) took the role of TPK secretary, and Simplisius Derry was chosen as treasurer. Things proceeded
smoothly, with funds distributed for productive economic activities.
One day, however, Rafael arrived at Simplisius’ house, asking to borrow KDP funds so that he could start a honey
business in Java. Simplisius turned him down, but after constant pressure from Rafael, who argued that as TPK head he
had authority to allocate KDP funds, he gave in. Simplisius agreed to lend Rp. 4 million, and Rafael signed a receipt
dated 30th October, 1999. Simplisius wanted to go to the nearest town to obtain an official stamp for the transaction, but
Rafael refused, arguing that he was not a kid and would return the money one month later. Feeling guilty about lending
program money, Simplisius reported what had happened to the sub-district facilitator (FK)
In late January 2000, Rafael returned from his business trip to Java. Angry that Simplisius had told the FK, he started
abusing him, shouting “La’e Acume” (dog’s penis in the local language). Simplisius returned to his house, feeling offended.
Events took a turn for the worse when the Itwilkab (District Investigations Office) arrived to carry out an investigation.
However, their focus was on Simplisius rather than Rafael; they insulted the treasurer, saying that he should make sure his
kids got a good education so they didn’t end up as stupid as him, while they were polite to Rafael. Simplisius became
extremely angry and went home to fetch his machete: “I want to kill them both [Rafael and the head of the investigations
team].” Only intervention from his wife, made him back down. Rafael moved to Java with his wife and children and has still
to return or pay the money back.
In this case, despite threats of violence, no-one was physically hurt. However, in another case
in Flores Timur district—the only violent KDP case in the areas covered by the newspaper
survey—corruption resulted in injuries to two people (Box 3.5).
Box 3.5: Corruption and Revenge in Flores Timur

Latonliwo village, Tanjung Bunga sub-district, Flores Timur district
The Head of the LKMD has stolen KDP funds. The village had been awarded Rp. 49,834,150 for an 850 meter long
drainage system. The construction goes ahead. But after only 413 meters are built, and only Rp. 10,150,000 has
been spent, the money has run out. Most of the remaining money is still with the LKMD head, except for Rp. 1 million
that he has lent to a friend. This is not the first time money has gone missing from a village project in Tanjung Bunga.
Previously funds to build twelve rainwater collection tanks went missing with only 44% of the awarded money spent
on the tanks. The KDP technical consultant at the district level (KMKab) knew what had been going on and decided
to take action, bringing the case to the attention of the public. One night the LKMD Head brought three friends—a
Sub-village Head (Kepala Dusun), Head of the TPK (Implementation team), and his son—to visit the KMKab and his
friend, a local community leader. When there, they started to beat them in revenge for their temerity in highlighting
their corruption and in order to cover up the accusation. Escaping, however, with just some bruises, the KMKab
reported the case to the police.
Reported in Dian, February 7, 2002
The Importance of Socialization and Links to Conflict

Analysis of the cases above shows the intrinsic and complex relationship between socialization
and conflict. In general, the relationship is positive: increased knowledge of the rules, processes
and aims of KDP tends to limit the number of program malfunction conflicts. Socialization is
particularly important for a number of reasons.
First, effective socialization ensures that program participants and beneficiaries understand the
intentions of the program and how the program will be conducted and implemented. Where the
38
aims and/or decision-making mechanisms for development projects are not clear, individuals or
groups will not see the project processes or outcomes as being “fair”. This, as we explore in
more depth below, can provide multiple bases for conflict. Rumors, misinformation, political
maneuvering, and exclusion of groups with mandated rights are all, in themselves, triggers and
sources of conflict. Indeed, this is particularly important for participatory programs, where the
number of people involved in decision-making and implementation is much higher than in more
centralized or pre-determined projects.
Second, socialization, both for program facilitators and beneficiaries, is intrinsically linked to
good program implementation. Of course, poor implementation can be purely the result of human
error. However, in many of the cases above (and more often than not), poor implementation was
a result of a lack of understanding. Poor socialization can thus lead to conflicts at the competition
or implementation stages.
Third, socialization, in improving accountability and transparency, can allow for ongoing
bottom-up monitoring which can prevent program malfunctions from occurring, and grievances
from building. With adequate socialization, when deviations occur, official monitors and
program beneficiaries are aware of the channels for recourse and upholding accountability
and transparency.
As such, effective socialization can prevent many conflicts that emerge from problems of either
omission or commission. Good socialization can result in the early discovery of malfunctions
(intentional or unintentional), thereby preventing grievances from forming. However, the cases
also show that effective socialization can trigger conflicts. In particular, in many of the corruption
cases, it was the knowledge of how processes should work that led to tensions. Indeed, in the
sole violent KDP example, this was the case: the exposure of corruption led to injuries to two
people. Empowering villagers by building countervailing forces and power (Gibson and Woolcock
2005), may lead to conflict in the short-run. Yet, over the longer run, such conflicts can be
positive. The case went to court, and, arguably, future cases of corruption are less likely. As
such, an aim of KDP should not be to prevent all conflict from arising; rather, the program should
ensure that where conflict related to program malfunction does occur, it stimulates future
improvements in the program.
2.3 The Link between KDP and Existing Tensions: Interaction Conflict
Development programs and projects can “do harm” not only by directly triggering conflicts but
also by feeding into existing ones. Projects operate in social contexts in which power relations are
constantly being negotiated, and development projects constitute a vital resource that can be utilized
by different actors in this process. Interaction conflicts occur when development projects (KDP or
others) interact with pre-existing local tensions, power structures and conflicts, triggering conflict
escalation and, in some cases, violence.
This can happen in a number of ways. The money associated with projects, which is often vast
compared to existing village budgets, can be tapped personally or for the benefit of a person’s
family or kinship group. It can also provide a basis for the strengthening or extension of systems of
patronage, with development resources being used by elites to buy the support of particular
39
individuals or groups in order to build their (the elites’) own power. Examples of this were
provided in the previous section of this chapter. Positions, such as those of village program
facilitator or on the implementation team, can also be the subject of political competition. In
such cases, KDP becomes a resource, material and/or symbolic, that is fought over.
Individuals in our research villages often used the financial resources, legitimacy, and access to
other forms of power that comes with having “won” a development project as a means for
strengthening their position and building their support base. The data revealed a number of
conflicts that involved the interaction of KDP with pre-existing forms of competition. In such
conflicts, KDP cannot be said to have been the primary cause; however, the program was a
trigger that contributed to the escalation of tensions. In order to ensure that programs such as
KDP do not have negative impacts on local conflict, it is necessary to consider the ways in which
the program interacts with the existing sets of power relations in the villages in which it operates.
A number of different forms of interaction conflict can be identified in our data:

Elite political power battles (e.g., village head elections)
Control over development resources
Resistance from elites to democratization
Power Battles between Elites and KDP

This was the most common form of interaction conflict. In almost every village in our sample,
there were cases where development projects had become ensnared in local power battles between
different elites within the village. In one sub-district in Pamekasan, Rp. 1 million was taken from
the budget for all the villages that were successful in obtaining KDP funding. The money was put
towards the campaign budget of one of the dominant political parties for the 2004 legislative
elections. A case concerning conflict over the development of water resources in East Java
intersecting with the local village head elections provides another prime example. In Panagguan
village in Pamekasan, a KDP proposal to provide water facilities in one village was used as a
political tool. The Village Head dominated the KDP proposal process to ensure that one of the
hamlets where he was seeking support in the elections (but which at the time supported his
opponent) gained access to a water pipeline and hence gave him their patronage.
“At the time, much of the Danglebar hamlet community did not support the Klebun (Village Head) … leading up to the
village head election, the Klebun turned the water in Danglebar back on [through a KDP funded project] by connecting
PVC pipes there to channel water. In this way, the Klebun obtained the support of the people of Danglebar in the
village head election.”
Villager, Panagguan, Proppo, Pamekasan (742)
“I prioritized the clean water project in Dusun Danglebar so that their support would come to me [meaning that the
citizens of that dusun would support him in the election]. I had already calculated it many days beforehand. Many
villagers of the dusun tended to support my opponent.”
Village Head, Panagguan, Proppo, Pamekasan (756)
This led to a heated conflict between election candidates, and the eventual damage of the water-
pump storage facility. Implementation of the project was delayed by the Village Head until after
the election to ensure the hamlet in question voted for him. Such elite capture resulted in a swing
in the vote in the target hamlet in his favor, as well as the creation of a new policy by the Village
Head that the beneficiaries had to pay for access to the water, which should have been freely
40
provided as a part of the KDP proposal. The village ultimately suffered sanctions from the program
for the delay in implementation, and was not able to compete for funds the following year.
Similarly, in another case in Pamekasan, preexisting tensions between elites played themselves out
through the attempted manipulation of KDP (Box 3.6). However, in this case, the Village Head had
encouraged transparent processes throughout program implementation. As such, good socialization
meant that the conflicts came to a head fairly quickly; with the help of local leaders (both formal
and informal), the problems were efficiently and effectively resolved.
Box 3.6 Road Surfacing Project - Desa Tattangoh, KDP-I Year 3 (2000)
Conflict and social tension in Tattangoh village in Pamekasan arose over three related issues during the third year of KDP
in that village. All were colored by the general rivalry between two different power blocs within the village, led by the Village
Head and the BPD (Village Representative Council) respectively.
The first problem arose when the Village Head appointed the TPK (Implementation Team) without going through the required
selection process. One villager in particular was disappointed at being passed over for a position on the TPK. The second
problem occurred between the disappointed villager and the appointed TPK Head, who both managed rival crews working
on the KDP road surfacing project. The TPK Head declared a certain day to be a holiday, but on that day the TPK Head’s
crew began work as usual. The disappointed villager was so incensed at the TPK Head’s deceit, which deprived him and
his men of payment for a day’s work, he threatened carok (ritual revenge dueling) on two separate occasions but was
prevented from carrying out his threats both times by local leaders and the Village Head.
The third matter was over the sourcing of the stones used in the road surfacing. Accusations that the stones were not being
bought fairly from local quarrymen led to a blockade of the work site, organized by another villager linked with the disappointed
villager and identified as a member of opposition to the village government. Negotiations between parties mediated by the
Village Head and other community leaders successfully resolved the confrontation.
Development Resources and KDP

The cases above show how groups seek to capture development resources in order to bolster their
position within the community. Access to development resources sustains, and in cases expands,
systems of patronage. However, in other examples the motivation for elites to capture development
resources is simpler: they want the resources themselves. Take, for example, a case from Tengku
Leda village in Manggarai (Box 3.7).
Box 3.7: Village Elections in Desa Tengku Leda, Manggarai
KDP helped fuel tensions between the Village Head and others in Tengku Leda village, Manggarai. A friend of the Village
Head, Martinus Rudung, who had been a strong supporter in the village head election campaign, was appointed to a
position in the KDP management team, as were other friends.
Initially tensions emerged between those who had supported the winning candidate and those who had supported other
candidates for the village head position. When the village held a party to celebrate his success, there was some debate
over whether or not the supporters of the other candidates who ran for village head should be invited. Laurens Guntur, who
had supported another candidate, thought that he would be elected as the Head of the LKMD (Village Community Resilience
Council) and a technical consultant for KDP. However, he got neither of the positions.
After a while, however, the nature of tensions within the village changed. Having won the election, the village head took
control of KDP finances. At first his friends, who now constituted the local KDP management team, were fine with this. Yet
it soon became clear that the village head was not sharing information about the project finances with others on the team;
neither the treasurer (Domi Abraham) nor the KDP field manager (Martinus Rundung) knew anything about KDP finances.
Feelings arose that the village head was trying to capture the program himself. Tensions worsened when the money left
over at the end of the KDP financial year was not divided up (in the form of cement, sand and cash) evenly as was previously
agreed. This is what triggered the feelings of antipathy towards the Village Head.
Religious/Community Leader, Tengku Leda, Lambaleda, Manggarai (M. 1023)
41
In poor areas of Indonesia, where resources are extremely limited, many people seek access to
the state (in political positions or as civil servants) for self-enrichment. The salaries for these
positions are extremely low, in some cases not much higher than those for laborers or successful
farmers. Rather than remuneration, motivation for seeking state positions is often related to the
potential for bribes and/or for bringing back projects to the person’s village or kinship group.68
In Flores, competition for civil service positions has been so high that in cases riots have occurred
when people have not been accepted in the annual round of recruitment. In 2004, the central
government decided to take back responsibility for decisions over recruitment from the local
administrations in Flores because the system was perceived as being too corrupt. Large-scale
protests took place from candidates who had paid bribes to local officials for a “guaranteed” job.
Development projects offer opportunities for corruption, and resources to be directed in inequitable
directions.
The Resistance of Elites to Democratization

A third form of interaction conflict relates to resistance by elites to the normative frameworks
that projects such as KDP bring with them. KDP is a democratization project; it aims to promote
transparency and widespread participation in decision-making. Unsurprisingly, this is often met
with resistance by incumbent elites who want to maintain the power balance in their favor. The
control of decision-making over development resources is one of the ways in which elites maintain
positions of power, and hence elites are often extremely reluctant to cede this task to others.
Many Indonesian villages are still quasi-feudal ‘mini kingdoms’. In such instances, Village Heads
and traditional leaders see themselves as playing the role of benign dictators, where they look
after the welfare of the people and ensure their security. Leaving aside the question of whether
or not this is really in the best interests of “subjects” within the village, it is clear that KDP (and
indeed the reformasi movement) envisions a different model of social structure and control. The
post-1998 democratization, and the administrative decentralization that followed, created an
environment of flux. In many parts of rural Indonesia, this has led to tensions between the way
things were done in the past—where local village and sub-district elites held the purse strings
and were not held accountable by villagers—and new normative systems which emphasize
widespread autonomy over decision-making, and checks and balances on forms of political power.
Local culture and custom, and appeals to tradition (adat), have been utilized by elites seeking to
cling to power.69 Programs like KDP promote transparency and greater attention to community
demands, but tensions arise at this juncture of ‘new and old ways of seeing and doing’, a common
outcome of social change. The clash between two normative worlds, and the resistance of elites
to change, can and does lead to tensions.
Data was collected in the key informant survey on levels of elite involvement in KDP and in non-KDP
development projects. By comparing reported levels of elite involvement in non-KDP and KDP
projects, we can get a sense of the extent to which KDP changes elite roles in these types of projects.
As Table 3.2 shows, in some areas there is a big difference between elite involvement in KDP and
68
This was clear in research conducted on the 2004 local legislative elections in East Java, NTT, Maluku and Bali
(Barron, Nathan and Welsh 2005).
69
For a cross-national analysis of the use of culture symbolism as a mechanism for the maintenance of power, see
Wolf (1999).
42
other development programs. For example, in Pademawu, East Java, over twice as many informants
reported disproportionate elite roles in non-KDP projects (87.6 percent of informants)
as did in KDP (37.5). In other sub-districts there was less of a difference, with Badegan (at the
other extreme) reporting the same (high) levels of disproportionate elite involvement in both
KDP and non-KDP projects.
Table 3.2: Elite Involvement in KDP Compared to

Non-KDP Development Projects
Degree of Sub-District (non-KDP disproportionate

difference/change elite involvement-KDP disproportionate
elite involvement)
% citing difference in elite role Province (district)
High (>40%) Pademawu (87.6-37.5) 50.1% East Java (Pamekasan)
Lambaleda (60.0-20.0) 40.0% NTT (Manggarai)
Medium (20%-40%) Ruteng (62.6-25.0) 37.6% NTT (Manggarai)
Nita (54.1-16.7) 37.4% NTT (Sikka)
Pasaen (68.8-37.5) 31.3% East Java (Pamekasan)
Slahung (56.3-31.3) 25.0% East Java (Ponorogo)
Proppo (100.0-78.2) 21.8% East Java (Pamekasan)
Low (0%-20%) Talibura (50.1-31.3) 18.8% NTT (Manggarai)
Jenangan (18.8-6.3) 12.5% East Java (Ponorogo)
Paga (56.3-43.8) 12.5% NTT (Sikka)
Cibal (38.9-27.6) 11.3% NTT (Sikka)
Badegan (66.7-66.7) 0.0% East Java (Ponorogo)
Source: Key Informant Survey (treatment sites only)
The fieldwork showed that elites have the autonomy to capture the KDP process in the way they
do other projects; whether they do or not is a partial measure of KDP’s efficacy in providing
constraints on elite power. If elite roles are significantly different between KDP and non-KDP
projects, we can assume that KDP is creating incentives for compliance with more participatory
and open decision-making processes. In contrast, if the difference in elite involvement is minimal
(and especially if elite involvement in KDP and non-KDP projects is high) we can assume that
there is resistance at play; i.e., that elites develop strategies to maintain disproportionate
involvement in development decision-making and resource distribution.
How does this relate to the level of conflict triggered by KDP? Is conflict more likely to result
from compliance (where elite roles change significantly, and hence decision-making presumably
becomes more uncertain) or from resistance (where elites try to maintain their disproportionate
role)?
Table 3.3 shows how changes in elite involvement correlate with the reported extent to which
KDP forums triggered conflict. The results show a clear trend for East Java, with sub-districts
with little change in elite involvement considerably more likely to report conflict. This suggests
43
that where resistance to a changed role exists among elites, conflict is more likely. The evidence
from NTT is less clear.
Table 3.3: Change in Elite Involvement and Level of KDP-triggered Problems/Conflicts

Province Change in Elite
Involvement Conflict triggered
East Java High 10.5%
Medium 39.9%
Low 70.5%
NTT High 21.4%
Medium 15.3%
Low 22.3%
Thus KDP-related interaction conflict is more likely, at least in East Java, when elites actively
resist local democratization. A case from Madura (Box 3.8 below) shows the intersection between
elite interests, conflict, and development programs, and the problems that can result from elites
seeking to cling to power.
In the case there was little response to community demands for shifting the abattoir before it was
burnt down in frustration, and there was little consultation with the community about their desires
and needs during the process of rebuilding the abattoir. This conflict then transformed into one
of a power battle between elites to monopolize and benefit from the spoils of development
funding from the district government coffers: the use of development funds was unclear, certain
contractors were favored (with no transparent materials acquisition process), and many political
maneuverings went on behind closed doors. The result was that the abattoir is not being used,
the conflict has never been resolved, and public resources have been wasted. This is a common
story of the interaction between development programs and conflict, particularly in periods of
democratization as villagers become more vocal, make more demands, and learn from introduced
projects which promote transparency.
44
Box 3.8: Abattoir Destruction and Violence in Madura
At the end of 1998, an incredible stench once again wafted into the central district of the Banyupelle village. Blood,
faeces, and refuse from the slaughter of animals taking place at the centrally located abattoir festered and rotted,
eventually draining away into the river nearby. The river was used by the local people to bathe, wash, and for air
wudhu (ritual ablutions water) for sholat (ritual prayers).
“The problem of the abattoir waste had been reported countless times by the community … Because there
was no follow-up and because reformasi (the reform era) was in full swing … the people were brave enough
to destroy the abattoir.”
Veterinarian, Banyupelle, Palengaan, Pamekasan (67)
The community’s long-standing frustration with the impact of the slaughtering activities eventually intensified when a
comment was made by one of the Kyai (muslim clerics) in his sermons that the river water was najis (unclean) so that
it could not be used to purify oneself for daily prayers. The district government made few attempts to resolve the
problem and ignored the complaints of the villagers.
In 1998, a mob of about 500 people destroyed the abattoir, toppling the walls, prying up the floor and completely burning
down the roof. Only about ten percent of the abattoir floor remained. This incident resulted in the destruction of
public property and significant levels of community tension, and it triggered a series of elite level disputes.
Lacking an alternative site, the slaughterers continued killing their cattle in the ruins of the roofless abattoir, as much
for practical reasons as to make a political statement. The Village Head (Klebun), renowned for his links to criminal
groups, politically opposed the head of the League of Slaughterers group. While discussions about building a new
abattoir began, no-one could agree on a location. The Klebun made a non-consultative executive decision on the
new site, not far from his home, but a long way from the public services required by the slaughterers. Land was
purchased by the Klebun at a low price and sold to the district government for a profit. Construction of the new
abattoir began, corners were cut, and rumors circulated about corruption of funds. Uncertainty over the construction
process annoyed the slaughterers who felt excluded from the decision-making process.
In the year 2000, the community again expressed their discontent through the destruction of the remaining floor of the
old abattoir. The head of the slaughterers eventually built an abattoir on his property. While the community was
finally satisfied, the new public abattoir went unoccupied and elite tensions increased. Meetings were held to try and
resolve the problem, but the elites continued with their own agendas.
“The first meeting discussed the plan to occupy the new abattoir, the second was held after the slaughterers’ demands
that the abattoir floor be widened and electricity be supplied were met, whereas the last meeting was held to plan the
occupation (of the abattoir) after the demands for a roof were met. But even after that, they didn’t occupy it because
they had further demands for security. They asked for police to be on guard from six until the morning, after they had
cut up meat.”
Village Head, Banyupelle, Palengaan, Pamekasan (58)
Intrigues and blunders continued to occur in elite government and non-government circles. The District Office
followed its own agenda as did the Klebun. All the while, the slaughterers were carrying out their activities on private
land, resulting in a loss of taxation income and public resources being wasted on an empty abattoir. A final agreement
was reached to occupy the abattoir at a meeting hosted by a Kyai. Security forces, government staff, and the village
elite were present. However, following the meeting, the Kyai stated: “Um…the slaughterers don’t want to occupy it.
They say a Kyai was robbed [there]”. Despite assurances from the Klebun that he would guarantee the security of the
abattoir (for a small price, of course), the building remains unoccupied to this day. Extensive public resources were
wasted and the issue is still hot in the community to this day.
45
Limiting Interaction Conflict

In the field, we found that almost all development projects, including KDP, helped trigger interaction
conflict. However, they did not always trigger such conflict. Analysis of the cases above, as well as other
examples we found, provides some lessons as to when interaction conflicts are most likely to occur.
First, and most simply, interaction conflicts are most likely when conflict or social tensions of the
types noted above are already present. Where tensions between different elites within a village, for
example, are already high, there is an enhanced risk that projects such as KDP will get pulled into
the conflict. Where there are already issues related to land ownership or use, development projects
can easily make things worse. The basic implication is clear: before a program enters an area, it is
necessary to understand which forms of local social tensions/conflicts are present, so that program
staff can ensure that the program does not fuel conflict that is already latent.
Second, KDP interaction conflict was more likely in areas where existing value systems/cultural
norms ran contrary to the principles embodied in the program. Where different sets of rules and
norms clashed, conflict was more likely.70 In particular, where elites did not buy into the principles
of grassroots participation, transparency, etc., there was a greater likelihood that they would subvert
the program. The program, in turn, would then interact with other tensions on the ground. For
example, in Proppo sub-district, Pamekasan, KDP rules do not fit easily alongside local norms.71
The village heads in this region have hegemony over many areas of decision-making on village
administration, planning and development (Barnawi 2003). Patronage systems operate, with the
loyalty of villagers bought. This system of social organization does not sit easily alongside the
principles of democratization and equitable participation endorsed by KDP. It is not surprising that
there are more problems with the program in places like this than in other areas where there is less
resistance to the principles and practices KDP brings.
Third, where there was weak socialization, monitoring and implementation of the program, or
where responses to program deviations were inadequate, there was an enhanced risk that the
program could interact with existing issues. Poor program functioning, as we discussed earlier,
allows for grievances to build, which over time can erupt into conflict. Program malfunction
conflict is serious enough, but when it interacts with other tensions on the ground, there is the
potential for more serious unrest. Ensuring that the program functions as intended is the best
defense against interaction conflict.
3. Why Do KDP Conflicts Not Escalate?
As we have outlined above, there are a number of different forms of conflict that are related,
directly or indirectly, to KDP. However, in most cases tensions related to KDP do not escalate
into fully-fledged conflicts. Why is this so? There are two primary reasons why conflicts do not
tend to grow in size or become violent. First, KDP has in-built mechanisms in the form of
70
See Barron, Smith, and Woolcock (2004) for a discussion of how incoherence and incompatibilities with respect to
‘the rules of the game’ affects the likelihood of conflict.
71
As noted in Table 3.2 above, 78.3% of survey informants reported that elites played a disproportionately large role in
decision-making in KDP. For other development projects, the rate was 100%.
46
forums, facilitators and other complaints mechanisms that provide an outlet for tensions and
opportunities for redress. Second, the principles the program embodies (accountability,
transparency, participation) make it less likely that misunderstandings will occur. Conflicts
escalate only where the program does not function well.
3.1 Problem Solving in KDP Forums
One of the primary reasons that KDP conflicts do not tend to escalate is the availability of
forums, people and processes, where problems can be dealt with before they grow in size.
While KDP does trigger conflict, in general the tensions it does create were constrained by the
same forums and processes from which they emerged.
Box 3.9: KDP Forums and Meetings
As outlined in Box 2.1 above, the KDP process involves forums or meetings at the hamlet, village and kecamatan
(inter-village) level: the Penggalian Gaggasan, Musyawarah Desa and Musyawarah Antar Desa, respectively.
These meetings are held to elect staff, brainstorm and decide on proposals, to account for funds, and to deal
with specific problems that arise during implementation. Formally, KDP stipulates that approximately 20 meetings are
held, depending on the number of hamlets in each village (see Figure 3.3). However, villagers often decide extra
meetings are required; if problems do occur, in some cases as many as seven extra meetings have been held.
The meetings are open to all. Villagers are encouraged to join the discussion, debate the merits of proposed
projects, and participate in decision-making. In some villages, a “one person, one vote” system is used, with each
person placing a corn kernel in the envelope of their preferred project. Other villagers have introduced a weighted
voting system, where villagers can preference their votes by spreading their corn kernels across preferred projects.
From discussions with KDP team
Indeed, KDP forums are commonly used for dealing with KDP-related conflicts/problems: 92%
of survey respondents in East Java reported the use of KDP forums for addressing KDP-related
problems; 96% did so in NTT.72 Forums at the sub-district level were more likely to be used for
solving KDP conflicts than those at the village level (95% compared to 85%), reflecting the fact
that more conflicts emerge at this level.73 Further, respondents reported very high success rates
for the forums at solving KDP-related conflicts: 84.1% of informants in East Java reported that
KDP conflicts were successfully solved in KDP forums; 72.1% reported the same in NTT.74
In-built tensions—that is, forms of competition that are part of the KDP design—were more
likely to be addressed in forums than malfunctions associated with KDP (see Table 3.4). This
was particularly true in East Java, where only 21% and 11% of informants reported that forums
were used to deal with problems relating to KDP staff and corruption, respectively. There are a
number of reasons for this. First, as discussed below, competition-related tensions are much
easier to deal with than those relating to corruption or other malfunctions. There are few avenues
of recourse within forums when malfunctions occur, aside from the sanctions that can be enacted at
the end of the project cycle. Second, forums are held relatively infrequently following the allocation
of funding. Other processes and facilitators need to be accessed for problems arising from
implementation.
72
Sample sizes: East Java (n = 119); NTT (n = 130).
73
In East Java, 91.9% reported use at the village level, and 95% use at the sub-district level (sample sizes: 99 and
20). In NTT, 96.4% reported use at the village level, and 94.8% at the sub-district level (sample sizes: 111 and 19).
74
47
While village and kecamatan forums were equally likely to address in-built conflicts, kecamatan-
level forums were substantially more likely to deal with program malfunction conflicts than were
village forums.75 The qualitative fieldwork demonstrated the reasons for this. Villagers view the
sub-district level forums and facilitators as having greater authority to deal with problems with
the program’s functioning. Program staff members in the kecamatan are employees tasked with
running the program in a fair and appropriate way. At the village level, in contrast, facilitators
are in effect volunteers. Further, as noted earlier, power imbalances are less prominent in sub-district
than in village meetings. As a result, the likelihood that a particular group will challenge a program
process or outcome is higher at the sub-district level than in the village. Third, those who attend
sub-district meetings are more likely to understand how the program is meant to function, and
hence whether there has been a deviation from program rules, than those in the village.
Where forums were used to address KDP-related problems, they were more likely to be successful
at addressing in-built conflicts. As Wong (2003) has rightly noted, issues related to corruption in
KDP are generally more difficult to solve than problems relating to the inherent competition in the
KDP process.76 However, almost 70 percent of informants in East Java and almost 80 percent in
NTT still said that these problems were usually successfully resolved, although as noted above they
are addressed less often.77
Table 3.4: Types of KDP Problems Dealt with in KDP Forums and Success Rates
Specific Source of Type of Tension Dealt With in KDP Forums Successfully Resolved
Problem/Conflict (in most cases) (% respondents agreeing) in Forums
(% respondents agreeing)
Problems over understanding
the project In-built 88% 96%
Decisions on projects to go to In-built
the kecamatan forum or malfunction 87% 98%
Decisions made on funding at
the kecamatan forum In-built 72% 90%
Problems over decisions on In-built
procurement/implementation or malfunction 69% 92%
Problems over KDP
staff/facilitators Malfunction 39% 83%
Problems over corruption Malfunction 33% 74%
When combined with insights from the qualitative fieldwork, this suggests that while it may be
difficult to resolve protracted problems of this type, which have already risen up the project
bureaucracy, a large proportion of corruption cases are addressed early on (at the village and/or
kecamatan level), where they are normally able to be resolved satisfactorily.
75
Across the two provinces, as the kecamatan-level, 65% of informants said forums were used to deal with problems
over KDP staff/facilitators, compared to 12.1% at the village level. For corruption, the figures are 35% (kecamatan
level) and 6.1% (village level). Sample sizes: village level (n = 99); kecamatan-level (n = 20).
76
Wong (2003, p. 17) argues that it “was much more difficult [for the complaints handling unit in KDP] to resolve cases
related to the misuse of funds and corruption than those involving procedural violations”.
77
These figures are averaged across problems relating to corruption and to facilitation.
48
In East Java and NTT, survey respondents indicated that in-built problems were resolved simply
because there was no other place (East Java) or because it was the appropriate place (NTT).
Both these answers demonstrate the extent to which the forums are perceived to be a natural place
to address these types of problems; it also suggests that KDP’s in-built processes (and the
authority and legitimacy the program bestows upon KDP facilitators), in addition to KDP staff’s
personal attributes, help contribute to successful resolution.
The pattern is slightly different, and somewhat inconclusive, for KDP malfunction problems,
with the successful resolution of corruption problems (East Java) and other problems (NTT)
attributed to the role of KDP staff. This seems to suggest that these problems required significant
assistance from KDP staff and/or facilitators in getting forums to address these more difficult, and
often sensitive, problems. It is not hard to imagine that when forums address these problems,
which often implicate key KDP actors, a little extra facilitation is required to arrive at a successful
outcome. In some cases, outside actors are also required (see Section 3.4 of this chapter).
When KDP-related problems were not resolved in the forums, it was largely due to a lack of
facilitation in the meetings, rather than to norms that these problems should not be solved in
KDP forums. This suggests that whereas both the inherent legitimacy afforded the KDP institutions
and processes and the particular characteristics of facilitators are important in understanding why
different types of problems are taken to the KDP forums, a focus on improving facilitation and
socialization more broadly would help in resolving KDP-related problems.78
Generally, then, various types of KDP-related problems can be successfully addressed by the
forums. KDP explicitly aims to introduce conflict in the form of competition over resources. As
such, some conflict is inevitable. This conflict is not necessarily negative; conflict that is channeled
through appropriate institutions and mechanisms that prevent it taking violent form can be an
engine for progressive social change, helping to build the social and civic skills associated with
collective needs identification and democratic decision-making. Even conflict over KDP malfunction—
for example, in cases of corruption—can have positive effects if it is addressed, including increasing
rights consciousness and demands for transparency amongst citizens. The data shows that in a wide
range of locations (with varying governance and conflict management capacity), KDP is able to
effectively channel conflict related to the program and, in doing so, is able to resolve the problems
it inevitably creates.79
3.2 The Role of Facilitators
Another internal mechanism by which KDP conflicts are controlled at an early stage is the KDP
facilitation network (see Box 3.10). KDP facilitators are used outside of KDP forums to deal
with conflicts related to the program, although to a lesser extent, and with less success, than was
the case for forums.
78
See Barron, Clark, and Mawardi (2004, pp. 49-51) for a fuller analysis and breakdown of the survey data.
79
As noted earlier, this is backed up by the newspaper conflict datasets (see Barron and Sharpe 2005).
49
Box 3.10: The Network of KDP Facilitators
As Box 2.1 noted, there is a network of KDP facilitators whose role is to ensure, at various levels, that the project runs
smoothly and to assist with problems. At the village level, two FD (Village Facilitators) are elected (one must be a woman
and one a man). Their primary role is to introduce the project to villagers and local leaders and to help to organize hamlet
and village meetings where project ideas and priorities are brainstormed. Also at the village level is the TPK (Activities
Implementation Team), the members of which are democratically elected by villages. The TPK takes responsibility for
coordinating the project, including tasks such as buying material and organizing labor in the villages.
At the kecamatan level, two FK (Sub-District Facilitators) are appointed, one responsible for social participation
issues and one for technical concerns. Together they try to ensure the project in the kecamatan as a whole runs
smoothly. FKs usually come from outside the local area but live locally and spend a lot of their time traveling between
the villages receiving KDP. They also work closely with the local Camat (Sub-District Head) and their staff.
At the district level, the KMKab (District Manager), who is appointed, provides general co-ordination with little direct
involvement with village-level activities.
From discussions with KDP staff, and Government of Indonesia (2002)
52.3% of respondents in East Java reported that KDP-related problems/conflicts were resolved by
KDP facilitators outside of KDP forums; 59.7% of respondents in NTT reported the same.80
These rates are high, although lower than that for KDP forums reported above.
Kecamatan Facilitators (FK) and Village Facilitators (FD) were the most active staff involved
in resolving KDP-related conflicts outside of KDP forums (see Figure 3.5). The extent to which
results were similar across kecamatan suggests to some extent it is the institutional role of the
KDP facilitators, rather than their personal qualities alone, that facilitates their ability to resolve
disputes outside of KDP forums, although in East Java there was also a role for informal or
formal leaders outside the KDP process.81 The main reasons why KDP facilitators were reported
to have been used to address conflicts outside of KDP forums, and the reasons for their success,
were that they were believed to have the resources and authority to solve the matter at hand
and, to a lesser extent, because the facilitators offered assistance when the problem emerged.
Despite (or even because of) the fact that most FKs come from outside the kecamatan to which
they are assigned, they seem to have little trouble establishing authority to resolve KDP-related
conflicts.82 In both provinces, and especially in East Java, a large proportion of respondents
reported that non-KDP staff (‘others’) helped resolve KDP-related problems and conflicts.
80
Sample sizes: East Java (n = 132); NTT (n = 134). In East Java, all of the problems solved by KDP facilitators
related to KDP. In NTT, three informants said the problems solved by KDP facilitators were not related to KDP (see
further discussion in Chapter 4).
81
Most of the ‘others’ relate to formal and informal leaders who are not formally KDP staff. For example, 35
respondents (25 in East Java, 10 in NTT) mentioned the village head and village officials; and 28 (22 in East Java,
6 in NTT) mentioned community leaders. In terms of ‘others’ that were directly related to KDP, 22 respondents
mentioned the kecamatan-level KDP team (10 in East Java, 12 in NTT) and 4 (all in East Java) mentioned the
provincial level KDP team.
82
See further discussion, and breakdowns of the survey data, in Barron, Clark, and Mawardi (2004, pp. 64-66).
50
Figure 3.5: People who Helped Resolve KDP-Related Problems/Conflicts Outside KDP Forums
Educating villagers about the role of facilitators is important. KDP has program facilitators at
the village, sub-district, and district levels; they are obvious points of contact for the community
when program-related problems arise. At the village level, the facilitators are local people
with whom the community is usually familiar. In three of the six KDP-related cases followed
in Pamekasan, East Java, program facilitators or consultants initially attempted to address the
grievances of beneficiaries and the broader community involved in the conflicts. In some cases,
forums were held to address the problems. In Ponorogo, in all six cases, program facilitators
and forums were involved in attempting to resolve problems at the onset. This is not to say that
they were successful in all instances, because they were not in most of the large conflicts
followed. However, even at a minimal level, the disputants could identify authorities and
places to express their grievances, or, in some cases, these authorities recognized that there
was a problem and sought to resolve it before it escalated.
Consider, for example, one of the largest drawn-out conflicts in Ponorogo. The case, which
involved KDP and actors from all variety of groups in the village in Slahung, related to the
District Consultant’s decision to change the technical drawings and design of the KDP-funded
bridge without consulting the community (see Section 2.2 of this chapter, above). In this case,
upon identifying the grievances of the landowner and other project beneficiaries, the village and
sub-district facilitators approached the aggrieved to try to resolve the key issues. While these
efforts failed, they did signify the beginning of a number of institutionalized initiatives to resolve
the dispute. Follow-up from leaders and other institutions outside of the KDP process, in
collaboration with KDP facilitators, were also instrumental in resolving the dispute.
3.3 Other Complaints Mechanisms
Many conflicts triggered by development programs are the result of program malfunctions and
frustrations with processes, with no or weak avenues for recourse. KDP has internal mechanisms
to deal with these early on, making the escalation of the problems less likely.
51
As we have seen, development projects inevitably trigger tensions since they involve the allocation
of finite resources. In places like Indonesian villages, where there are problems with corruption,
decision-making processes, and elite capture, these tensions are likely to be prominent. It would
thus appear to be obvious that programs and projects must plan for ways to deal with the grievances
and problems that inevitably arise. The fieldwork in East Java and NTT, however, encountered
no other development programs with complaints units or clear, institutionalized routine
mechanisms for dealing with problems arising from implementation.
The lack of such planning for the grievances that arise can lead to the programs having destructive
consequences. For example, if we examine further the case of the burning of the abattoir (see Box
3.8, above), there were no avenues of recourse for frustrated villagers, and when they did try to
complain early on to district officials in the Office of Animal Husbandry, there was no response to
allow for the relocation of the abattoir. The villagers also complained several times to the Village
Head who argued it was outside of his authority. Eventually, the abattoir was burned down by
aggrieved villagers.
When plans were made to rebuild the abattoir in a different location, there was no point of
contact readily available to allow villagers to communicate with the district office. The result
was that the remainder of the initial abattoir was burned down. Even when contact was made,
no action was publicly taken, or, if it was, the results were never reported back to the community.
The problems continued to escalate and more and more people, each with their own interests,
became involved.
In a number of cases in Flores, conflict emerged because of government corruption of development

resources. Some of these cases escalated because of a lack of complaints mechanisms to deal
with the resulting tensions. Box 3.11 outlines two conflicts in Sikka, Flores, related to the
government’s rice for the poor program.
One of the reasons these conflicts escalated was the lack of a mechanism within the OPK/
Raskin projects for communities to register their discontent.83 Indeed, in such cases there are
few in-built control mechanisms to prevent protests over government decision-making from
escalating. Given the extent to which participation is discouraged, and to which courts are
unlikely to convict local politicians of corruption, communities often have few available effective
outlets for dissent. Thus demonstrations in many cases turn violent, as frustrations rise.
83
Although it does seem that the government’s rice for the poor program (formally OPK, now relaunched as Raskin)
is particularly prone to corruption; see Olken (2006).
52
Box 3.11: Rice for the Poor? Two Cases of Conflict over Rice Distribution in Sikka
Both the government’s OPK rice and Raskin programs, which are aimed at helping poor communities, have resulted in corruption.
A key part of the ‘safety net’ constructed in 1998 after the Asian financial crisis, the OPK program (or Special Market Operation)
was a program whereby poor households received subsidized rice. In 2002, OPK was replaced by a similar program, Raskin
(Rice for the Poor). Corruption has not only taken place at the village level, where the poor community has to struggle to collect
the funds required to buy Raskin, but also at the kabupaten (district) level. Similarly, at the provincial level, Dolog’s (the
government warehouse for foodstuffs) involvement in the transportation of rice throughout NTT has resulted in other irregularities.
Two cases in Sikka, Flores show how corruption in the program can trigger broader conflicts:
15th December 2000, Nele Wutung village, Sikka:

The village and sub-village heads made their way to the Dolog office in Maumere (the district capital) to collect the village’s
2,940kg of rice. During their trip home, they stopped to unload 200kg (four sacks) of rice at the house of the sub-village head
of Kode. The remaining 2,740kg were finally unloaded at the Nele Wutung village government office. The villagers who had
been waiting at the village meeting hall heard the news that four sacks of rice had been unloaded earlier. They suspected that
the rice had been stolen. When they questioned the Sub-village Head he said that he had been ordered by the Village Head to
unload the rice at his house. Meanwhile, the head of another sub-village, Delang, said that the Village Head had used her own
money to purchase the rice directly from the Dolog. The villagers became frustrated, not only with the contradictory answers
but also with the fact that the Village Head was not in her office. They waited at the village hall; by nightfall approximately 62
households had still not received any rice. The community soon became rowdy, and staged a protest in front of the village
office. That night the Village Head arrived and tried to clarify the situation. However, the Village Secretary had no record of the
apparent transaction (200kg of rice) allegedly ‘bought’ by the Sub-village Heads. The atmosphere that night was tense and
fiery. Two groups formed in the village: those who supported the Village Head, and those who were in opposition. Around
2.00am flames from the village office could be seen. Both sides blamed the other for the burning: one side saying that the group
opposing the Village Head had committed the arson in frustration; the other saying, the Village Head’s supporters had done it
to hide the receipts of money “borrowed” from the village budget. The case is still unsolved.
5th July 2002, Watu Gong village, Sikka:

The community was waiting for discounted rice to be delivered from Dolog. Several hours passed without any sign that the
rice was on its way. The rice eventually arrived at 2pm. The village government, which had been assigned the task of
distributing the rice, believed that a scoop was the most accurate means of distribution. However, the villagers did not agree,
preferring that scales be used to ensure everyone received their fair amount. They were worried that the Village Head
would try to cream off some rice for himself if a scoop was used. The two parties could not agree. By 4.30pm the rice had
still not been distributed and tensions were rising. Six disappointed villagers decided to storm the village office, each taking
off with a sack of rice. Seeing this, many other villagers followed suit: before long, the atmosphere was chaotic with
widespread looting of rice. Police, on their way home from a neighboring sub-village, happened to be passing by. Seeing
the developing chaos, they fired shots into the sky. Slowly the situation was brought under control.
The Importance of Complaints Mechanisms

KDP, on the other hand, has a dedicated unit to handle complaints, with points of contact at the
regional level (through the Regional Management Unit – RMU). Complaints can be submitted in
person, or anonymously by post. While the role and function of this unit has evolved with the life
cycle of the program, it primarily handles large cases which are outside of the capacity of local
facilitators, somewhat diffusing tensions. Many of these cases relate to corruption within the
program. Between August 1999 and May 2005, the Handling Complaints Unit in NTT addressed
74 corruption cases, 8 of which took place in Manggarai and 4 in Sikka. In East Java, between
September 2001 and May 2005, the Unit addressed 131 such cases, 3 in Pamekasan and 13 in
Ponorogo.84
84
Data from the Handling Complains Unit of the National Management Consultants (NMC), Jakarta.
53
In many cases, follow-up of reported cases had a positive impact in diffusing tensions. None of
the instances which involved the complaints unit in East Java became violent, and many were
resolved. For example, the RMU had followed through on one case in Pasean, Pamekasan, where
program staff members were accused of misappropriating funds. When the staff members in
question were caught, they gave the funds back to the project. The RMU also followed-up on
cases of damaged water pipes and a case where villagers were facing sanctions from the program
for not meeting the implementation targets. The villagers spoke openly of the positive role the
RMU had played in resolving the problems and of their frequent visits to the field. In the interviews,
however, villagers also demanded more feedback on the outcome of cases forwarded to the
complaints unit.
These complaints mechanisms are extremely important. Where no complaints unit existed, where
there was a complaints handling unit but there was no action on complaints, or where there was
no feedback mechanism for the outcomes of complaints handling, then conflicts involving development
programs were more likely to escalate. All programs, and especially those dealing specifically with
conflict resolution and post-conflict reconstruction, require internal complaints handling mechanisms,
action processes, and systems for reporting back. KDP has these mechanisms, although the fieldwork
shows that they could still be improved. More resources need to be provided for the complaints units
to visit villages on a regular basis (both those experiencing and those not experiencing problems) to
provide the opportunity for communities to air grievances at a publicized time.
In general, all development programs, projects, and other initiatives should have clear complaints
mechanisms in place which:
are accessible to beneficiaries;
take responsibility for program decisions and communicate these to villagers;
are willing and able to listen to villagers and hear their complaints;
clearly publicize their availability and functions.
3.4 The Role of Outside Actors in Solving KDP-related Conflicts
While KDP does have internal mechanisms that allow for program-related problems and conflicts
to be dealt with, in some cases it is necessary for outside actors to get involved. This is the
case both for problems relating to competition within the program (Box 3.12) and for program
malfunction problems.
The qualitative fieldwork showed that facilitators alone often could not solve problems related to
KDP malfunctions. While facilitators and program structures are essential for access and
preventing violent conflict in the first instance, they are not always successful at dealing with
significant program malfunctions such as deviations in program processes, corruption, problems
with facilitators, or the kinds of problems that sit outside of everyday KDP interactions and
processes.
54
Box 3.12: Competition and KDP: Disputes in Manggarai, Flores
August 2003, Bangka Lelak village, Kecamatan Ruteng: A conflict has emerged over which proposal should be put
forward to the MAD (inter-village meeting). Three proposals were submitted from three different hamlets: a proposal
for clean drinking water (Tongger hamlet), a proposal for a village road (Mbohang hamlet), and another village road
(Manu hamlet). At the meeting, the different hamlets agreed that the proposal for the road from the group from
Mbohang should go forward. However, the group from Manu hamlet disagreed. A heated debated ensued; agreement
was impossible.
The group from Manu expressed their disappointment by tearing down the KDP information board. The next day,
the residents of Manu carried the board to the Camat’s (Sub-district Head’s) office to register their objections. There,
the Camat explained KDP mechanisms, procedures and regulations. After they had the regulations and process
explained, the Manu residents understood and were satisfied with the explanation.
Kecamatan Facilitator (FK), Bangka Lelak, Ruteng, Manggarai (M.1030)
Two people from Kecamatan Pagal were not satisfied because their proposals were rejected by the verification team.
Consequently, they reported their complaints to the Bupati (District Head). The Bupati ordered the Camat to handle
the case. The Camat, a representative from the district-level PMD (Village Community Empowerment body within the
Ministry of Home Affairs), the Village Head, and the LKMD (Village Community Resilience Council) solved the problem
in the sub-district office by accepting their proposals.
Head of LKMD, Pagal, Cibal, Manggarai (M.1005)
Often it was necessary for other actors (informal leaders and/or the state) to get involved. In
Pamekasan, all the large conflict cases required intervention from an outside actor: one case
involved the District Head reaching an informal agreement with the Village Head; the other involved
the Village Head and community and religious leaders intervening in the dispute. In the case in
Slahung, Ponorogo, outlined above (Section 2.2 of this chapter), the media, local leaders and
government officials were all embedded in the resolution process. In two villages in Sikka, Flores,
parties external to the program stepped in when repayments from KDP micro-credit dried up:
“At the KDP forums at the dusun (sub-village) level, the Village Head and several respected community figures from
Hobuai hamlet had to step in and take action when it came time to collect the repayments. They wrote a latter to the
UPK in Talibura as well as to the FK and Camat (Sub-District Head) to indicate their regrets … they also requested
that all of the parties reconsider their proposal…”
Primary School, Werang, Talibura, Sikka (S.1123)
“We once formed a Team for the Collection of Delinquent Loans. The team included members of the police and
military. We went to all the villages and after two weeks we had collected Rp. 60 million. I think the team was very
effective, but we didn’t have enough operational funds to pay the large lump sum requested by the team. That’s why
we have only done it once until now.”
Sub-district Head, Paga, Page, Sikka (S.1134)
“The surprising thing is that when we went out with the police, military, Camat, Village Head, BPD and TPK, they
actually wanted to repay their loans!”
Secretary of the UPK, Paga, Paga, Sikka, (S.1125)
In Cibal sub-district in Manggarai, a Tim Sukses (Success Team), made up of members from
various formal and informal institutions, was formed to overcome problems with delinquent loans.
The cases demonstrate that while there are many problems that facilitators and forums can deal
with themselves, resolution of problems is often easier where there are strong links between the
program, informal leaders, and the state.
55
In East Java, the most successfully resolved KDP-related cases involved program facilitators
working together with outside formal and informal leaders. In one case that was successfully
resolved in Ponorogo, the village facilitator (FD) used creative means to find ways for the savings
and loans group to pay back their funds. He employed the program principles while adapting the
repayment system to the local conditions and seasonal nature of the livelihood of local people, and
worked together with the local group leader to reach agreement; as such, his local knowledge and
the technical resources provided by the program were central to the success of his initiative. In the
other successfully resolved case (the Slahung bridge case outlined earlier), outside actors— the
district parliament (DPRD), the village parliament (the BPD), the media, members of the district
executive, and the security forces—all sat down with program facilitators and beneficiaries to seek
an agreement, and then to enforce it.
“…Eventually a meeting was held in the district hall. The district administration, district parliamentarians, and the
district consultant all attended. At the time, the village technical assistant couldn’t say much. The FK supported him,
but he wasn’t particularly strong. Then the district consultant supported the technical assistant. But the community
wanted the bridge construction to return to the original proposed design. After this was agreed, it was over.
The same theme emerges in both KDP and non-KDP conflict cases that have escalated. In
many such cases, both formal (state) and informal (non-state) actors are needed for the problem
to be successfully resolved. Where there is synergy between the state, informal leaders, and
program staff, development conflicts are much more likely to be resolved.
4. Comparison of KDP and Other Development-Related Conflicts
Q: In your opinion, what is the difference between KDP and other projects?
“...There is a huge difference. With KDP the community can be directly involved, through putting forward proposals,
planning, and implementation ... In putting forward proposals, they can be involved and give their opinion, starting
from the hamlet level to the village level. Then, in the planning they can also get involved … and in implementation
they can be involved as labor, which adds to their income … In other projects, only certain people are involved in the
meetings, it’s enough with just the Village Head and his officials … even then not everyone attends … the saying goes,
‘It’s enough to only have four people at a meeting’ … Yes, at most the Village Head, the Secretary, the Head of the
Village Development Board … and in the implementation of the project it is usually only outsiders.”
BPD (Village Representative Council) Chair, Padegelan, Pademwau, Pamekasan (1209)
As we have seen, KDP does trigger conflicts, but these conflicts are generally solved, or at least
dealt with, before they escalate. Almost without exception, they do not become violent. How
does this compare with conflicts related to other development projects in Flores and East Java?
As Table 3.6 shows, KDP conflicts are less likely to result in violence than either those related to
government administration and service provision, or other donors’ programs. While only one of
forty-two KDP conflicts reported in local newspapers between 2001 and 2003 became violent
(2%), 36 government-related conflicts resulted in violent impacts (5%). The newspaper data
only picked up one violent conflict related to other donors’ programs, but this is also
proportionally a greater amount (also 5%).
56
Table 3.6: Development Conflicts and Their Impacts
Type of Program # Conflicts # Violent % Violent # Killed # Injured # Properties

Conflicts Damaged
KDP 42 1 2% 0 2 0
East Java 10 0 0% 0 0 0
NTT 32 1 3% 0 2 0
Other Government 713 36 5% 4 38 11

East Java 619 20 3% 2 23 6
NTT 94 16 17% 2 15 5
Other Program 19 1 5% 0 1 0
East Java 17 0 0% 0 0 0
NTT 2 1 50% 0 1 0
Total 774 38 5% 4 41 11
Source : KDP & CCN dataset
If we look at all the development case studies collected in the field, it is clear why KDP conflicts
tend to be less violent than non-KDP development conflicts. As discussed above, KDP has
forums, facilitators and other internal complaints mechanisms to handle conflicts as they arise.
When it works well, KDP also provides space for non-program actors (local government, etc.)
to play a role in resolving program-related programs; other programs, which often have more
rigid hierarchies and less room for “creative implementation” strategies, provide fewer
opportunities for benign intervention when things go wrong. Comparative analysis of the
cases shows one other reason why KDP conflicts are less likely to turn violent. Because KDP
provides opportunities for communities to define their needs, programs are less likely to clash
with local priorities (see Box 3.13, below).
The research followed a number of other non-KDP development conflicts, in particular in East
Java. These included conflict over forest development policies in Dayakan, problems related to
mining refuse in Sampung, and the abattoir case in Banyupelle, Pamekasan, discussed above. In
these cases, there was no facilitator proximate to the villagers who could assist with problem
resolution, nor were there forums available to discuss the problem. In all three cases, transparent
and accountable processes were missing. Those at the higher levels did not adequately respond
(if at all) to the complaints and grievances of the beneficiaries, even after repeated attempts. In
these three cases, acts of civil disobedience and violent property destruction resulted, as seen in
the burning down of the forest and the abattoir, and the blocking of the culverts that diverted the
mining waste.85
85
See Barron and Diprose (forthcoming) for the cases and some analysis.
57
Box 3.13: Aid and Conflict in Madura
Three of the government’s development projects that turned violent in East Java related to aid promised to internally-
displaced Madurese fleeing from the conflict in Central Kalimantan. The cases show that when beneficiaries of aid
are left out of decision-making and needs identification processes, things can sometime get out of hand.
10 May 2001: Hundreds of internally-displaced people (IDPs) from Central Kalimantan riot over distribution of food
aid in Desa Torjunan, Sampang. Officials promised each refugee Rp. 10,000 (approximately USD 1), to be distributed
at the Village Head’s house. However, they underestimated the number of IDPs likely to appear, and when funds ran
out the IDPs rioted and vandalized the house. They were angry because officials were using IDP figures submitted
by the Village Head rather than those counted by the IDPs themselves. They also accused the Village Head of
corruption, stating that they only received Rp. 19 million of the Rp. 25 million allocated for food aid that day.
10 September 2002: “At least 1000” students and IDPs clashed with police when a demonstration to the district
government in Bangkalan district got out of hand. Demonstrators were demanding an explanation about the distribution
of food aid to refugees and requested that an official address the crowd directly. The government offered demonstrators
the opportunity to send representatives to meet with the government, which they refused. Police were called and a
clash ensued. 18 students and IDPs were injured during the clash, two of whom were hospitalized.
10 July 2003: Fifty IDPs from Central Kalimantan, now staying in Ketapang Barat village, Sampang, vandalized
temporary accommodation (barracks) funded by aid from Kuwait, with money channeled through the local government.
The construction was 80% complete, but was put on hold following the vandalism. Following this, representatives of the
refugees met with sub-district and district officials and made several demands: that no further barracks be built, as they
limited options for IDPs hoping to return to Kalimantan and because existing accommodation was sufficient (and in many
cases unused). Rather, they argued, aid funds should be used to provide food staples and education for IDP children.
From newspaper articles from KDP & CCN dataset
In contrast, in most of the KDP cases followed in this study, mechanisms were in place to address
the issues as and where they arose. For example, in the case in Jenangan, Ponorogo, the KDP
district consultant unilaterally changed the technical drawings for a bridge that was to be funded by
KDP. The community protested, and meetings were held to help resolve the problem. This
resulted in quick action from higher level program officers, informal leaders and the media. There
was state, program, non-state, and community participation in the resolution process, and the
case was resolved peacefully. Similarly, informal leaders in Proppo were important actors in
resolving a case where KDP project workers protested against nepotism by one of the technical
team leaders by blockading the road.86 In another case in Pasean where there were problems
with disbursing funds, KDP facilitators quickly intervened and the problems did not escalate.
The comparative analysis thus makes it clear why KDP, when it works as intended, causes less, and
less serious, violent conflict than other development programs. In cases where KDP does not
work as intended, KDP is less successful at managing the tensions it generates.87 However,
where the program works well, there are rarely problems. KDP has a clear structure of forums
and facilitators (with secondary functions such acting as locations and points of contact for
discussion of problems), it aims to promote transparency and inclusion through socializing program
processes and results, and it has systems in place for ensuring accountability and effective handling
of complaints. KDP also provides mechanisms for communities to define and prioritize their needs,
hence making it less likely that development projects will be resisted by community members.
86
This case, and all the others referred to in the paper, are available online at www.conflictanddevelopment.org
87
See the analysis earlier in this chapter, and the material discussed in Chapter 7.
58
Cumulatively, all of these factors contribute towards limiting KDP-related conflicts, both through
minimizing the potential for conflict and through providing opportunities for redress when problems
do occur.
5. Summary
This chapter has shown that development and conflict inevitably go hand in hand. Development
projects and programs introduce resources into poor areas, and inter-group competition over such
resources can lead to tensions. Programs like KDP, which explicitly aim to empower
marginalized groups, also introduce new rules and norms about how decision-making should be
made, in so doing impacting on local power balances and structures. Resistance of elites to such
changes is another basis for conflict. Accordingly, those designing and implementing development
projects must be aware of the ways in which projects can trigger conflicts, or interact with existing
ones, and must develop strategies and mechanisms for dealing with development-related conflicts
as they arise.
The research found that ‘micro-conflicts’ related to KDP are common, but that these rarely
escalate and almost never turn violent. There are three forms of development-related disputes.
First, KDP introduces competition within and between villages over which proposals should be
funded. This can lead to tensions, in particular when groups feel that the decision-making process
was not transparent or fair. However, the research found that, over time, groups tend to accept
the validity of the competition process and, as a result, its results. Only where the program does
not function as intended (e.g., where one group has captured the process) do larger problems
emerge. The second form of conflict relates to these and other ‘program malfunctions’.
Malfunctions can either be problems of omission or commission. The first is a result of poor
socialization or implementation; the second concerns situations where there has been deliberate
and active malfeasance from program staff or local elites. The latter is more serious than the
former, with cases of corruption providing a basis for larger community unrest.
A third form, interaction conflict, occurs when development projects (KDP or others) interact
with pre-existing local tensions, power structures or conflicts, triggering conflict escalation and,
in cases, violence. Projects operate in theatres in which power relations are constantly being
negotiated; development projects constitute a vital resource that can be utilized in this process.
In cases, interaction conflicts involve actors using the project for patronage purposes, raising
tensions between competing local elites. In others, elites attempt to capture the project for self-
enrichment. Other cases relate to the resistance of elites to the norms of widespread access to
decision-making, transparency, and accountability that KDP brings. Where resistance is greater,
there is more potential for conflict.
Despite the numerous ways in which projects can cause conflict, we found that KDP-related
conflicts almost never escalate and/or turn violent. This is in large part because the program has
effective in-built conflict resolution mechanisms. Project forums, facilitators, and other complaints
mechanisms, mean that tensions can addressed before they grow in size. Resolution success
rates of KDP-related problems are very high. This is in marked contrast to many other development
projects, which do not have such mechanisms. As a result, conflicts relating to other governmental
and non-governmental projects are far more likely to turn violent.
59
60
Chapter Four
Development Projects as Part of ‘the Solution’?

The Impact of KDP on Non-Project Conflict
In the last chapter, we showed that KDP is very good at solving problems relating to the program
itself. While KDP, like any development project, does cause tensions and on occasions trigger
conflict, almost all of these conflicts are dealt with through the program’s in-built conflict resolution
mechanisms. In this respect, KDP often works better than other development projects. Yet is there
any evidence of increased conflict management capacity in KDP areas in terms of non-project
conflicts? Are the KDP conflict resolution mechanisms used to deal with other conflicts that do not
relate to the program? Are there other social spillovers that result in improved conflict management
capacity? If so, in what ways does the program appear to have made a difference? And for what
types of conflicts, in what contexts?
The next four chapters seek to answer these questions through comparative analysis of conflict
levels and outcomes, conflict management processes, and indirect impacts on what we term
“conflict environments” in areas that have received the program and areas that have not. In this
chapter we look, at a fairly macro level, at conflict levels and outcomes. We do so by using the
newspaper datasets, which record all reported incidents of conflict in twelve districts in the two
provinces between 2001 and 2003.88 Merging in data on where KDP was operating, and when,
allows us to look comparatively at conflict trends in sub-districts (kecamatan) that had the program
(treatment sites) and those that did not (control sites), as well as in areas before and after KDP
implementation. Conflict levels and outcomes are but two indicators of conflict management
capacity. Underlying structural factors and exogenous forces may lead to different levels of
conflict in areas that have similar conflict management capacity. However, comparative
examination does provide some insights on the strengths and limitations of the program in changing
local conflict dynamics. Chapters 5 and 6 will examine in more depth the processes, direct and
indirect, by which such change is affected.
1. Forms of Conflict in East Java and NTT
Before examining the impact of KDP on conflict and conflict outcomes, it is necessary to
provide a brief sketch of the forms and impacts of conflict in our research areas. Conflict is
common. We recorded 1840 discrete conflict incidents over the three-year period in the
twelve districts, 591 of which were violent.89 Conflict in our research areas resulted in 275
88
As noted earlier, we conducted the newspaper survey in our four research districts plus their neighboring districts.
We therefore collected data on three “clusters” of districts: the Ponorogo cluster (Ponorogo plus Madiun and Magetan);
the Pamekasan cluster (Pamekasan plus the three neighboring districts on the island of Madura); and the Flores cluster
(all districts on the island of Flores). Over the course of the research, two of the five Florenese districts split, making
a total of seven districts on the island. However, for coding purposes, we use the old district boundaries.
89
It should be noted again that we used a wide definition of conflict that includes relatively minor disputes. We did so
in order that we could compare conflicts over similar issues but with different outcomes; see the discussion in the
Chapters 1 and 2.
61
deaths in the 2001-2003 period, 158 of which were in East Java, 177 in NTT. The death total was
highest in 2001 (120 deaths), after which death totals declined (91 deaths in 2002; 64 in 2003).
Death totals ranged from 4 deaths in Ponorogo district (East Java) to 52 deaths in Manggarai (NTT)
(see Table 4.1).90 While violence is relatively common, and its cumulative impacts significant,
with a few exceptions most of the violent conflicts remained relatively small-scale; 47% of conflict-
related deaths in East Java and 38% in NTT were the result of conflicts between individuals.91
Table 4.1: Conflict-related Deaths (2001-2003): KDP & CCN dataset
Districts 2001 2002 2003 Total

Madiun 6 3 2 11
Magetan 3 3 3 9
Ponorogo 0 4 0 4
Pamekasan 11 5 1 17
Sampang 14 15 9 38
Sumenep 17 12 5 34
Bangkalan 20 13 12 45
East Java 71 55 32 158
Ende 1 11 3 15
Flores Timur 2 15 5 22
Manggarai 41 2 9 52
Ngada 2 1 4 7
Sikka 3 7 11 21
NTT 49 36 32 117
Source: KDP & CCN data set
There are four main types of conflict in the research areas, plus a number of other conflicts that
do not easily fit into any of the broad categories. First, there are physical resource conflicts.
These include disputes over the ownership of, access to, or use of either natural or man-made
resources. By far the largest proportion of these types of conflict pertains to land—private land,
state-owned land, or communally-held (adat) land. Between 2001 and 2003, 218 conflicts of
this type took place in the research areas, 70 of which were violent. These resulted in 73 deaths,
109 injuries, and 13 damaged properties.92
Second, administrative conflicts. These include disputes over procedures or the management
of the provision of services or other resources. Thus the category includes conflicts over
government development projects and other forms of government funding and service provision.
It also includes disputes over donor projects and KDP, as well as disputes over the management
of companies. We analyzed many of these disputes in the last chapter. A second form was
disputes about administrative jurisdictions within government. Between 2001 and 2003, 932
conflicts of these types took place in the research areas, 53 of which were violent. These resulted
in 5 deaths, 81 injuries, and 16 damaged properties.
90
Other impacts of violent conflict, such as property damage and injuries, also varied considerably by district and did
not always correlate with death levels. See Barron and Sharpe (2005).
91
The big exception is communal land conflicts in Manggarai. These conflicts are between groups and have much
larger impacts each time they occur.
92
Clark (2004) gathers together many of the case studies relating to land conflicts that were followed in the research.
Forthcoming volumes will be produced on the other main forms of conflict.
62
Development Projects as Part of ‘the Solution’? The Impact of KDP on Non-Project Conflict
A third form of conflict was that relating to political position and influence. This includes positions
(state or non-state) primarily at the village or district levels, as well as disputes within and between
political parties for influence over decision-making. Between 2001 and 2003, 102 conflicts of this
type took place in the research areas, 8 of which were violent. These resulted in 2 deaths, 7 injuries,
and 9 damaged properties.
Fourth, and the most common form of violent dispute, was vigilantism and retribution conflicts.
Conflicts here included revenge attacks and lynching relating to a diverse range of phenomena:
theft, witchcraft, sanctioning of social ‘deviants’, sexual indiscretion, murder, broader identity clashes,
damage to property, accidents, humiliation, etc. They could take the form of individual personal
vigilantism/retribution, or group-based vigilantism/retribution. Between 2001 and 2003, 549 conflicts
of this type took place in the research areas, 430 of which were violent. These resulted in 178 deaths,
443 injuries, and 289 damaged properties.
Finally, a range of other conflicts were found in the field, the most common being those related
to domestic violence and other intra-family arguments. Between 2001 and 2003, 39 ‘other’
conflicts took place in the research areas, 30 of which were violent. These resulted in 17 deaths,
27 injuries, and 10 damaged properties.
The forms of conflict vary considerably between areas. Whereas conflict is common in both
provinces, the forms it takes (in terms of how actors organize themselves, where they live, who
they target, and what the preferred modes of violence are) differs greatly. As Figure 4.1 shows,
different forms of violent conflict are more common in the different provinces. Vigilante conflicts
are far more common in East Java, whereas resource conflicts, primarily over land, occur much
more frequently in NTT.
There are many reasons for the variation in conflict between provinces and, indeed, between
different regions within the provinces. Institutional, economic and cultural factors all
contribute.93 Within each area, serious disputes tend to be over a given issue (e.g., land) and
there are also commonalities within areas in how such conflicts are ‘expressed’ (e.g., the ways in
which actors participate in conflictual action, the symbols and strategies they employ, and so
on). In both East Java and Flores, the most violent forms of conflict are those prescribed, and
seemingly driven, by local cultural behaviors and norms: the history and development of rival
silat (martial arts) groups in the Ponorogo research area; Madurese culture that condones and
even insists on bloody retribution over matters of honor; the communal battles over traditional
land ownership and usage rights in parts of Flores, where actors split along lines of ethnicity and
lineage. However, local leaders play a large role in helping shape the specific ‘cultural realm’
which regulates conflictual action. Further, effective intervention by local leaders (formal and
informal) and state institutions (e.g., the security sector) can prevent culturally-legitimized forms
of violence from escalating (Barron and Sharpe 2005; Diprose forthcoming).
93
For basic frameworks see Barron, Diprose, Madden, Smith, and Woolcock (2004) and Smith (2005).
63
Figure 4.1: Number of Violent Conflicts by Type in East Java and NTT (2001-2003)
Number of Violent Conflicts
Source: KDP & CCN dataset
2. Conflict in KDP and non-KDP Areas
Are kecamatan that have received KDP more or less likely to experience violent conflict than
those which have not? And is there any difference in the forms of conflict in KDP and non-KDP
areas? Using the data on conflict levels in our twelve districts, we find the evidence about any
differences between KDP locations and places that did not receive the program to be inconclusive.
While levels of conflict and violence are lower in KDP areas in East Java, the opposite holds for
NTT. For some forms of violence, the presence of KDP is correlated with lower rates; for others,
the opposite is true. Collectively, the macro evidence does not point towards a significant
programmatic impact on levels of conflict and violence.
2.1 Average Levels of Violent Conflict in KDP and non-KDP Areas
Figure 4.2 shows comparative levels of violent conflict in KDP and non-KDP areas for East
Java and NTT. The ‘KDP’ bar shows the mean number of violent incidents that took place in
sub-districts in the years in which they had KDP; the ‘non-KDP’ bar shows the mean number
of violent incidents in years in which the sub-districts did not have KDP.94 The mean is of the
average number of conflicts in one year in KDP and non-KDP areas.
94
We matched the conflict data by year to whether a sub-district had had KDP in the previous year. We lagged that data
in order to ensure that the full cycle had finished, a necessary step if we assume that KDP only has a significant impact
over a full program cycle. To illustrate, take, for example, a sub-district that received KDP in cycle 2 (2000/2001) and
cycle 3 (2003/2004), but not in cycle 1 (1999/2000). The conflicts recorded in Years 2 and 3 of our conflict data (2002
and 2003) would be included in the ‘KDP’ total. However, the conflicts in the first year (2001) were recorded in the
‘non-KDP’ total because the program had not yet operated in that sub-district at that time. Thus one particular sub-
district will contribute towards both the KDP and non-KDP totals unless it received the program for at least three years.
If a sub-district had KDP for three years, it would contribute three years of data to the KDP sum; if it had the
64
In East Java, marginally more violent conflict took place in non-KDP areas than in areas with the
program (with a mean of 1 in the former and 0.92 violent conflicts in the latter). In NTT, areas
with KDP reported higher levels of violent conflict (1.04 versus 0.74).95 If we take all conflict,
violent and non-violent, similar trends are observable.
Figure 4.2: Mean # Violent Conflict Incidents per Year per Sub-District in East Java and NTT (2001-2003):
KDP and non-KDP
Mean Cases Per Kecamatan Per Year
If we look at most other indicators of violent conflict, we find similar patterns with lower rates in
KDP areas in East Java, and lower rates in NTT in non-KDP areas (Table 4.2). The big, and
statistically significant, exception is that there is a higher probability of an area having a case that
results in a death in KDP areas than in non-KDP areas in East Java.
Table 4.2: Mean # of Cases with Impact per Year per Sub-district in NTT & East Java (2001-2003): KDP and non-KDP
East Java NTT
Cases with a… Mean KDP Mean non-KDP Mean KDP Mean non-KDP
Injury 0.47 0.60 + 0.72 + 0.56
Property Damage 0.12* 0.23 + 0.20 + 0.16
Death 0.33* + 0.24 0.35* + 0.19
* Statistically significant at 3% level
+ denotes higher probability of given impact (compared to other group) Source: KDP & CCN dataset
program for two years, it would contribute two years of data to the KDP total, and one to the non-KDP total; and so
on. However, once a sub-district had had KDP, all subsequent years were counted in the KDP total, as we assume that
KDP programmatic effects will hold. If, for example, a sub-district had KDP in Year 1 and 2, but not in Year 3, we
would still count all three years of conflict data in the KDP total. Doing the analysis this way allows us to factor in for
conflicts that took place before the program arrived in an area. If we simply compared KDP and non-KDP areas, these
conflicts would bias the comparison. See further explanation in Annex C.
95
The difference in East Java between KDP and non-KDP areas is not significant; in NTT it is significant at the 10% level.
65
2.2 Changes over Time
We can disaggregate the data further to examine comparative levels of violent conflict in areas
which have had the program for different lengths of time.96 As Figure 4.3 shows, in East Java
there are similar levels of violent conflict in sub-districts which have not had KDP and those
which have had it for one or two years. Only in areas in their third year of the program are
rates of violent conflict noticeably lower. There is a less clear pattern for NTT.
Figure 4.3: Mean # Violent Conflict Incidents per Sub-district per Year in East Java (2001-2003),
by Year of KDP
Mean Casses Per Kecamatan Per Year
What do these results tell us? First, it is clear that KDP, in and of itself, does not have a
significant positive impact in reducing levels of violent conflict at an aggregate level. Indeed, if
anything the evidence points in the other direction. Violent conflicts are more likely in areas that
have KDP in NTT, although not in East Java, and conflicts resulting in a death are more likely
in KDP areas in both provinces. Can we say that KDP is leading to higher levels of violent and
destructive conflict?
The short answer is: not really. There are many factors that determine an area’s conflict profile.
Violent conflict is a rare event and is not normally distributed. The research has shown that there
needs to be a confluence of a range of different phenomena for a dispute or social tension to
escalate into violence. Existing social, political and economic structures, and historical legacies,
96
The data should be interpreted carefully. The bar for ‘One year’, for example, includes data only on conflicts that
took place in sub-districts during the first year the program was there; the bar for ‘Two years’ includes only data for
sub-districts in their second year of the program. Again, the data is lagged. As such, a sub-district which has had KDP
for three years would contribute one year of data to the ‘One year’ bar (for 2001), one year to the ‘Two years’ bar
(2002), and one year to the ‘Three years’ bar (2003). Because the ‘Zero year’ bar includes three years of data for places
that did not receive the program, two years for those which had it for one year, etc., the sample size is larger, even in
NTT where most places have now received the program (see Annex C).
66
all help explain why some areas are violent and others are not. KDP, and the rules and norms it
brings, is but one factor in determining an area’s “conflict environment.”97 The reason most of
the results are not statistically significant is because of massive variation between sub-districts in
terms of the prevalence and impact of violence. This is true in both KDP and non-KDP areas,
and points towards other local factors playing a large role in determining levels of conflict.
Indeed, we would also expect a bias against finding, at an aggregate level, lower levels of conflict
in areas with KDP. KDP was targeted at the poorest areas of Indonesia, with the sub-districts that
received the program first being the poorest.98 Previous research on conflict in Indonesia has
shown links, albeit complex ones, between forms of poverty and of conflict incidence (Barron,
Kaiser and Pradhan 2004). The international literature backs this up.99 Given that KDP was not
randomly assigned, and that we are not controlling for other structural factors in the community, we
might expect a bias to exist towards higher conflict in KDP areas.
The lack of evidence that KDP, in and of itself, positively influences aggregate levels of conflict,
however, should not be interpreted as meaning that KDP has no influence at smaller units of
analysis. It does mean that development agencies should resist temptations to believe that CDD
projects can, ipso facto, be expected to deliver a ‘peace dividend’ in moderate conflict
environments. As previously noted, however, KDP’s key organizational and decision-making
innovations occur at the community level, where it is designed to provide a flexible structure
that can be adapted to the specific features of local contexts. As such, we would be more
likely to find any impacts it generates with respect to local conflict trajectories and conflict
management capacity at the community level, and for those impacts to be highly contingent on
the nature and extent of its interactions with those contexts.
A range of forms of conflict thus exists in East Java and Flores, and local patterns of conflict
vary based on local conditions. A large number of such local factors help dictate the incidence
and form of conflict. In order to ascertain more precisely the impact of the program, it is necessary
to examine in greater depth how particular cases of conflict play out in areas with and without the
program, and—specifically—whether and how KDP interacts with key aspects of the social context
within which the program is embedded. We explore these issues in the following chapters.
97
See discussion in Chapters 6 and 7 below.
98
A long list was prepared of poor sub-districts using data from the SUSENAS survey. This list was cross-checked
locally. This process allowed for local knowledge of poverty incidence to be factored in. However, at the same time,
it did mean that in places political calculations helped determine the list.
99
In the international literature a number of researchers and theorists have identified absolute poverty as a source of
conflict. Burton (1990), for example, argues that violent conflict is a result of the denial of basic absolute needs.
Addison (1998) has used the African experience to hypothesize that slow growth is a predictor of violent unrest.
Galtung (1969) argues that where underlying structural inequality exists, “positive peace” is unattainable. William
Easterly (2001) has used empirical evidence from Africa to demonstrate that poverty, when combined with inequality,
can easily lead to violence.
67
68
Chapter Five
The Direct Impacts of KDP on Conflict Management:

The Role of Forums and Facilitators
The evidence is ambiguous as to whether there is more or less violent conflict in KDP locations
than in areas that have not received the program. Given the range of factors that contribute
towards the presence of violence, and the fact that KDP is deliberately placed in poorer areas,
it is hard to identify, at least at a macro/aggregate level, whether the program helps to make
areas (sub-districts) more robust to conflict or not. In order to ascertain the program’s impact
on conflict management capacities at the micro/community level, it is necessary to also examine
in more depth the ways in which the program is utilized (or not) for conflict management, as well
as the impact of the program on a range of variables that help to shape the likelihood of conflict
emerging and/or escalating.
This chapter does the former: it looks at the ways in which KDP may directly impact conflict and
conflict management capacity. There are two potential direct effects that KDP could have (see
Figure 5.1). First, the spaces it introduces. The KDP forums at the sub-village, village, and
sub-district levels) may be used for conflict resolution purposes or for directly addressing the
underlying social tensions that could escalate into conflict. In most cases, effective conflict
mediation requires a dialogue between disputing parties. This needs to happen in a place that is
perceived to be neutral—hence the fact that many long-standing international disputes are
mediated in places far away from the homes and power bases of the disputants.100 The same is
necessary for local conflict resolution; disputants need to feel they are in a ‘safe’ place that
putatively favors neither party. In particular, in resolving disputes where there is no technical
‘divide the pie’ solution, and hence where creative “adaptive” agreements are necessary, there is
the need for what Heifetz (1994) calls a “holding place” to contain the stresses each party faces
in compromising. Do KDP forums provide such a space?
Second, KDP introduces people who may play a role in conflict mediation. The program places
facilitators at the sub-district level (FK) and requires the election of village level facilitators
(FD). Other key staff members include the district program manager (KMKab) and the village
activities implementation team (TPK). The FK and KMKab are usually not from the areas in
which they are stationed. Are they able then to effectively play the role of neutral outsider in
arbitrating disputes between villagers and villages, not just those that relate to the program but
others, too? In contrast, FDs are generally locals, being from the village in which they serve.
Their position has authority, yet they are not official government employees. Does this give them
100
Take, for example, the separatist conflict in Aceh. Peace talks between the separatists and the Indonesian Govern-
ment took place in Finland. There are a number of reasons for this, including attempts by the separatists (GAM) to
internationalize the conflict. However, another reason is mistrust by the separatists of the Indonesian Government,
with fears that they may not be free from coercion if the talks were held on Indonesian soil. Indeed, given the fact that
GAM representatives to earlier talks were promptly arrested upon returning to Aceh, they may have a point.
69
the credibility and legitimacy to mediate in disputes that are unrelated to the program, and/or
that others (with their clearer linkages to the intents of the state) may not be willing or able to
effectively mediate?
Figure 5.1: Potential Direct Impacts of KDP on Conflict
Supervision missions and reports from field staff prior to the study seemed to suggest that KDP
forums and facilitators were, in many cases, playing these roles. In North Maluku, for example,
an area that has experienced high levels of conflict and violence,101 KDP supported the reconciliation
process, through the participation of kecamatan facilitators (FK) in the reconciliation process in
Southern Ibu. The program supported a football competition between villages as part of the
reconciliation process, and KDP forums were used to promote relations between villages that
previously had little interaction.102 The provincial KDP team formed a Reconciliation Team composed
of Christian and Muslim community leaders, customary leaders, youth, university students, and KDP
kecamatan facilitators. In Lampung province, a series of thefts led to a vigilante response, with
Javanese villagers burning down 60 houses in a Lampungese sub-village (Barron and Madden 2004).
As part of the reconciliation efforts, an inter-village meeting was held resulting in the use of KDP
funds to rebuild the destroyed houses, a measure that would help promote reconciliation between
groups within the village and with other villages.103 In Central Kalimantan province, KDP staff played
a role in trying to prevent the massacre of the Madurese by Dayaks from escalating (Smith 2005).104
Indeed, an internal review of KDP in high conflict areas of Indonesia concluded:
“KDP provides a useful framework for negotiations and consensus-building. The meetings and
procedures provide open fora for communications and discussion. In locations such as Aceh
and the Malukus, communities are able to come together despite the conflict situation and reach
common ground. KDP provides a non-violent forum to mediate differing interests and reach a
broad consensus (through “musyarawah”) on what is in the best interests of the community.”105
101
The UN estimates there were 2,794 deaths from collective violence between 1990 and 2003 in North Maluku
(Varshney, Panggabean, and Tadjoeddin 2004).
102
See World Bank (2002).
103
Based on field visits in January 2002.
104
For background on the Central Kalimantan conflict, see also Bertrand (2004).
105
Government of Indonesia (2002).
70
The Direct Impacts of KDP on Conflict Management: The Role of Forums and Facilitators
The research in East Java and NTT found, however, that there was little aggregate evidence that
KDP forums or facilitators were playing a direct role in the management of conflict unrelated to
the program. KDP forums are not often used for dealing with non-project conflicts, although
when they are they tend to be successful. More markedly, KDP facilitators almost never
address non-KDP conflicts outside of KDP forums. In East Java and NTT, KDP plays only
a very small direct role in local conflict management and resolution; as we shall we, its
indirect impacts, however, are more significant.
1. The Use of KDP Forums for Managing Non-Project Conflict
KDP introduces planning and decision-making forums at a number of different levels (outlined in
Box 3.9). Are these forums used in East Java and NTT to address conflicts that are not related to the
program?
The key informant survey found that KDP forums are used occasionally, but not regularly, for
addressing conflicts unrelated to the program. This use is higher at the sub-district level. The
results, when combined with evidence from the qualitative fieldwork, show that on occasions KDP
forums are ad hoc venues for conflict resolution, but that their use for dealing with non-project
conflict has not been institutionalized.
In East Java, 13% of village informants surveyed, and 20% in NTT, reported that KDP forums
had been used in their area to solve conflicts not related to the program.106 In both provinces,
reported use for non-KDP problems was higher at the kecamatan level, where 20% of informants
reported use in East Java, and 47.4% did so in NTT.107 However, this is as much a function of the
way in which the question was asked as an indication of higher use at the sub-district level.
Reported rates are higher at the kecamatan level because informants at this level would know
about the times KDP was used for such purposes across a wider range of villages (i.e., all or most
in the kecamatan) compared to village informants, who reported only on forum use in their own
village. Where forums are used for dealing with non-KDP issues, evidence from the fieldwork
shows that this is more likely to be at the village level.
The fact that KDP forums are not used regularly for dealing with conflicts unrelated to the
program is confirmed by the qualitative research. In almost every village, respondents were
clear when asked about the extent to which issues not related to KDP were discussed in KDP
forums:
“They only talk about KDP during the KDP meetings.”
Farmer, Magepanda, Nita, Sikka (S.606)
“The KDP forums aren’t used to discuss other issues outside of KDP.”
Village Secretary, Tengku Leda, Lambaleda, Manggarai (M.1024)
“In the KDP forums in Desa Cumbi, we have never raised any other issues besides issues or conflicts involving KDP.”
Village Facilitator (FD), Cumbi, Ruteng, Manggarai (M.1036)
106
107
Although sample sizes are small: East Java (n = 20); NTT (n = 19).
71
“In the KDP forums, both the Musbangdes (the village decision meeting), and the UDKP (the sub-district meeting,
also known as MAD), only KDP problems are discussed. However, there was a case where UDKP was used to discus
the problem of village development…but, problems outside of KDP have not been resolved in KDP forums, because
there would definitely be interruptions from the participants because they would consider it irrelevant.”
Golkar member, Padellegan, Pademawu, Pamekasan (1204)
However, while the majority of informants claimed such, KDP forums are on occasion used for
conflict mediation purposes. In particular, at times, and especially in East Java, conflicts related
to other development projects sometimes make their way into KDP forums. In NTT, other conflicts
between villagers are, on occasions, discussed (Figure 5.2).108
Figure 5.2: Types of non-KDP Problems/Conflict Addressed in KDP Forums, by Province109
% of each Province
While KDP forums are rarely used for dealing with non-KDP problems, when they are used they
are usually successful. In East Java, 62.5% of informants who said that the forums dealt with non-KDP
problems said that when they were used they were successful; in NTT, 60.6% reported success.110
It should be noted that this is a lower success rate than for KDP-related problems.111
Why were KDP forums used so infrequently? And why, when they were used, was it for these
types of problems? Answering such questions requires an understanding of people’s perceptions of
the forums, of conflict resolution norms, and of the role of local elites and program staff.
108
The small samples should be noted. Only respondents who had said that KDP forums were used for addressing non-
KDP problems/conflicts (33 in NTT, and 24 in East Java) answered this question.
109
Note: the percentages are those who cited KDP forums were used for a given type of problem/conflict, out of only
those who reported that KDP forums were used for non-KDP issues. Hence the small sample sizes.
110
111
As noted earlier, 84.1% of informants in East Java, and 72.1% in NTT reported that KDP-related conflicts/
problems dealt with in KDP forums were successfully resolved.
72
1.1 Problems and Jurisdictions
The primary reason why KDP forums are not used for solving most kinds of conflict unrelated to
KDP is simple: villagers have other institutions and actors that they deem more suitable for these
purposes. In almost all villages, people know where they should take a given type of problem/
conflict. Particular people are seen as having the legitimate jurisdiction for addressing different
types of conflict.112 Take, for example, inter-village land conflicts in Manggarai, Flores. Disputants
know that the conflict should be taken to the Camat (sub-district head) who normally tries to deal
with the problem in an adat manner involving traditional leaders. If this is not successful—as is
often the case—the conflict will rise to the district level government and/or into the formal court
system. In contrast, domestic violence cases will usually be addressed within the family or, if a
homicide is involved or the case is a repeat offence, by the police (normally at the sub-district
level). In Ponorogo, East Java, conflicts between silat (martial arts) groups are normally dealt with
by the police, in consultation with the group leaders. KDP forums are not seen as having legitimate
jurisdiction for these types of problems.
“Resolution [between silat groups] is carried out only by the leader of the group with the problem; it can also be
mediated by a third party. I was once a mediator when there was a problem between SH Terate and BS [Bintang
Surya – two silat groups] … Problems [between inhabitants] are usually settled in routine hamlet meetings … When
there’s an urgent problem, a special meeting will be held…”
BPD Chair, Wates, Slahung, Ponorogo (1334)
A range of different actors and institutions are seen as having a legitimate role in mediating
disputes. In the research areas, the government, at the district, sub-district, village, and sub-
village), the security sector (police and military), the judicial system, traditional adat and other
community leaders, the Church, Kyai (Muslim/Islamic clerics), and others all play a role in
mediating different forms of disputes. Given the extent to which perceptions of “the correct
place” to deal with problems are bound up in social norms—which have in most cases developed
over long periods of time—it is unsurprising that relatively recently-introduced forums would
not be seen as being the appropriate place to handle most kinds of disputes. When survey
respondents were asked for the reasons why KDP-related conflicts were solved in KDP forums,
the top two reasons were that there was either nowhere else to address these problems, or that
KDP forums were the most appropriate place for them to be addressed. In contrast, only in one
sub-district (Cibal in Manggarai) was either of these reasons given as explaining why non-KDP
conflicts were solved in KDP forums.
Of non-KDP conflicts, those relating to other development projects are the most likely to be
addressed. This is the case for two reasons. As we noted in Chapter 3, there are often few
avenues where problems that emerge from government or donor development projects can be
taken. In the past, local leaders and officials had hegemonic control over development resources.
With the democratization of life at the village level, accompanied by a stronger system of checks
and balances, conflicts over development projects are arising in ways not true in the past.113
112
For more on who is seen as legitimate for addressing which types of conflict, see the two provincial reports: Diprose
(forthcoming) and Satu and Barron (forthcoming).
113
The introduction of the Village Representative Council (Badan Perwakilan Desa – BPD), an elected body that holds
the village head accountable, has changed the uni-polar system of power in village life (see Antlov 2001). The
73
Given that many government and donor projects do not have effective in-built complaints and
conflict management mechanisms, there is often an institutional void as to where these problems
can be taken and discussed. In some cases, KDP forums step in to fill this vacuum.
KDP forums are primarily designed to deal with the kinds of problems that tend to emerge from
development projects: complaints over processes of resource allocation, jealousy over who gets
a project and who does not, and so on. Given that the problems that arise from different
development projects are similar to those that emerge from KDP—the ‘program malfunction’
conflicts we identified above—it is hardly surprising that KDP resolution mechanisms can in
some cases play a role in dealing with them.
In NTT, a number of respondents also said that KDP forums were used to deal with conflicts
between villages, and disputes relating to theft. There were a number of examples of KDP forums
being used to discuss theft, problems relating to the control of livestock, and other village problems.
In these cases, it was the absence of other opportunities to bring people together that was cited as
the reason why KDP forums were used. Given the declining frequency with which village
musyawarah (group decision-making meetings) are held in many villages, KDP forums were seen
as a comparatively rare opportunity to bring the whole community together in one place:
“Problems relating to village development were raised in the KDP forums in Wairterang and Egon. These were problems
which were not related to KDP, but were raised in the KDP forums. As KDP staff members, we normally give the Village
Head an opportunity to chair the forum if there were any problems which were not related to KDP [which needed to be
discussed]. We only gave suggestions. Or, if we didn’t feel brave enough we just kept quiet. We just let them resolve
their problems themselves in the KDP forums.”
Head of the KDP Financial Management Unit, Talibura, Talibura, Sikka (S.1123)
The evidence supports the thesis that people will only bring conflicts to KDP forums that they think
will be able to be effectively addressed there. Reported successful resolution rates for conflicts
brought to KDP forums are high: 86% for development-related problems, 100% for problems relating
to theft and conflict with other villages, 96% for conflict between villagers, and 74% for other
conflicts brought to the forums.114
Thus KDP forums are most likely to be used when there are no other existing institutions to deal
with the problem, and/or when people believe KDP is able to deal with the particular conflict. In
general, institutions exist in East Java and NTT to deal with most types of conflict, although by
no means do they always do so effectively. There is no evidence that KDP displaces other
problem solving forums—that is, that where existing institutions exist, KDP usurps their role.
Rather, the places where people take different conflicts are likely to be determined by the
institutional history in the area. Within districts, there was little marked difference (by design)
between our treatment (KDP) and control (non-KDP) kecamatan in terms of the institutional
lay of the land. Where there were differences, this was far more likely to be a function of the
results, while generally positive, are that many latent conflicts are often made more explicit, with dissenting villagers
having a new and legitimate institution through which to voice their grievances. However, there are few effective
mechanisms to deal with inter-institution conflict at the village governmental level.
114
Averaged across the two provinces (n = 67).
74
historical development of local cultures, and the broad processes of socio-economic change that
helped form them, than of any specific outside development intervention.115
It is an open question whether institutions still exist to deal with most problems and forms of
conflict in those parts of Indonesia which have been hit by extremely high levels of violent conflict.
The relative lack of use of KDP forums may be partially due to the nature of conflict in East Java,
a ‘low’ conflict province where social relations and institutions have not broken down to such an
extent that mechanisms such as KDP are used as a last resort for problem solving (Diprose
forthcoming). While we do not have comparative data on KDP and conflict management in any of
the highest conflict provinces, there is evidence that, at the district level, areas with lower capacity
to manage conflict use KDP more for this purpose.116 KDP forums were used more in areas with
low preexisting capacity to manage conflicts (i.e., in Manggarai and Pamekasan districts) than in
those with higher capacity (Sikka and Ponorogo). As we will explore more in Chapter 7, there is
evidence that where no existing institutions exist to deal with a particular type of problem, in some
cases KDP forums fill that role.
1.2 Resistance from Local Elites and Facilitators
A second reason why KDP forums are rarely used for managing non-program conflicts is resistance
from both facilitators and local leaders for them being used in this way. In a number of cases, when
people would bring up problems in the program’s forums that were not related to KDP, facilitators
or local government would say that the forums were not the correct place to discuss them. In many
instances, program facilitators were reluctant to sanction the use of KDP forums for non-project
related conflict resolution.
There are a number of reasons for this. First, in some areas such as Manggarai, facilitators were
averse to taking perceived “risks” by allowing discussion of conflicts. There are explicit sanctions
for people who contravene decisions made in KDP meetings; inappropriate behavior can result in a
kecamatan or village being excluded from participating in KDP in the following year. For this
reason, KDP officials tend to be extra careful and, therefore, try to avoid taking unnecessary risks.117
Second, facilitators and leaders often did not want to “mix KDP and non-KDP issues” and were
concerned that the forums may fail if they were to discuss complex and sensitive issues not directly
related to KDP and which involved people participating in the KDP forums (Didakus 2004).
They may have a point. Arguably, one of the reasons that KDP has been able to keep going in
many high conflict areas, such as in Aceh, during periods of high tension and unrest is its political
115
This is not to say that outside intervention has not been important. Indeed, policies of the central state, particularly
during the New Order era, did much to shape local institutional environments. In particular, Law 5 of 1979 went a long
way towards homogenizing village structures across Indonesia. Rather, outside interventions have had little effect in
determining difference in local institutional structures.
116
Although this does not hold at the village level (see Chapter 7).
117
This results not only in a reluctance to discuss contentious issues in KDP forums. We found examples in the field of
KDP officials pushing for economic productive funds to be allocated to those who were already reasonably well-off,
knowing that they were more likely to be repaid (Ghewa 2003). Creating an environment that strikes a balance be-
tween having clear rules (and sanctions for breaking them), but which allows for “altruistic deviance” from KDP staff,
is a major challenge.
75
neutrality. In the case of Aceh, both the Government and GAM (the separatist movement) supported
the program because it was seen as benefiting the poor people of the region; if the same forums had
been used to talk about conflict-related issues, it is unlikely they would have been as successful or
allowed to operate in the same way.118 While conflict issues are certainly not as sensitive in either
Flores or East Java as in Aceh, if particularly tense conflict cases (such as some of the inter-village
land conflict in Manggarai) were discussed, it would arguably negatively impact on the extent to
which agreement on other non-conflict issues could be reached in the forum.
While some problems are best not dealt with directly, there may be cases where forums can play
a positive role. In particular, conflicts related to development projects, or between individuals
(rather than groups) where the negative consequences of unsuccessful resolution are likely to be
less, could potentially be addressed in forums. Given the high success rates of KDP when used,
facilitators and local leaders could be encouraged to use KDP forums for discussing other
conflict-related problems in cases where other institutions do not exist to deal with them, and/
or where it is unlikely to affect the functioning of the program. That they do not do so, as we
discuss below, is a result of a lack of understanding about the positive roles KDP can play in
conflict management.
2. The Role of KDP Facilitators in Conflict Management
The KDP institutional structure involves having facilitators and staff at a number of levels, from
every participating village to a national secretariat in Jakarta (see Box 3.10). Third party intervention
can take place by choice (if an actor decides to intervene), by circumstance (if an actor happens to
be present when a conflict occurs), or because the disputants take the problem to a particular actor.
What role do KDP staff and elected facilitators play in helping to mediate conflicts that are not
related to the program?
Survey respondents were asked about the role played by KDP facilitators and staff in problem
solving/conflict resolution outside of KDP forums.119 The evidence from both the key informant
survey and the qualitative fieldwork shows that facilitators/staff do play a role outside of KDP
forums. In East Java, 52% of survey respondents said that KDP facilitators had successfully
resolved conflicts/problems outside of KDP forums; in NTT, 60% reported the same.120 However,
almost without exception, the conflicts resolved by KDP facilitators and staff outside the forums
118
We are grateful to Sentot Satria for discussions on this point. A casual discussion between Scott Guggenheim and
Michael Woolcock on the striking capacity of KDP to endure in such an unlikely setting as conflict-ridden Aceh was in
fact one of the sparks that launched the present study. More formally, see Barron, Clark and Daud (2005), which
outlines the extent to which KDP was in many places the only local development project operating in Aceh. Interest-
ingly, at the time of writing it appears that KDP will be used to distribute “reintegration assistance” to conflict-affected
persons in Aceh.
119
Their involvement in these activities within forums was covered above.
120
In East Java, 40.2% said facilitators did not solve problems outside of forums, with 7.5% saying they did not know
(n=132). In NTT, 38.1% said facilitators did not solve problems outside of forums, with 2.2% saying they did not
know (n=134). It is important to note that these figures refer to KDP facilitator success and not use. Informants were
asked if “conflicts or community problems [were] resolved or partially resolved by a KDP facilitator.” Thus these
figures do not include engagement that was unsuccessful.
76
related directly to the program. Whereas KDP forums do sometimes address conflicts that are
unrelated to KDP, staff and facilitators almost never take the role of conflict mediator for such issues.121
As Figure 5.3 shows, 100% of respondents in East Java, and 96.3% of respondents in NTT, who
had said KDP facilitators resolved problems outside of forums, said that the problems were related
to KDP. In NTT, KDP facilitators’ involvement in non-KDP conflict was isolated to two kecamatan,
Lambeleda in Manggarai and Talibura in Sikka.122
Figure 5.3: Were Problems/Conflicts Resolved by KDP Facilitators Outside of KDP Forums Related to KDP?
% of each Province
Province
Informants, both KDP staff and villagers, confirmed the findings. KDP facilitators hardly ever
get involved in directly mediating conflicts that are not related to the program. Why is this so?
Comparative analysis of the cases followed in the study provides insights into why different
actors play different conflict mediation roles, and why in some cases their attempts are successful
while in others they are not. Effective conflict mediators need to have three qualities—
legitimacy, willingness, and capacity—if they are to successfully intervene as third parties in
conflict. Examining the extent to which different KDP facilitators have these qualities provides
some explanation as to why they rarely successfully intervene in conflicts that are not directly
related to KDP.123
121
Across all types of problem (KDP and non-KDP), sub-district facilitators (FK) and village facilitators (FD) were the
most likely KDP staff to be involved in dealing with conflicts outside of KDP forums, unsurprising given their proxim-
ity to the communities they serve, and their facilitation mandate.
122
The rates of KDP facilitators addressing non-KDP conflicts may be higher than these results suggest. Respondents
who answered that facilitators did address problems outside the forums were only asked if the problem/conflict was
related to KDP or not. Respondents were not given an option of answering that facilitators addressed both KDP and
non-KDP conflicts. Hence, where facilitators addressed both types of conflicts, this may not be picked up.
123
This section draws heavily from the discussion of the “efficacy of intermediaries” in Barron, Smith, and Woolcock
(2004, pp. 30-32).
77
2.1 Legitimacy
For mediators to be successful, they must be seen to be legitimate. Two types of legitimacy are
of relevance. First, institutional legitimacy: the mediator must be seen to have legitimate
jurisdiction over the specific problem. In other words, it must be clear to all concerned that it is
appropriate for a given dispute to be adjudicated by a particular third party. Where a gap exists
between a given dispute and the (perceived or actual) jurisdictional mandate of a particular third
party, legitimacy is likely to be low. As we noted above, KDP forums are often not seen by
disputants and others in the community as being the appropriate places to deal with problems
that are not related to KDP, or, to a lesser extent, other development-related problems. The same
is true—indeed, to an even greater extent—for KDP facilitators. Facilitators, and other KDP
staff, are seen to have a mandate for dealing with problems relating to the program. However, in
almost all cases, the same is not true for other aspects of life and for other forms of conflict.
Other actors and institutions (government at its various level, the security sector, community
leaders, etc.) are deemed to have authority to mediate in disputes in such areas.
Second, in some cases the personal legitimacy of KDP staff was a factor limiting their role in
conflict mediation. In cases where the facilitators were not seen to be of high professional and
moral repute—and there were a number of these—people were less likely to take problems to them.
For example, in some cases KDP officials have lost their credibility and some of their authority
because of perceived corruption. In Lambaleda sub-district in Manggarai, the sub-district facilitator
(FK) is currently on trial, accused of corrupting approximately Rp. 80 million (USD 9,000).
Whenever the accusations started, this began to seriously affect his personal standing and authority
and hence had a negative impact on the program:
“Repayments began to dwindle after they found that the FK and the Financial Management Unit had corrupted Rp. 80
million of funds. If the community had wanted to pay their installments, where would they have taken them?”
Village Secretary, Satar Punda, Lambaleda, Manggarai, (M.512)
It is not surprising that communities will not see such individuals as having the legitimacy to
intervene in conflicts that are not related to the program. Another reason for the perceived lack of
legitimacy of facilitators and KDP staff was the lack of input that some communities had in
determining who their program representative was. Village facilitators (FD) are meant to be elected
in an open process; in most cases, FDs were indeed fairly elected but in some the process was not
seen to be legitimate:
“As for as who would be chosen as FD … it all depended on who the Village Head liked; there was no democratic
election.”
Teacher, Tengku Leda, Lambaleda, Manggarai (M.505)
This was also true for others involved in the program, for example local verification teams:
“The technical staff at the kecamatan (sub-district) confused those of us at the village level … Perhaps if the TTD
(Village Technical Staff) was from the village, or members of the community were appointed, the result would have been
much better. The TTD was taken from the kecamatan level even though they had already selected graduates from the
STM (Technical High School) from around here.”
TPK Member, Tlontohraja, Pasean, Pamekasan (S.615)
78
In other cases, such as in Desa Padellagan, in Pademawu sub-district, Pamekasan, the community
trusted the FD to the extent that he was also asked to run other development projects, as well as
village administration.124 We will explore this more in Chapter 7. However, it suffices to say that,
in general, local level KDP staff are not seen as having the institutional legitimacy—or, in many
cases, the personal legitimacy—to address conflicts that are not related to the program.
2.2 Willingness to Act to Seek a Solution
Second, effective mediators need to be willing to act constructively and proactively to help disputants
find solutions. As we noted above, in many cases facilitators did not feel it was their role to deal
with conflicts unrelated to the program.
In Talibura village, Sikka, there was an attempt by a villager to use the KDP forum at the village
level to discuss an unresolved land case. However, the FK said the KDP forum should only be used
for KDP matters. This is not to say his decision was not wise: the FK in question argued that KDP
was successful in Talibura—both in its operation and in building inter- and intra-community relations
(see next chapter)—because it remained a relatively neutral, ‘de-politicized’ forum. If the forum
also dealt with other conflicts/tensions, the shared acceptance of the legitimacy of the KDP process
would have been undermined. However, the result was that the problem was not dealt with within
the KDP structure and respondents reported that they were less likely to ask the facilitator to help
with conflicts in the future.
If facilitators are to be expected to play a direct role in conflict resolution (of non-KDP problems),
then they must be trained to play such a role. However, before proceeding with such an agenda,
serious thought needs to be given by the KDP design and implementation team about the potential
trade-off between possible enhanced conflict resolution and the potential impact on the perceived
neutrality of the program. No doubt this trade-off would be even more pronounced in areas where
conflict-related cleavages are more marked and all-encompassing than in our research districts in
East Java and NTT.
We should note here that in some cases facilitators were themselves part of the problem rather than
the solution. Some of the conflicts we followed revolved around distrust of program facilitators,
with savings and loans groups not wanting to repay their funds to the KDP Financial Unit but,
rather, wanting to store the funds in the village. In Ponorogo, when savings and loans groups could
not repay their loans, conflicts ensued between program facilitators and beneficiaries. In other
examples across all four districts, program facilitators became parties to disputes when they were
seen to be co-opted by elites. Furthermore, in some cases where facilitators sought to reduce
tensions between hamlets competing for funds, they were viewed as favoring one party or the other,
thereby heightening tensions.
This is one good reason why there should never be only one facilitator assigned to a conflict
region. It is necessary for there to be at least two facilitators (at each level), so that if one is
perceived as being biased, this can be ‘neutralized’ by the existence and performance of the other.
124
Ex-Village Facilitator (FD), Sumedangan, Pademawu, Pamekasan (1225).
79
2.3 Capacity
A third requirement for effective conflict mediators is that they possess the capacity to solve the
problem. Capacity has a number of elements. It involves having the resources—human, financial
and administrative—to solve the problem. It also includes the understanding necessary to be able
to effectively arbitrate a conflict in a specific context. This includes both technical understanding
(e.g., of conflict resolution skills) as well as contextual or local knowledge, what James Scott
(1998) has called ‘metis’.
KDP facilitators are embedded within an institutional structure which is relatively well funded
and capably staffed. If they need back-up, they can call on a range of colleagues at a number of
levels (sub-district, district, Regional Management Unit, national). In theory then, they should
have the resources to solve the problem. However, facilitators at different levels face difference
constraints which help explain their lack of capacity in dealing with non-KDP conflicts.
Sub-district Facilitators (FK)

At the sub-district level, time pressures are such that sub-district facilitators (FK) have little time
to be able to play an extensive role in issues not related to the project, especially if they cover
rural areas where considerable travel is necessary.
A more important weakness in some cases is their lack of local understanding. FKs are deliberately
chosen from areas outside where they work. In principle, this improves their perceived neutrality,
but it does mean that at times they are not sensitive to local conditions. In Tlontohraja village,
Pasean sub-district, Pamekasan, for example, the FK did not understand the local culture well and
hence did not coordinate with local community figures. This led to problems in the implementation
of the program. In Pasean sub-district, both villagers and KDP staff told us that it was difficult for
sub-district facilitators to perform their tasks well because the FKs were not from Madura and
hence could not understand certain aspects of Madurese culture:
“If the FK came from Pasean the program implementation would certainly be more effective because they could
understand the character of the community. Most of the FK come from outside the district so that many don’t feel at
home living here. Like FK Sidik who only stayed a few days in Pasean and after that said goodbye to go home to
Malang and didn’t return again…”
Development Coordinator (PjOK), Sana Daya, Pasean, Pamekasan (1248)
However, being from outside the area was not an absolute barrier to being accepted in the
community. The new FK in Pasean (Yuyun, Sidik’s replacement) made efforts to integrate herself
and to adapt to the local culture by changing her dress style:
“… FK Yuyun has begun to be able to follow the culture of the Madura people. In the beginning she never wore a
jilbab (head covering) when she came to the village, but after a while she began to wear a jilbab too. This is a very
effective strategy to approach the village community when carrying out program education and awareness-raising
because the community feels that the FK is not an outsider but has already become a part of their life by wearing
clothes similar to those worn by the people [here]…”
BPD Chair, Sana Daya, Pasean, Pamekasan (1259)
80
Yet local knowledge is complex, deep and diffuse. While Yuyun had made important symbolic
efforts to gain the trust of the local community, something that should not be underestimated,125 in
other aspects her understanding of the local situation was not as deep. Indeed, others complained
that Yuyun had not adapted to local conditions:
“… What Yuyun intended was to try and implement the procedures in the PTO (Operational Technical Guidelines),
but she hadn’t adjusted them to community conditions. In fact, the impression given was that she was forcing her
wishes arbitrarily onto others.”
TPK Chair, Tlontohraja, Pasean, Pamekasan (1250)
This is not to say outsiders cannot be accepted in the areas in which they work. Facilitators can
do a very good job, even when they are from completely different parts of Indonesia. But in
terms of sensitive issues such as the resolution of conflicts, those with local understanding have
a major advantage (Gibson and Woolcock 2005).
Village Facilitators (FD)

At the village level, facilitators have more time and generally have the local knowledge. However,
often they do not possess the technical skills to play a productive role in conflict resolution. Training
for FD staff is minimal; they receive little, if any, instruction in conflict management. Many of the
village facilitators are relatively young; this not only impacts upon the sense of authority they
command, but also upon their conflict management “wisdom”, a skill that is learnt through life. In
both provinces, the program contributed to a noted rise in the civic skills of village facilitators (FD)
through their experience in the program. In Padellegan village, the former FD has become Village
Head. In Sikka, another former FD said he had learned much from his position:
“Yes, my step into the BPD (Village Representative Council) was determined when I was still the FD. Before this I was
just an ordinary villager, just one of the little people.”
Former Village Facilitator (FD) and Member of the BPD, Loke, Paga, Sikka (S.1135)
Thus while facilitators may not be seen as legitimate or able to solve non-KDP problems in the
short run, their positions can help to build capacity and understanding in the long run. Indeed,
targeting ex-FD may be an appropriate strategy for finding appropriate people to support in
building village-level conflict mediation capacity.
Where facilitators were successful in solving problems outside of forums, survey respondents
cited the following reasons as being most important: they had the resources to solve the problem
(most important reason in NTT, second most important in East Java); they had the knowledge to
solve the problem (second most important in NTT, third in East Java); and they had the authority
to solve the problem (third in NTT, first in East Java).
125
On the importance of culture and symbols for addressing public goods (such as the maintenance of peace), see
Rao (2005).
81
3. Summary
In both East Java and NTT, the direct impacts of KDP on conflict management are minimal.
KDP forums and facilitators are rarely used for addressing conflicts unrelated to the program.
Where they are used, it tends to be in an ad hoc manner. In none of our research locations had
KDP been institutionalized as a regular conflict resolution device. There are a number of reasons
why KDP does not frequently get used for non-program problem solving.
First, other institutions exist at the village level to deal with conflicts that are not related to the
program. Where these institutions do not exist (for example in lower capacity areas, and for certain
forms of conflict such as that relating to other development projects), KDP forums and facilitators
are sometimes used. This suggests that KDP can act as a complement to existing conflict mediation
mechanisms. It may be that in the parts of Indonesia with the highest levels of violent conflict,
where many local institutions have collapsed, KDP can play a role in resolving certain kinds of
disputes, such as those relating to administrative procedures and resource allocation.126
Second, in some cases facilitators are perceived not to have the personal legitimacy to handle
disputes. In particular, where they are seen to have been corrupt, or to have unfairly favored one
group over another, it is unlikely that they will be called on to mediate contentious issues. For
program facilitators to be effective, they need to be seen to be honest, independent, and neutral.
Third, in many cases facilitators are unwilling to address conflicts and problems brought to them that
are not related to the program. In large part, this was because they are risk averse. Many facilitators
are scared of the sanction that might follow—for them and/or for the communities in which they
work—if they deviate from the PTO (Operational Technical Guidelines). There needs to be more
flexibility built in to these guidelines to allow for creative “altruistic deviance” by facilitators.
Fourth, in some cases KDP facilitators are not called on to mediate disputes because of gaps in
their capacity. These gaps differ by the position of the facilitator. Sub-district facilitators (FKs)
tend to have the technical skills needed to help mediate issues, but often they do not have adequate
time or the local knowledge necessary to understand the positions of the disputants and to win
their trust. Village facilitators (FDs), on the other hand, tend to have time and local expertise/
legitimacy, but they often do not have the technical training/education necessary.
As it stands, KDP is thus not an effective mechanism for working directly on non-program conflict.
In some ways, this is a good thing—it allows it to remain a politically neutral space where communities
can work out their needs and priorities. As we will see in the next chapter, it is in this space that social
and state-society relations can improve, leading indirectly to enhanced conflict management capacity.
However, at the same time, there is scope for modifying the program to allow it to more effectively
manage local conflict. When non-project conflicts are addressed through the program, they tend to be
resolved successfully. This suggests that there are possibilities for KDP to play a larger mediation
role, in particular for development-related disputes. Improved training for facilitators (in particular at
the village level), and increasing the discretion of facilitators implementing the program (in particular
at the sub-district level), would improve this aspect of the program. At the same time, it may be
necessary to have other complementary programs in place to directly address conflict.
126
This will be a key question for SPADA and those assessing its impact.
82
Chapter Six
Indirect Impacts of KDP on Conflict Management:

Changes in Group Relations, Behavior, and Norms
The previous chapter examined the ways in which the people and spaces associated with KDP
have a direct impact on social tensions and conflicts that emerge that are not related to the program.
We found that in most cases KDP had little direct effect on conflict management. However, there
are other ways in which the program may lead to changes in the level and impacts of local
conflict, and in how they are managed when they arise. This chapter will examine three ways in
which KDP may indirectly impact on conflict management capacity: through changing social
structures, forms of behavior, and norms and perceptions in the areas in which it works.
Development projects like KDP aim, albeit implicitly and in often unacknowledged ways, to
reengineer inter- and intra-group and state-society relations. KDP strives to empower villagers
by promoting participation in decision-making (particularly from marginalized groups), whilst
improving the quality of that decision-making through the promotion of greater transparency
and accountability. This necessarily involves attempts to transform local structures and norms,
which for our purposes collectively define the conflict environment of an area (see Figure 6.1).127
If KDP affects the underlying conflict environment, it will likely have indirect impacts (positive
or negative) on the likelihood that conflict will arise, the form it will take when it does, the ways
in which it is dealt with, and hence the potential for escalation.
In this chapter, we examine three different possible indirect impacts of KDP on local conflict
environments:
First, through the program’s effect on inter-group relations;

Second, through its effect on the political and social behavior of individuals and groups (and
thence its impact on intra-group and state-society relations); and,
Third, through the ways in which it impacts on people’s perceptions and normative frames,
e.g., villagers’ expectations, understandings and values with respect to violence, peace, and
collective problem solving.
We test for the presence and strength of these effects before looking for correlations between
these indirect effects and conflict outcomes. We use data from the key informant survey,
qualitative fieldwork, and newspaper archives. For the former, we bring in data from our ‘control’
locations.128
127
In this sense, programs like KDP are inherently political: they promote certain forms of behavior and cognitive
understandings, with the explicit intent of displacing other forms of social order.
128
We should note here how we are defining our control locations. Three out of four sub-districts are not pure controls.
Whereas these places had not received KDP when the qualitative fieldwork was conducted, by the time the second
round of the key informant survey was conducted (February 2005), they had had one cycle of KDP-II. (The
83
Figure 6.1 Direct and Indirect Impacts on Conflict and Conflict Environments
1. Group Relations
One factor that may play a role in determining levels of conflict is the nature of social interaction
between groups. In his now famous work on conflict and civil society in India, Varshney (2002)
argued that it was the extent of interaction between different identity groups—in his case
one exception was Sampung sub-district in Ponorogo, East Java, which had not received the program). This was
unfortunate but unavoidable given KDP’s scale. In many parts of Indonesia, almost all poor areas have now received
the program. In NTT, the only areas that have not received the program are rich or urban areas. As such, by late 2004
it was impossible to find ‘pure’ matches for our treatment sites. Nevertheless, we can still use these areas as control
locations for a number of reasons. First, we hypothesize that KDP effects are cumulative over time; hence by disaggre-
gating by years of KDP, we can get a sense of the impacts of KDP over time (indeed, the analysis below shows that, in
general, this hypothesis holds true). Disaggregating by years of KDP also evens up the sample sizes. Second, we argue
that one cycle alone of KDP is insufficient to have any real impact on social relations and structures. Given the slow
speed at which norms change, we cannot expect real changes in the first year. When we selected our treatment areas,
we chose only places that had KDP for two years or more. As such, the control site data does help us to ascertain and
isolate program impact.
84
Indirect Impacts of KDP on Conflict Management: Changes in Group Relations, Behavior, and Norms
between Muslims and Hindus, and, in particular, the extent to which such interaction was
institutionalized—that determined why some Indian cities were peaceful and others violent, despite
similar Muslim/Hindu population shares. The assumption is that attitudes and norms of behavior
towards other groups change (for the better) with frequent interaction; as groups which previously
saw their interests as separate begin to interact, opportunities arise for cooperation and collective
action resulting in common rewards. Hence individuals start to emphasize points of similarity as
they work together.129 Varshney’s work was conducted in urban India. In places such as rural
Flores, where pre-modern kinship systems continue to frame social relations between individuals
and between groups, opportunities for enhanced interaction may break down the rigidity of primordial
culture, religion and class structures.
Evidence from our study, as well as previous work on conflict in Indonesia, indicates that a large
proportion of violent conflict is group-based. Of all the violent conflicts we picked up in the
newspaper datasets, 50.6% had a group as at least one of the actors; for the most violent conflicts—
those that involved two or more deaths—72.5% involved at least one group.130 Even where conflicts
are between individuals, for example conflicts between two people over access to water, group-
based identities tended to determine in part whether or not the conflict would escalate. This is not
surprising; humans are social creatures whose lives are embedded in networks—social, political
and economic—from which they deduce their identities. As numerous authors (e.g., Horowitz
2000) have noted, group-based identity can be ascriptive (based on race, language, clan, caste, etc.)
or prescriptive (village location, political affiliation, etc.). In many cases these overlap. Where
different groups emphasize their differences, conflict is more likely.131
If KDP provides more opportunities and a regulated institutional framework for interaction
between groups which would not have taken place before, it may improve relations between
groups, hence making destructive conflict between them less likely. KDP seeks to encourage
two different types of interaction. First, intra-village interaction in group forums, sub-village
musyawarah (meetings), women’s forums, and in project implementation teams which involve
members of the community from different groups. Second, inter-village interaction, when
representatives from different villages come together to discuss and prioritize village
development proposals.132 Does the KDP process help to improve the relations between different
groups? And do these improvements lead to a reduction in levels of violent conflict?
129
Varshney is but one of a number of scholars who have started to draw attention back to how state-society relations,
and systems of social order, help in determining the propensity to which, and the mechanisms by which, individuals or
groups take violent action (see King 2004).
130
From analysis of KDP & CCN dataset.
131
Indeed the causality works both ways. Groups which emphasize narrow identities are probably more likely to
conflict with neighboring groups, because the emphasis on the particularity of a given identity in turn increases the
extent to which neighboring groups are seen as “different”. In turn, emphasizing difference can be a strategy to
mobilize a group for conflict, often for political or economic reasons. See Barron, Smith, and Woolcock (2004, pp. 21-
29).
132
See Figure 3.3, above, for the program cycle.
85
1.1 Forms of Difference in East Java and NTT
First it is necessary to outline briefly the different groups and group cleavages that exist in East
Java and NTT. Individuals in both areas identify themselves within a number of different groups.
In most of the villages in our study, groups are not completely polarized; unlike in other parts of the
country, interaction does take place in everyday life between individuals from different identity
groups. However, the different identity cleavages outlined are, at times, used as bases for group
mobilization, in some cases for conflictual purposes.
East Java
In the East Javanese research districts, there are a number of identity cleavages which underlie
social tensions and which on occasions get utilized when groups are mobilizing for conflict. Where
cleavages overlap, problems tend to be bigger. Though East Java (unlike NTT) is largely ethnically
and religiously homogenous, key cleavages which define identity groups include the following:133
geographical differentiation (hamlet, natural topography, location in reference to main road,

martial arts geographical base, etc.);
martial arts group (silat) affiliation;
competition between criminal groups (bajingan);
support of different Islamic clerics;
familial/kinship relations;
political allegiances to village and district elites;
state/non-state allegiances (that is groups in the community versus the state);
differentiation between rich and poor hamlets (often along occupational lines);
differentiation between program beneficiaries or between program beneficiaries and program
facilitators (for both state and non-state programs).
First, geographic cleavages are important. In the research areas, these cleavages tend to exist
at a low level. Thus people feel affiliation to their own hamlet (dusun) or village (desa); a
sense of group identity at the sub-district (kecamatan) or district (kabupaten) level does not
exist. In some cases, geographic identities relate to topographical rather than administrative
boundaries. For example, Slahung sub-district in Ponorogo is divided into two distinct regions:
mountains and plains. In Pademawu sub-district in Pamekasan, there are distinct regional
identities based on local topographical conditions, with the district split into a coastal area, a
dry field area, and an urban area. In Slahung sub-district, Ponorogo, geographic location has
become an identity marker as result of development policy gaps between areas, within and
between villages (Anggraini and Rasyid 2004).
Second, gang cleavages are also important in both districts, although the nature of the gangs
differs. In Ponorogo, martial arts (silat) groups are extremely important markers of identity. The
silat groups in the Ponorogo area have their genesis as organizations of resistance against the
Dutch in the early twentieth century. The two dominant groups (SH Terate and SH Winongo)
133
These are outlined in greater detail in Diprose (forthcoming). See also Annex A for more detail on the contexts
(political, economic, cultural) of the research sites.
86
stem from the same organization, Setia Hati, which split in the late 1940s. The groups are driven
by strong quasi-mystical ideologies, packaged in a way that relates strongly to kejawen (syncretic
Javanese Islam) spirituality with which the community is already familiar. Leaders justify the
existence and actions of the groups in spiritual terms. The triggers of silat battles are often cultural
in nature, for example traditional campursari (Javanese dance) performances. The groups are
particularly strong in marginalized villages, were locals have unfulfilled hopes and expectations;
the martial arts groups, and the cultural ideologies they embody, provide a social outlet which can
free them from the otherwise rather mundane routine of their everyday lives (Probo 2003b).
Affiliation to the groups maps on to geographic identities; Winongo has its base in towns, whereas
Terate is stronger in more rural areas in outer Ponorogo.
In Pamekasan, membership of, or relationships with, different mafia-like criminal groups (bajingan)
provide a strong identity marker. In many villages in Pamekasan, bajingan have gained credibility
as problem solvers and ‘enforcers’. They have relationships to Village Heads and to Kyai (Muslim
clerics), who need the bajingan to maintain law and order (Ashari 2003).
Third, political allegiances are important. This includes affiliation to political parties as well as to
particular individuals at the village or district levels. As discussed earlier, village head elections are
often extremely tense, in part because candidates have spent large amounts of money on campaigning,
and in part because they do not want to “lose face” by suffering defeat, especially in Madura (Box
6.1). At the district level, political allegiances tend to be to particular figures (e.g., candidates for
the district head position) rather than parties per se (Barron, Nathan and Welsh 2005).
Box 6.1: The Village Head Elections in Palengaan Laok
In March 2000, the Village Head elections became heated. Prior to the elections, a rumor circulated that the winning
candidate had false educational qualifications, so the candidate’s campaign team mobilized the mass of supporters,
including bajingan, to protest at the sub-district office. They said that if his qualifications were not verified, then someone
would be murdered. The situation was so heated that his qualifications were verified quickly.
The votes were counted on election day and the candidate accused of having false certificates, and who also had the
backing of one of the senior Kyai in the region, won. During the morning of election day, symbols for the party of the
winning candidate had appeared around the village. This contravened election law. Thus, the four losing candidates
wrote a letter of complaint to the District Head, copying it to the other government and security agencies, and even
to the Head of the District Court. But there was no answer. Receiving no response, the losing candidates mobilized
500 people and protested at the district office. When they finally met with the District Head, he stated that there
were no regulations prohibiting the use of party symbols on election day (even though the District Head had signed
the district regulation which stated the prohibition), and that the disputants had not registered their complaints with
the election committee.
“ Indeed, there was already a syndicate agreement between the District Head, the parliament, and the Kyai
[who supported the winning candidate] to resist the complaints … Until Kyai Z [another senior Kyai in the
region] intervened at the time and appealed to the District Head that what was ‘true should be endorsed and
what was wrong should not’…”
BPD Chair, Palengaan Laok, Palengaan, Pamekasan (7)
Eventually the District Head held a meeting with Muspika (the district representatives from all the security and government
agencies), where the only person who disagreed with the District Head was the Head of the District Court. The losing
candidates employed a lawyer and asked the police to investigate, but the investigation was half-hearted, they had little
evidence to take to court, and hence they did not proceed.
87
Fourth, differences between the rich and poor, and between rich and poor hamlets (dusun), are a
pronounced cleavage in some areas. Villagers are readily able to identify who is poor and who is
rich within villages.
In both Pamekasan and Ponorogo there is not great ethnic or religious heterogeneity and hence
while both of these are important components of people’s identities, they do not constitute significant
markers of “us-them” identification. All the kecamatan in East Java where research was conducted
are almost 100% Muslim. However, in certain areas such as Kecamatan Janangan in Ponorogo,
affiliations to different religious organizations are important:
“… Once there was another Islamic group that came to this village. This mosque [pointing to the mosque in front of
his house] was about to be occupied by Islam Jaulah, you know the ones who have a tradition of going out for a few
days each month to spread religion. So one of their groups from somewhere once stayed at the mosque here, but
the community here opposed them. This is a public mosque; it’s an NU mosque, 134 what’s more. How could they
want to take it over like that?”
Village Secreatry, Panjeng, Jenangan, Ponorogo (1313)
It is also important to note that there are other salient identities in East Java which form along
more prescriptive lines. In the analysis below, we address them in the section on relations between
‘other’ identity groups.
NTT
In Flores, the most noticeable form of group identity is ethnic in nature. Ethnic diversity is
extraordinary: one survey estimated that there are 28 different ethno-linguistic groups in Flores,
and 61 in NTT province more broadly (Grimes et. al. 1997). Ethnic identity works at a number of
levels. At a reasonably macro level, five main identities exist that correspond with the district
boundaries on the island: Manggaraian, Endenese, Ngadanese, Sikkanese, and East Florenese. Within
each of these, different tribes exist. For example, in Manggarai district (our ‘low capacity’ district
on Flores), sub-district borders map on to the boundaries of different suku (ethnic groups). Yet the
continuing strength of clan and kinship systems has meant that ethnic differentiation takes place at
a very local (particularistic) level: 29.4% of the people in NTT province report themselves as being
of an ethnic group which has less than 109 other people in the province (Suryadinata et. al. 2003).
Religious identity is also an important marker. While Catholicism is the dominant religion within
Flores, unlike in East Java other religious groups make up a considerable share of the population.
In some villages, like Paga in Sikka, indigenous inhabitants tend to be Catholic, whereas newcomers
are more likely to be Muslim, Hindu or Protestant.135 Local terms have been coined to distinguish
Muslims from Catholics.136
134
Nahdlatul Unama (NU) is one of the largest Islamic mass organizations in Indonesia. It has over 35 million, and
represents an important segment of the progressive modernist Islamic movement (Hefner 2000).
135
Didakus (2004).
136
Muslims are generally referred to as ‘ata nggobhe mite’, whereas Catholics are ‘ata serani’. The former literally
means ‘to wear a black hat’. The latter has its roots in the word Kristiani (Christian). Local terms for other religions
have yet to be developed.
88
Class and vocation are also salient identity markers. Groups exist that represent the interest of
different occupational groups such as fishermen, farmers, and so on. However, in some of the more
remote villages in Manggarai, wealth differentials are smaller.
In most areas in Flores, political affiliation is not particularly important to group formation. Even
more so than in East Java, parties command little loyalty from villagers. Indeed, many of the
candidates for the April 2004 local legislative elections had previously stood under different
parties. It was not uncommon to see villagers in Catholic villages donning the t-shirts they had
received for free from Islamic parties such as PAN. Indeed, PAN commanded a significant share
of the vote in some homogenous Catholic villages. The reason that people had voted for the local
PAN candidate, we were told, was that he was a respected community figure and was seen as
having the ability to bring money and projects back to the village. Indeed, patronage politics is
strong in Flores, with candidates expected to bring back resources (legally or illegally) to the
communities who elect them. This leads in turn to local ethnic politics, where particular suku
(clans) support “their man” in order to ensure that they have access to district-level decision
making and the benefits it confers.137
[Q. Why did candidates stand in the election?] They were in the party because there was an opportunity offered by the party.
The problem is that the candidates didn’t know the ideology of the party … people are easily mobilized. The primordial criteria
is still there – in conflicts between parties, candidates, and in other areas of conflict like land conflict. The parties and
candidates manipulate those feelings to get elected…”
Vice-Principle, Ledalero Seminary, Maumere, Sikka
Unlike in East Java, and much of Indonesia, there does not seem to be a strong gang culture.
Incidents of fighting between individuals do not tend to escalate into broader inter-group disputes.
“They generally resolve the fights themselves.”

Deputy Head of the LKMD, Tanah Rata, Kota Komba, Manggarai (M.14)
“The cases involving youths are generally triggered by drunkenness. They get drunk, they fight and then once they are
sober they are reconciled.”
Head of a Catholic Prayer Group, Tanah Rata, Kota Komba, Manggarai (M.16)
“The fights are generally triggered by alcohol. They will normally hug and make peace once they are sober again. I even saw
some of them crying because they regretted their actions.”
Farmer, Koting A, Maumere, Sikka (S.52)
Unlike in other places in Indonesia, the evidence seems to show that disputes fuelled by the
consumption of large amounts of alcohol in Flores do not transmute into broader conflicts and
worsen relations between groups.138 The reason is that the youths tend to come from the same
group, they live together in the same area, and they know each other. Flores does not yet have a
prominent gang culture. As such, small incidents do not harden pre-existing cleavages, and those
who get into drunken fights tend to recognize them for what they are.
137
See Barron, Nathan and Welsh (2005).
138
This is very different to evidence from elsewhere. Barron and Madden (2004) show, in the case of Lampung, how
drunken clashes often develop into larger clashes and how they lead to worsened relations between groups. There is
evidence that this also takes place in other areas such as West Kalimantan (Parry 2005).
89
1.2 The Impact of KDP on Group Relations
How does KDP impact upon the different forms of constructed group difference we have just
outlined? A number of informants claimed that the program had led to increases in trust between
different groups:
“Yes, KDP has had positive effects … the community has been able to mature. Also, we can see that since the implementation
of KDP a feeling of unity has been cultivated. Perhaps this is because people within the groups help each other. There is a
feeling of closeness [between them].”
Village Facilitator (FD), Pagal, Cibal, Manggarai (M.1004)
“…For instance, when we were working together to build this road, people who could afford to gave cigarettes to those without.
That can strengthen relationships…”
Krajan Hamlet Head and TPK Chair, Kemiri, Jenangan, Ponorogo (1306)
“…Problems between martial arts groups did not affect the KDP program; in fact there are members of silat groups who
participated in the program, for instance in Pintu village and Kemiri village.”
Development Coordinator, Kemiri, Jenangan, Ponorogo (1309)
Informants noted two different mechanisms by which levels of interaction are increasing. First,
musyawarah (consultative meetings aimed at moving forward through generating consensus)
are increasingly rare in many of our research villages. The reasons for their demise vary by
region. In Sikka and, to a lesser extent, in Manggarai, Flores, the encroachment of the Indonesian
state down to the village level during the Soeharto era, accompanied by explicit attempts to
‘normalize’ and standardize (i.e., Javanize) governance structures across the archipelago has in
many ways ‘modernized’ village life.139 Accompanied by profound social changes—urbanization,
increased population movement, etc.—this has meant that the role that community forums play in
day to day life has in many cases declined. While multi-stakeholder forums such as the Rakorbang
do now exist at the district level (though with varying degrees of success; Ashari 2005), few
opportunities exist at the sub-district level for collective decision-making and problem solving.140
Thus inter-village musyawarah, in particular, are very rare. KDP, in establishing a decision-making
forum with authority to allocate resources, fills a vacuum.
“Competition has positive effects … the community in a sub-district can all come together and learn the method and
process for obtaining assistance. So they get to know the community out there [meaning the communities in other
villages].”
Village Secretary, Panjeng, Jenangan, Ponorogo (1313)
The qualitative findings from East Java highlighted that many of the interactions in the villages
visited were intra- rather than inter-group; shared interests tended to exist within groups rather
than across groups (Diprose forthcoming). There were three key exceptions in the sub-districts
139
These changes are actually a combination of legislative/political action and broader trans-national processes of
change (globalization, modernization, etc.) Of the many legal tools used to these ends, Law 22/79 on village gover-
nance was probably the most important (see Antlov 2001; Evers 2000).
140
The kecamatan (sub-district government) is only an arm of the district government apparatus, with no autonomous
authority or decision-making power.
90
visited: the village level meetings (musyawarah, although these usually only involve elites);
KDP forums and implementation activities at both the village and sub-district level; and higher
level ad hoc meetings (set up by leaders to resolve problems) which became repeated and
institutionalized but could potentially dissolve with a change of leadership.141
Second, and less directly, KDP facilitates group interaction through improving transportation
networks. This may seem a tenuous link, but it was one brought up again and again by informants,
particularly in Manggarai where the minor road network is almost non-existent. Indeed, analysis
of the non-KDP development projects that are going to our villages, as well as of the APBDes
(Anggaran Pendapatan dan Belanja Desa – Village Budget), shows that KDP is by far the biggest
player in terms of small-scale local road construction.142 Where transportation networks are poor,
KDP was cited as having improved interaction:
“KDP has been very helpful for the community, especially in terms of transportation. People never dreamed that they
would eventually be able to pass through Desa Lando in a motorized vehicle. The problem was that the land there is
so steep and rocky. But after the implementation of KDP in 2000, a road which could be used by motorized vehicles
was built.”
Villager, Lando, Cibal, Manggarai (M.1063)
“KDP has improved transportation. With improved transportation the villagers have become more mobile and, in turn, interaction
has increased.”
Kecamatan Facilitator (FK), Ruteng, Manggarai (M.1030)
“KDP has had a positive impact in this kecamatan. Now, the isolated regions can be reached by road. This has allowed the
community to market their produce. Also, the people’s economy has strengthened. Similarly, it has become a lot easier for the
community to interact because they are a lot more mobile…”
UPK Chair, Pagal, Cibal, Manggarai (M.1002)
But can we link increased interaction to improvements in inter-group relations? It could be that
encouraging groups with different identities (and interests and value systems) to interact could
fuel tensions, particularly if the environment in which such interaction is taking place is a competitive
one. Data from the key informant survey, backed up with cases from the qualitative fieldwork,
suggests that relations between groups across a range of identity cleavages are improving. And
comparative analysis of our treatment and control locations shows that KDP is playing a role in
driving such change.
141
On the latter, see Ashari (2005).
142
Take for example the case of Cibal sub-district in Manggarai. According to data obtained in the field, 23 small roads
in 17 villages (of 27 total) were funded over the four years that KDP was operational. Over the same period, informants
knew of only one other project, P2JD (Proyek Peningkatan Jalan Desa – Village Road Upgrading Project) that focused
on local infrastructure. The government project improved roads in only three villages in the sub-district.
143
In East Java, 78.68% of the population is Javanese and 18.07% Madurese. No other group makes up more than 1%
of the provincial population. See Suryadinata, Arifin, and Ananta (2003).
91
Ethnic Relations
Ethnicity is an important marker of difference in NTT, less so in East Java. This is not surprising
given the heterogeneity of self-reported ethnic groups in the former compared to the latter.143 In
NTT, the majority of informants in KDP treatment locations (58%) reported improvements in
relations between ethnic groups since KDP arrived.144 Survey informants most commonly stated
that it was the inclusion of more ethnic groups in decision-making compared to the past that had
led to improved relations. 89.6% of those informants who reported improvements said KDP
had played a role.
It should be noted that this is perceptions data and that there may be in-built biases, for example
in the incentives of KDP staff and elites to favorably report on the program. However, if we
bring in information from our control villages in Flores, and disaggregate responses by the number
of years an area has had KDP, the evidence appears to strongly back the claim that KDP improves
relations between ethnic groups over time. As Figure 6.2 shows, the extent of positive change in
ethnic relations steadily increases as KDP stays in a location for longer periods. In the one-year
KDP villages, 38.3% reported improvements in ethnic relations.145 For two-year villages, 50%
reported improvements; for three-year villages, 62.8% reported improvements. In villages that
had had KDP for four years, 68.8% of informants reported that ethnic relations had improved in
that time period. Similarly, the percentage of people reporting that KDP had improved ethnic
relations “a lot” increased over time.
Figure 6.2 Changes in Ethnic Relations in NTT by KDP Years
Source: Key Informant Survey (treatment and control sites)
144
26.1% said relations had improved a lot, 32.2 % said they had improved a bit, and 36.5% noted no change. 2.6%
reported a worsening of relations, and 2.6% said they did not know if there had been any change. Sample size (n =
115).
145
Informants in treatment sites (i.e., two-, three- and four-year KDP villages) were asked for changes since KDP came
to the area. For informants in control sites (i.e., zero- and one-year KDP villages) the time period in question was
derived by taking the rounded mean number of years of KDP in the research locations in that district. Thus, in Pamekasan,
informants were asked for changes in the past three years; in the other three districts, they were asked for changes in the
past two years.
92
This suggests three things. First, ethnic relations are improving across all our research locations
in Flores. Only six informants (3.7%) reported a worsening of ethnic relations. This correlates
with the drop in the number of conflicts with a reported ethnic basis over the 2001 to 2003
period.146 Second, KDP is playing a role in driving such change. Reported rates of improvement
are markedly higher in areas that have received KDP for more than one year.
“The implementation of KDP, through the coordination of groups, has increased levels of trust between different clan,
ethnic and religious groups as well as between the rich and poor.”
Sub-district Head, Ruteng, Manggarai (M.1027)
Third, this improvement increases over time. The longer KDP operates in an area, the greater
impact on ethnic relations it has.
Religious Relations
Similarly, KDP appears to have led to an improvement, albeit a less marked one, in relations
between religious groups in NTT.147 In our treatment sites, of those who had an opinion, more
than half (or 39% of all respondents) noted some improvement in religious relations, with the
rest saying religious relations had stayed the same. Of those who reported favorable change in
the treatment sites, 89% attributed it to KDP. Again, the fact that more groups were involved in
decision-making than before was the most commonly attributed reason for improvement in
relations between groups.
Bringing in the control (one year of KDP) locations reveals a marked difference between the
treatment and control locations. However, the effect of time seems less marked: only 14.9% of
informants in the one-year villages reported an improvement, compared to 39.1% of informants
in the treatment villages.148 Having said that, the percentages reporting improvements was fairly
similar across the two-year, three-year, and four-year KDP villages (see Figure 6.3).
146
KDP & CCN dataset. Of violent conflicts in 2001, 31% had an ethnic basis (22 of 71); 19.3% did in 2002 (16 of 83),
and 11.0% did in 2003 (8 of 73).
147
Not surprisingly, in our religiously homogenous East Javanese locations, informants said either that KDP had not, or
they did not know if KDP had, made any difference to relations between different religious groups.
148
Sample sizes: one-year villages (n=47); other villages (n=115).
93
Figure 6.3 Changes in Religious Relations in NTT by KDP Years
Class Relations
In both provinces, the majority of respondents (with opinions) indicated that there had been
improvements in levels of trust between the rich and the poor. In East Java, 38% stated this,
whereas in NTT 55% noted some improvement. When asked why relations between rich and poor
had improved, informants in East Java most commonly cited the presence of more bridging forums;
in NTT the most popular reason was that more groups were involved in decision-making than
before. In East Java, 93% of respondents who reported a favorable change said that KDP was
responsible; in NTT, 95% reported the same.
“I think that KDP is indeed prioritized for the poor. Relations between business groups increased, for example the
tempe [fermented soya bean cake] maker stored their products at the warungs [food stalls]. Relations between the rich
and poor, and between religions, also improved…”
This is an important finding, given that in the research areas rich-poor cleavages either existed
between occupational groups within villages, between elites and other members of the community,
or between hamlets. These three major cleavages were prominent again and again in the cases
followed in this study.
However, whereas in NTT improvements are closely correlated with the length of time the program
has been in an area, the same is not true in East Java. In NTT, there is a noticeable trend of
improving levels of trust by years of KDP; in East Java, the pattern is less clear, although there
appears to be larger positive changes in non-KDP areas (see Figure 6.4).
94
Figure 6.4 Changes in Class Relations in NTT and East Java by KDP Years
Political Affiliation Relations

KDP appears to have little impact in our research areas on relations between people with affiliation
to different political parties. Most informants in East Java, and all in NTT, noted that they did not
know if there had been changes in levels of trust between people with affiliations to different
political parties. This is largely because political parties are neither pervasive at the village level
nor connected to the KDP process.
“There was no influence from the political parties on KDP; nor, on the other hand, did KDP influence relations between political
parties. Political parties only appear towards the general elections. Outside they do not appear at all. It seems they only work
at the time of the general elections …”
Another informant said that relations had improved between political parties but that KDP had not
influenced this. In his view, improved relations between political parties was due more to community
awareness that no matter what political party was in power, it would not provide a positive influence
on the lower levels of society:
“In my opinion there’s been an increase in relations between political parties, but this hasn’t been due to KDP; in fact, there has
been no relationship between the political parties and KDP … This increase occurred because the community here has
become aware that no matter who leads, the common folk remain just that, the common folk. People who say they’re going to
help the ordinary folk, they forget their promises once they gain a seat. So for the people here it’s simple really. If political
parties want to give aid, they’re voted for, if not, then not … Relations between silat groups have increased ... there’s been no
influence from the KDP. The community and figures have become aware that the presence of groups increases village and
hamlet security. Before perhaps indeed there was a problem because these groups almost all arrived here together.”
BPD Chair, Wates, Slahung, Ponorogo (1334)
95
Relations between Silat Groups

As we noted above, identities in East Java and NTT are multi-faceted and polyvalent. A range
of other identities are important. In particular, affiliation to silat (martial arts group) is of
importance in Ponorogo district.
In East Java, seven survey respondents noted that the relationship between silat groups had
changed since KDP had been in operation. Of these, three (43%) noted an improvement in
relations, and only one said that the relationship between silat groups had worsened. However,
the primary factor was not KDP. Indeed, no survey respondent attributed the improvement in
relations between groups to the program. The qualitative fieldwork revealed that the primary
reason for the decline in silat violence in Ponorogo was the efforts of the District Police Chief,
who in 2001 began holding regular meetings between silat group leaders. In turn, group leaders
agreed to hand over troublemakers to police for prosecution. This approach was successful
because silat groups are highly organized and hierarchical, and once senior silat group figures
took ownership of the problem they were able to use their influence within the system. Other
theories exist. One goes that the District Head, Markum Singodimedjo, himself an honorary
member of the SH Terate silat group, deliberately stirred up silat violence in preparation for
the 1999 district head election, in order for him to present himself as the strong man capable of
handling the situation. According to this theory, silat conflict was ‘allowed’ to die down once
he had been elected (Probo 2003b).
Nevertheless KDP appears to have helped build relationships between silat groups, even if it did
not trigger or primarily drive such processes. In Jenangan, Ponorogo, members of silat groups
were involved in implementing KDP projects at the village level.
“…Problems between martial arts groups did not affect the KDP program. In fact there are members of silat groups
who participated in the program, for instance in Pintu village and Kemiri village.”
Sub-District PMD Head, Jenangan, Ponorogo (1303)
Cases of conflict between silat groups occur frequently throughout Ponorogo and few bridging
forums exist between these groups. However, members of silat groups were able to become
involved in KDP, albeit on a personal rather than institutional level (Rasyid 2004).
2. Behavioral Change
A second indirect mechanism by which KDP may affect change is through its impact on people’s
behavior. KDP is in many ways a democratization initiative masquerading as an anti-poverty
project. It aims to deliver small-scale infrastructure to marginalized communities. However, the
mechanisms through which it does so are aimed at empowering communities, socially and politically
as well as economically. Indeed, the program is conceived by its auteur as a way “to trigger and
support processes in which villagers exercise discretion in solving self-identified development
problems” (Guggenheim 2006). Villagers are encouraged to participate in development project-
related decision-making. The aim is that this will lead to changes in non-project behavior, in the
process indirectly strengthening the state.149 Attendance and participation in KDP may make
149
An important critique of community-driven development projects like KDP is that they set up parallel structures to the
state and, in doing so, undermine it (see Tendler 2000). This raises vital policy questions. In the Indonesian context,
96
other forms of participation (e.g., in other elements of village governance and decision-making)
more likely. Are there any spillovers in terms of quality of governance? And is there a link between
more democratic decision-making and improved conflict management?150
At the macro level there appears to be a case for the latter. Cross-country evidence has shown
that democracies are less likely to experience civil war than authoritarian states (Hegre et. al.
2001). On the other hand, while democracies may be more likely to channel conflict in peaceful
ways, democratizing countries—i.e., those in transition—may not (Gurr 2000; 2001). A vast
literature has explored the links between conflict and societal transition.151 However, while a
rich empirical and theoretical base now exists for understanding national level transitions, and
their links to conflict, there is a lot less evidence on whether and how these processes play
themselves out at the local level.
In the literature on Indonesia, it is now commonplace to link the transition of 1998 to the outburst
of violent conflicts that occurred across the archipelago.152 The New Order era (which ran from
Soeharto’s accession in 1966 until his downfall in 1998) provided development and security during
most of its tenure, but at the expense of civil society institutions. Security sector institutions (the
police and military) and formal government institutions (at the village, sub-district, district, provincial
and national levels) were eroded by corruption, distorting the way in which decisions were made
and resources (both material and power) were allocated, to the point that, over time, the very basis
of their legitimacy was undermined. The collapse of the New Order left a conflict resolution
vacuum, with formal institutions not trusted or able to manage problems, and civil society too weak
to take up the slack. It also left different groups with a number of grievances and differential access
to the state (Bertrand 2004). The result was an institutional environment unable to address the
needs of many ordinary Indonesians. In many cases, groups who perceived themselves excluded
felt that there were few non-violent avenues for redress.
To a lesser extent, the political decentralization implemented in 2001 has also been attributed as a
factor in explaining outbursts of violent conflict.153 Decentralization provides both new opportunities
and threats. It offers opportunities for improving accountability and the delivery of services at the
local level. This, in turn, can mean that grievances can be addressed, and conflict more effectively
managed when it arises. At the same time, decentralization has created an environment of flux,
where groups battle for power and resources in a climate of contested and changing rules. With
when KDP was designed in 1997, it deliberately by-passed the state because: (a) the system of patronage and corrup-
tion that characterized the Indonesian state prevented the delivery of services to the local level; and (b) it was felt that
it would be easier to reform the state from the outside—e.g., by creating visible examples of processes that worked, and
hence demand for similar structures within the state—than by tackling state corruption directly (see Guggenheim
2006). Now that Indonesia is further along its path of democratic transition, questions arise as to how, and the extent to
which, projects like KDP should link in to formal state structures. These are questions that are being debated exten-
sively within the KDP and Social Development teams in Jakarta. The third phase of KDP (KDP-III) tackles this
question, as does the new SPADA project, which explicitly links bottom-up CDD processes to district-level decision-
making.
150
This is the hypothesis that informs the 2002 UNDP Human Development Report (UNDP 2002).
151
See, for example, Polanyi (1944), Moore (1967), and Bates (2000).
152
See, for example, Tadjoeddin et. al. (2001), Tadjoeddin (2002), Bertrand (2004), and Varshney et. al. (2004).
153
See International Crisis Group (2003, 2005), McCarthy (2004), and Cribb (2005).
97
more resources at the district level, there are incentives for local elites to attempt to capture decision-
making processes. Too often, corruption has been merely devolved rather than addressed (Bjork
2003; Aspinall and Fealy 2003; Morishita 2005).
Yet how do these processes play out locally? And how do external interventions interact with
ongoing processes of change? In the sections that follow, we will first briefly outline some of the
dynamics of local level political change in the post-1998 period. We will then examine the empirical
evidence regarding the extent to which such processes of change have impacted upon the political
behavior of villagers in our research sites. Finally, we will attempt to ascertain the extent to which
such changes are solely a function of changing context, or whether we can reasonably attribute
some impact to KDP.
2.1 The Dynamics of Local Level Democracy in Indonesia
Much has been written on the impacts of democratization and decentralization on Indonesian political
culture and practice. In April 2004, Indonesians voted for individual legislative candidates at the
national, provincial, and district levels, for the first time. This was followed by the election of the
national President in October of that year, and then the first direct election of District Heads (Bupati)
in mid-2005. The seeming smoothness with which these processes took place led one prominent
observer to remark on “Indonesia’s quiet revolution” (Rieffel 2004).
In terms of decentralization, attention has focused largely on changes at the district level. This is
understandable: district spending now accounts for 31% of total government expenditures, up from
17%, and over 2 million of 3.3 million civil servants have been transferred to the districts. However,
democracy and decentralization have also had a great impact at the village level. Institutional
changes have created new opportunities; these, when combined with other factors such as a free
and change-oriented press, have helped create cultural and behavioral changes.
Antlov (2001, 2003) outlines the institutional changes. Law 22/1999 on regional governance replaced
Law 5/1979 on Village Governance. This prompted a vast array of changes. Village Heads could
now be called by traditional names. The appointed Village Consultative Assembly (LMD) and
Village Community Resilience Board (LKMD) were replaced by an elected Village Representative
Council (BPD), in effect a local legislature separate from the local executive (the Village Head).
The BPD has the power to approve the village budget, previously a task of the district legislature,
and they can propose the dismissal of the Village Head. Villages are allowed to raise local sources
of funding.
These developments have vastly altered the dynamics of village political life. There is now a real
counterbalance to the Village Head. Indeed, in many cases there is genuine political competition
between the two village-level political power blocks: the Village Head (and his supporters) and
the Village Council (BPD, and its supporters). Given that the BPDs have numerous members
(normally nine), ordinary villagers have greater access to points of political leverage. The research
found that BPD membership includes a wide range of people within the village, including defeated
98
Village Head candidates, traditional leaders, and young activists. Political power is thus diffused
among different groups within the village.154
Democratization has also led to cultural changes, with villagers far more willing to participate in
local political life. The research found that communities are increasingly willing (and able) to
protest perceived wrongdoing from their political representatives. The protest culture that has
emerged in the post-1998 period has percolated down to the village level; in many villages, mass
protests against misuse of funds are now common. In Nele Wutung, in Sikka, Flores, for example,
villagers held demonstrations in front of the DPRD-II (district parliament) and the Bupati’s (District
Head’s) office in Maumere, in order to protest the corrupt practices of the village head.
“The whole community protested and reported the case to the Bupati (District Head). However he wasn’t able to
resolve the problem so it was taken to court.”
FGD with male group, Nele Wutong, Maumere, Sikka (S.09)
“The villagers went straight to the police. They took the rice with them as proof…there were throngs of them [villagers going
to the police in protest].”
Former KSP (Saving and Loan Group) Chair, Nele Wutung, Maumere, Sikka (S.17)
There is also evidence that villagers are getting more politically savvy. A study of the 2004 legislative
elections in four provinces found that money politics was still strong (Barron, Nathan, and Welsh
2005). Candidates and parties would give potential voters gifts such as ikat clothes, rice, and
generators. Villagers would gratefully take the gifts, yet this did not seem to affect voting choices
on election day. “We take what they offer,” one villager in Sikka told us, “and then vote for who we
want to anyway.”
It is into this context of flux, and uneven instances of political empowerment, that KDP, a program
with democratization aims, and other development programs, enter.
2.2 The Impact of KDP on Local Democracy
How does KDP affect these processes of democratization? We examine two indicators of behavioral
change. First, we look at differences in the number of groups who are participating in village
meetings over time, and differences between those participating in KDP meetings compared to
others. Second, we examine changes in village-level decision-making. Given the changing context
outlined above, it is extremely difficult to separate out the influence of KDP from other contextual
effects. Is an observed effect attributable to the program or to the broader constellation of events
that have marked Indonesia’s transition? The inclusion of ‘control’ locations does allow us some
leeway to disaggregate program effects from the broader processes of social and political change
that have been sweeping Indonesia over the past seven years. We find that, in general, positive
changes in local democracy are observable across all our research locations, but that they are more
marked in KDP areas.
154
A forthcoming volume will bring together our case studies on village politics and conflict. It should be noted that a
movement has developed to limit the degree of autonomy that can be exercised at the local level. Law 32/2004, dilutes
some devolution measures (e.g. BPDs will now be appointed rather than elected), while strengthening others (e.g.
provisions for the direct election of Bupati). It is still unclear how changes to the role of the BPD will play out, or even
if they will be fully implemented.
99
Changes in Participation
Informants in KDP areas noted that people come to KDP meetings who do not usually get
involved in other village meetings: 60% of informants in East Java, and 81% in NTT agreed.
Interestingly, the most commonly cited change in both provinces was an increase in women’s
attendance and participation, followed by that of poor and ‘ordinary villagers’ (Table 6.1).
“From what I’ve observed, it seems every year there are more people attending you see, so it’s not just community
figures attending but also the number of ordinary village folk and women attending is increasing … Perhaps it’s because
they’ve begun to realize that the KDP program is actually for their needs, too…
KDP Financial Unit (UPK) Chair, Pasean, Pamekasan (1268)
In NTT, substantial increases in participation by minority ethnic and religious groups were also
recorded; that this was not the case in East Java can again be contributed to the province’s
relative ethnic/religious homogeneity.
Table 6.1: People Who Go to KDP Meetings Who do not Usually Go to Village Meetings
East Java NTT

Group (% respondents citing (% respondents citing
increased participation) increased participation)
Women 54.5 94.6
Poor People 30.3 93.5
Ordinary Villagers 40.9 87.1
Minority Ethnic Groups - 78.5
Remote Hamlets 28.8 77.4
Minority Religious Groups 1.5 53.8
Opposition to Village Government 24.2 26.9
Others 24.2 1.1
Source: Key Informant Survey (treatment sites)
Women and Participation
“Some men say that women are already given a chance, and they just don’t want it. But that is not true. If they were given
a chance, women would take it.”
Women’s Rights Activist, Maumere, Sikka
“Participation of women in KDP is only around 5% because it is a tradition in Madura that women are not allowed to mix
with men in a single forum, so it is not because their awareness is lacking, but rather the tradition.”
Villager, Padellagan, Pademawu, Ponorogo (1215)
“After the KDP meetings, where many of the women were invited, the women in Sana Daya experienced many problems.
They usually stayed quiet at meetings, now they’ve begun to propose things. Perhaps this can be interpreted as
indicating that after KDP women have become bolder. For example, there was a women’s Koranic recital group reading
Yasin [a book of the Koran]. Just one person regularly turned up. But after KDP arrived many of them began to come
and offer something…”
Female program beneficiary, Sana Daya, Pasean, Pamekasan (1255)
100
“Women have only just started to come together [in meetings] since KDP. PKK (Family Welfare Organization, run by women)
has been around for a long time, but only a few people would come, the people who came to the savings and loans routine
meeting yesterday afternoon, those ones. Women were unwilling because, first, back then it wasn’t clear what PKK’s activities
were … Aside from that, there was a cynical view of women who attended meetings in the village hall. ‘Well aren’t you like a
civil servant, attending meetings at the village hall, how much are you receiving in wages, who wants to be told to come to the
village hall anyway?!’ That’s the way the women would jeer. After being treated like that, women were usually embarrassed
and wegah [reluctant] to go. But since KDP, women’s participation has improved…”
Midwife, Kemiri, Jenangan, Ponorogo (1312)
Across almost all the research locations, women are represented poorly in the public realm in
village life. Women face a number of barriers to participation, both cultural and structural.155
Enshrined social structures and institutional practices discourage women’s involvement, as women
face restrictions due to their limited resources, time, skills, and numerous familial obligations.
Too much work in the house, in the field, and with children leaves women with little time to go
to village talks and meetings, or to get involved in other elements of village decision-making,
including conflict resolution:
“Usually if the husband is going to the meeting, the wife will stay at home. If a meeting is about social matters, many
women will go. But when they are in meetings, they cannot open their mouth.”
Midwife, Nele Wutong, Maumere, Sikka
“Some women speak at village meetings, but it is mostly men. It is difficult for women to speak. They feel that men should
be speaking, not them.”
Retired Female Teacher, Bloro, Nita, Sikka (S.654)
“…. Women who are active and invited to meetings … are mainly teachers”
Health Worker, Padellegan, Pademawu, Pamekasan (1213)
“Normally only the men get involved in conflict resolution. They will attend [the forum] and eat there. The women don’t
attend because there are prohibitions … that’s just the way it’s been since the time of our ancestors.”
Woman in focus group discussion, Bloro, Nita, Sikka (S. 668)
When asked why his wife did not attend village meetings, a government official in Bloro village,
Sikka, said that she had too much work in the house. Most other men gave similar answers in
interviews. Numerous women explained their lack of involvement by arguing that they did not
have enough time and could not afford to leave their work. While interviewing women for the
research, children often ran around the house and women looked after them while also preparing
food. At the same time, children rarely interrupted interviews with men, and women usually took
care of refreshments (Cutura 2003).
Table 6.2 shows the role of women in two important village institutions—the Village Representative
Council (BPD) and adat (traditional) leadership—in a sample of villages in Sikka, Flores. As can
be seen, women rarely play leadership roles.
155
It should be noted, that in many (usually private) realms of village life, women do play a key role. In particu-
lar, it is common for women to be in control of household economics, looking after money and outgoing ex-
penses. Thanks to Michael Dove for this point.
101
Table 6.2: The Role of Women in Village Life in Sikka

Village BPD Membership Adat Leaders
Bloro Men only Men only
Magapanda Men only Men only
Nele Wutong Men only Men only
Watugong 6 men, 3 women 3 men, 1 women
Source: Cutura (2003)
The same is true for the other three districts. For example, across the six research villages in
Kecamatan Sampung and Badegan, Ponorogo, there was only one female representative in the
BPD (Anggraini 2004).
However, the qualitative fieldwork backs up the survey evidence that things are different with
KDP. Numerous informants (both men and women) told us that the participation of women—both
in attendance, and in active engagement once there—was much higher at KDP meetings than for
other village meetings (Box 6.2).
Box 6.2: The Role of Women in KDP in Flores
“Many women in this village are involved in KDP. They were all invited to meetings so they went because they heard
that KDP would give money to the village.”
Female Adat (Traditional) Leader, Watugong, Maumere, Sikka
“Now more women come to KDP meetings. They come because of money, because they think they will see some
benefit form it. Otherwise, women would not want to come.”
Teacher and BPD Member, Watugong, Maumere, Sikka (S.26)
“There are more women at KDP meetings than at village meetings because of the socialization process. Women also
think that KDP will help them personally.”
KDP Official, Bloro, Nita, Sikka
“The women used not to speak, but now they often speak up … the women at the musbangdes (village KDP meeting)
are part of a weaving group, but they have marketing difficulties.”
Village Head, Bloro, Nita, Sikka (S.652)
“The general level of community participation in the KDP meetings was quite good … there were about 100 people …
indeed there were more men, but there would have been about 30 women, too.”
Village Facilitator (FD), Pagal, Cibal, Manggarai (M.1004)
There are a number of reasons why the participation of women is higher in KDP meetings. First,
the way in which the program is socialized encourages the participation of women and other
marginalized groups. The research found that in many cases simply inviting women to come has a
large impact. For most village meetings, women are not invited; in KDP, they are explicitly
encouraged to come.
Second, as Table 6.3 shows, the type of meeting is a large determinant of the extent of women’s
participation. KDP meetings are open; as such, women’s participation is higher. The distinction
between open and restricted meetings was one that most villagers (and particularly those who were
not invited to restricted gatherings) were aware of.
102
Table 6.3: Type of Meeting and Women’s Attendance in Sikka, Flores

Type of Meeting Who is Invited? Women’s Attendance
Public, open to all All villagers welcome, word of Higher
mouth or public announcement - Women leaders and ordinary
villagers
- 10-40% women
Private, restricted Invitation letters sent to specific Lower
people, as determined by village - Only select women leaders
leadership and topic of discussion - 0-20% women
Source: Cutura (2003)
Third, women come because they have incentives to do so. At least one proposal from every
village must come from a women’s group; as such, women actually have a reasonable chance of
getting their initiatives funded. As the above quotes from the informants demonstrate, this makes
them far more likely to attend KDP meetings than other village meetings, where funding is less
likely.
All this is not to say that KDP is a panacea for women. Their participation is still lower than men;
they are less likely to hold important positions within the KDP structure than men, although the
requirement that each village must have a female facilitator does help. Where they do attend
meetings they are, in general, less likely to be vocal than are men. Yet the signs are encouraging.
Trying to promote the equal participation of women is working against norms that have developed
over centuries. It can hardly be expected that one program can change this overnight. Even so,
KDP is having an impact, albeit a slow and gradual one.
Spillover Effects?
We would hope that participation in KDP would encourage villagers to participate in other elements
of village political life. In order to understand whether this is the case, we need to disaggregate the
democratizing impact of the program from that of the broader processes of social and political
change, post-1998. The survey data shows that across all research villages (KDP and non-KDP)
there is evidence of more groups attending meetings than in the past. However, this change appears
to be more prominent in villages that have received KDP than in those which have not.
“KDP forums at the village and kecamatan levels can increase community interaction. I think that gradually more and more
community groups became involved.”
Camat, Slahung, Ponorogo (1323)
In East Java, there is a trend of increasing participation of different groups in all village meetings
the longer time KDP has spent in the village. In villages that had not had KDP, 41.7% of informants
reported more groups coming, compared to 61.5% of informants in villages that had had KDP.156
Further, the impact increases over time: 75% of villagers in four-year KDP villages reported that
more groups came to village meetings.
156
If we count one-year villages as control sites, then the percentage of respondents reporting more groups coming to
village meetings are 52.0% (control locations) and 61.2% (treatment locations).
103
In Slahung sub-district, Ponorogo, East Java, the qualitative fieldwork found that KDP forums
had become a vehicle for increasing community awareness of development programs and
increasing their involvement in program planning and development (Anggraini and Rasyid 2004).
Indeed, there had been calls from community members for similar involvement in other
development projects to that which they have in KDP, with KDP becoming the yardstick for
community participation:157
“KDP is the only program which openly involves the community. Before KDP there were intensive work groups from
BP3MD [a previous government development project], but the villagers just participated without receiving any explanation
first. KDP is different because it relies on a socialization process first.”
“I am brave enough to bet that KDP is better because the community is satisfied. Even though they haven’t all been
involved, I am sure that they are satisfied. That is because we did the work ourselves in an effort to better the village.
So, we made it really solid. It would be different if other people had made it. They wouldn’t have paid any attention
to detail.”
Head of TPK (Village Project Management Team), Tengku Leda, Lambaleda, Manggarai (M.503)
“…What’s even better is that in Crabak, Gombang and Seneop villages, KDP has encouraged high community initiative.
After experience with KDP, the community requested that any development project should involve the community like the
KDP … When the fourth year was proposed recently, there was an FD in Desa Broto who sketched a bridge. He had
studied from the FK and TTD [Tenaga Teknis Desa – Village Technical Assistant]. The drawing was very, very good, like an
architect’s sketch.”
Development Coordinator (PjOK), Slahung, Ponorogo (1341)
In NTT, the differences are less marked with similar (high) levels of reported increases across
villages that have had KDP for different periods of time, although there still appear to be greater
improvements in KDP areas.
Figure 6.5: Change in Number of Groups Coming to Village Meetings by KDP Years: NTT & East Java
157
See interview with religious community figure, Wates, Slahung, Ponorogo (1326) and interview with development
coordinator (PjOK), Slahung, Ponorogo (1341). This backs up other evidence from KDP supervision missions. See
Guggenheim (2006) for a good example from Sulawesi.
104
KDP and Local Decision-Making

Survey informants were also asked whether decision-making was more democratic now that in the
past.158 The evidence shows again that village-level decision-making is becoming more democratic
almost everywhere. As Figure 6.6 shows, 57% and 77% of informants in KDP areas in East Java
and NTT (respectively) noted that decision-making had become more democratic. Less than 5% of
informants in East Java, and less than 3% in NTT, said decision-making had become less democratic.
These figures show impressive impacts from Indonesia’s decentralization and democratization.
Figure 6.6: Changes in Village-Level Decision-Making since KDP: NTT & East Java
Don’t know
% of Respondents in each Province
Can any of these changes be attributed to KDP? The data shows that in East Java, villages that have
received KDP report greater changes: 67% of four-year villages say decision-making has become
more democratic, compared to 46% in villages that had not received KDP. In NTT, the evidence is
more ambiguous, although, again, informants in four-year KDP-villages were the most likely to
say decision-making had become more democratic (94%) (see Figure 6.7).
Figure 6.7: Changes in Village Decision-Making since KDP by KDP Years: NTT & East Java
Informants in KDP areas were asked whether decision-making was more democratic since KDP had arrived. For
control areas, we calculated the mean number of years that KDP had been in operation in the research kecamatan in
each district. Informants were asked for changes over this period.
105
We should note here that KDP does not always lead to such behavioral change. There were
numerous cases where both the reputation and impact of KDP was negatively affected by contra-
democratic actions by people involved with the program. Where the program does not operate
in the manner it is meant to (and especially where transparency is low), behavioral spillovers
do not occur. We explore this in more depth in Chapter 7. However, where KDP does work
well, it positively reinforces the processes of democratization and decentralization, which have
resulted in real increased participation from ordinary villagers.
3. Changes in Norms and Perceptions
A third hypothesis is that KDP improves local conflict management through changing the ways in
which people understand problems and the best ways to respond to them. Violence, as we
noted earlier, is not only a symptom of conflict but a learned response to it. When faced with
a dispute with another party, individuals or groups can pursue a number of courses of action.
They can try to reach agreement directly with the other party or they can take the case to an
external third party mediator. Alternatively, they can take action without involving an outside
party. Responses can be violent or not. How people respond to conflictual situations is
determined by a number of different factors. The presence and/or quality of intermediaries
will help determine whether people use them.159 The extent and contours of the animosity with
the adversary, and the extent to which the barrier between ‘us’ and ‘them’ is inflexible, will
also determine the form of action.
However, perhaps most important are the existing norms that regulate what an appropriate response
is to a given problem. Rule systems govern people’s understandings and behavior in response to
events, actions, or choices. Human interactions and exchange are shaped by incentives, which in
turn are structured by rules, whether formal laws or informal norms (North 1990, p. 3). Understanding
the rules that exist within a given society is vital if we want to understand the ways in which social,
political, and economic relations are structured, and hence the ways in which people and communities
interact (Barron, Smith, and Woolcock 2004).
We hypothesized at the study’s outset that development programs could help shape norms with
regards to how people understand and respond to disagreements and disputes. KDP emphasizes a
collective and inclusive process of decision-making and problems solving. In doing so, do the
routines and practices that constitute participation in the program help people learn how to address
conflicts and problems in non-destructive ways?
3.1 Changes in Problem Solving and Conflict Resolution
In the previous part of this chapter, we showed how decision-making and problem solving is
changing in Indonesian villages. More groups take part, including marginalized groups;
decision-making is seen as more democratic. But is it any better? Are communities and/or the
159
See the discussion in Chapter 5 as to the different qualities interveners need.
106
state able to handle conflicts in better ways than the past? Are peaceful solutions more commonly
reached? In this section we ask three related questions: Do people try to solve problems differently
than in the past? Are people better or worse at solving problems/conflicts than in the past? And, if
there are differences, can we attribute them to KDP?
Survey informants were asked two perceptions questions related to changes in the quality and
nature of problem solving and conflict resolution. First, they were asked whether they thought the
quality of village level problem solving had changed since KDP’s arrival. Second, they were asked
whether problem solving/conflict resolution was more violent or peaceful, or whether it had stayed
the same, since KDP had arrived.
As Figures 6.8 and 6.9 show, informants were generally positive about changes. In East Java and
NTT, 41% and 69% (respectively) of informants said that village problem solving had improved
since KDP’s arrival. Higher rates of improvement were reported by kecamatan level informants:
67% and 89%, said it had improved in East Java and NTT, respectively. Respondents also said that
problems and conflicts were now solved in more peaceful ways. 73% of Florenese respondents
said that village-level problem solving was now more peaceful. In East Java, the rates are lower
(35%) with a majority saying the nature of problem solving had stayed the same. However, even
here, less than one percent said that problem solving was more violent. At the kecamatan level, in
both provinces a large majority of informants reported that problem solving was more peaceful
than before (76% in East Java; 79% in NTT).
Figure 6.8: Has Village Problem Solving Improved Since KDP?
Don’t know
% Respondents in each Province
107
Figure 6.9: Changes in Village Conflict Resolution Since KDP
Don’t know
% Respondents in each Province
That reported changes in kecamatan-level problem solving and conflict resolution are greater
than village-level changes is not surprising. The research found that institutions exist to deal
with most problems and conflicts within the village. Both formal structures (the government
hierarchy, which extends down to the election of neighborhood heads – RT), and informal ones
(community leaders, adat leaders, etc.) exist, and people generally know which of these institutions
and actors is most suitable for dealing with which type of problem. The same is true to some
extent at the district level, where the executive, the legislative, the security sector, and the numerous
arms of civil society all play a role in addressing particular types of conflict.
However, a vacuum exists at the kecamatan (sub-district) level. The Camat (Sub-district Head) is
almost entirely an administrative position, implementing decisions made at the district level. The
police, although present, have little autonomy in decision-making at this level, and are vastly under-
resourced (Baare 2004; Meliala 2005; Riefqui 2004). Adat structures are also weak at this level.
Even so, a number of problems arise at the kecamatan which need to be addressed. The research
found that most problems within village could be successfully handled. However, where conflicts
work across village boundaries, they are more likely to escalate, in large part because few extra-
village conflict resolution and mediation mechanisms exist. Especially in places like Flores, where
the distance between the village and district capital is immense, the district level is too far away for
villagers to take problems and seek outside intervention.160 Conversely, district actors are often
reluctant to deal with conflicts in remote villages. Interviews revealed that the thinking of many
district level government figures is that if the problem is big enough, it will come to us. Unfortunately,
by the time conflicts reach the district level, they have often built in intensity, making resolution
much more difficult than if they had been addressed early on.
160
This is also true in other provinces; see World Bank (2004) and Stephens (2003).
108
3.2 Is KDP Responsible for Changing Norms?
The survey data does not point strongly to KDP areas experiencing larger changes in problem
solving norms than non-KDP areas. Indeed, there is no difference in reported rates of positive
change between treatment and control sites (Figures 6.10 and 6.11). For changes in conflict
resolution, 56.2% of informants in KDP areas, and 56.3% in non-KDP areas, reported
improvements. Similarly, 55.3% of informants in KDP areas reported improvements in problems
solving; in non-KDP areas, the rate was 56.3%.161
Figure 6.10: How Are Conflicts Solved in Your Area Since KDP Arrived by Years of KDP
Figure 6.11: Changes in Village Problem Solving by Years of KDP

161
For both questions: treatment areas, n = 226; control areas, n = 96.
109
It would thus appear from the survey evidence that norms of problem solving and conflict resolution
have changed over time, and for the better, but that this is due to other factors than KDP. However,
the qualitative fieldwork does show some concrete links between the program and the building of
civic skills that allow for better, and more peaceful, resolution of problems and conflicts.
One such impact has been the program’s effect on modes of interaction between elites. As we
noted in Chapter 3, in some villages KDP was seen as a resource to be captured by competing
power blocks within the village. Sometimes this triggered interaction conflict, where competition
over KDP resources led to an escalation of pre-existing tensions. However, in other cases the
forums and rules of the program provided a space where elites could learn how to compete in non-
violent ways:
“… A positive impact achieved after the implementation of KDP in Pademawu sub-district has been the increased
community knowledge of the meaning of democracy. This can be seen from the Village Representative Councils
(BPD) in the villages. The atmosphere of the KDP meetings has been ‘absorbed’ in each forum held by the BPD so
that the democracy in process in these meetings is incredibly dynamic… One example is the Pademawu sub-district
BPD ‘Working Together’ forum which has been formed, where the members are from the members of the BPDs in the
villages. This forum functions to unify the perceptions on Perdas [district regulations] which are implemented at the
village level. They meet for hours and hours until nighttime without even getting a little tired! ... Apart from that, at
the community level ... since KDP, the community has gained more knowledge of how to reach consensus in decision
making, while still having their own desires.”
Sub-district Head, Pademawu, Pamekasan (1201)
“I have watched the village administration and the BPD learn from KDP over the past year. The BPD was elected last year
in this sub-district and in most villages it truly represented each of the hamlets in the village. The members in most villages
were not the Village Head’s cronies but instead were ordinary villagers for once, and in some villages they even held
parties to celebrate the election results … I watched as the members of the BPD and village administration battled it out in
KDP forums. If the Village Head tried to dominate the process, members of the BPD would stand up and point this out,
quoting back program rules … This worked both ways though, where, for example, in one village I saw the Village Head
and other members of the administration stand up and do the same thing when the BPD dominated the process. So they
are learning from each other and the program.”
Sub-district Facilitator (FK), Pasean, Pamekasan
“…The community’s democratic maturity following KDP is manifest in the forums found in BPD that follow the KDP
pattern… They are more mature in dealing with differences of opinion between BPD members as well as with the Village
Head in conflict resolution… Usually the problem relates to land boundaries, it often happens in Padellegan village. In
order to solve the problem, the BPD and the Village Head sit together to mediate between the two disputing land
owners. Then the two land owners are truly satisfied with all the resulting decisions.”
UPK Head, Pasean, Pamekasan
As such, there is some evidence that KDP is helping contribute to changes in relations between
different arms of the state at the village level. Another form of impact has been on the expectations
of villagers :
“Through KDP, villagers are learning to accept difference of opinion in a more mature and peaceful fashion, compared
with previously. These skills are being transferred to negotiations in other village meetings and negotiations between the
BPD and government. There was also a case where there was a land dispute. The BPD organized a meeting between the
warring parties and the final resolution was accepted without protest, violence, or further tension. Normally, people in the
community hate to lose face in a public forum, causing community tension, and won’t interact on the street.”
Sub-district Facilitator (FK), Pasean, Pamekasan
110
Indeed, the community has benefited indirectly from the critical discussions about public policy,
evaluations, monitoring, proposal formulation as well as the technical side of construction. The
concept of learning by doing is the key to success:
“Before it [KDP], I never knew that there were so many clever people in the village … perhaps before KDP villagers’
potential had never been utilized, so their capabilities had never been evident.”
Head of Saving and Loans Group, Jenangan, Jenangan, Ponorogo (1304)
“As for positive effects, I think it can increase the community’s outlook, from one of ignorance to one of knowing. The
community can learn how to lobby other delegations so their proposals succeed.”
Teacher/TPK Secretary, Wates, Slahung, Pamekasan (1337)
4. The Impact on Conflict Outcomes
We have seen the impact of KDP on a number of variables that, we have argued, help determine the
local environment which makes conflict more or less likely, that help to dictate its nature when it
does arise, and which help shape local capacity to manage it in more or less violent ways. However,
demonstrating a causal link between KDP and changes in local “conflict environments” goes only
part of the way towards showing a link between KDP and local conflict management capacity. We
must also test the second part of the hypothesis: that changes in these environmental variables help
shape the outcomes of conflicts.
Establishing this link empirically is difficult. We sought to use the newspaper dataset on reported
incidents of conflict to look for correlations between (reductions in) violence levels and improvements
in group relations, or behavioral or normative changes. We found no such clear correlation: sub-
districts reporting larger improvements in the variables explored in this chapter did not see larger
reductions in levels of violent conflict than those where indirect impacts were not as great. Does
this disprove the hypotheses outlined earlier that improvements in group relations, increases in
participation and decision-making, and changing attitudes to problem solving and conflict resolution
can limit conflict from occurring and escalating?
In the remainder of this chapter we turn to case analysis and lessons from theory to understand
better the ways in which local structure and norms may affect conflict incidence and outcome. We
argue that one of the reasons for the lack of correlation between changes in the variables measured
in the survey and changes in levels of conflict is that conflict is explained, in part, by extra-local
factors. This limits the effectiveness of local approaches to limiting violent conflict when
implemented alone, without complementary approaches to get at larger drivers of conflict and
power interests. However, we argue that local factors do often matter in determining the likelihood
of whether small-scale ‘everyday’ conflicts escalate into larger ones.
4.1 Extra-Local Factors and the Incidence of Violent Conflict
It is unsurprising that we do not, at an aggregate level, find a link between changes in these
variables and levels of conflict over time. One important reason is that local conflict cannot be
explained entirely by local factors. Changes at the local level may increase conflict management
capacity, but local capacity can be overwhelmed by large-scale problems. It is clear that many of
Indonesia’s larger conflicts—in Poso, in Maluku, in Aceh before the peace agreement—have as
111
much to do with elite interests (at various levels) as they do with local cohesion and forms of
decision-making. This was also evident in some of the cases already discussed in the paper and in
others we collected. Two other cases, one in East Java and one in NTT, illustrate.
The Mining Company in Ponorogo

PD Sari Gunung, a company owned by the Ponorogo district regional government, carries out
limestone mining on government land in Sampung village. Institutionally, PD Sari Gunung is
responsible to the Ponorogo district government. However, on a day-to-day basis it has the authority
to make technical decisions.
“This is a regional government enterprise, so profits are not prioritized. In fact it has more of a social function. Every
year we must make a contribution to local revenue … I don’t know the history of it, but this is government-owned land
…. Because it belongs to district government, we do not have obligations to the village or the sub-district. We’ve
already provided funds to the district government. As to how much of these funds are allocated here, that’s up to them.
If it’s just about helping out, when, for example, the village or the sub-district has an event and then asks for a contribution
from here [PD Sari Gunung], well we give it. … But if special funds are being sought because we carry out mining here,
there aren’t any, because that’s the local government’s responsibility.”
Director, PD Sari Gunung, Sampung, Sampung (19)
Because the state enterprise makes payments to the district government, they argue that they
have no responsibility to the local peoples in the areas where they are situated, that this falls
instead within the realm of authority of the district government itself. In the 1980s, grosok
(limestone mining refuse) began to destroy several road sections in the hamlets to the west of the
Sampung main road. After a series of complaints throughout the 1980s, drains were finally built
to carry the grosok away. However, the residents did not know which entity had built the drains
because they were not consulted in the design. The drains that were built only shifted the refuse
problem to a different hamlet in the village further down the hill.
“At my place it was just the same. The waste from the market would come into the house. When it rained the grosok
mixed with gravel would be affected by rainwater. Well, that mixed with market waste and came into my house. Now it’s
like a ditch….”
Villager, Sampung, Sampung. Ponorogo (37)
Over time, the number of actors in the conflict grew, as did its complexity. It influenced the
allocation of development funds, and increased tensions between the different hamlets in the
village.
“There used to be aid, but it was given to other hamlets. The reason they gave was that the community here could not
take care of the environment, so it did not deserve to obtain aid… At this thanksgiving event many commented about the
aid. The community in the region over there [the other hamlets] said all sorts of things about the community here,
especially my family, because the biggest amount of grosok had piled up in front of my house but I just left it alone.
Initially I just stayed quiet; I let them say whatever they wanted. After many people had commented I said every year
when it rains the grosok is always carried down by the water. Not only mud but grosok is also carried by the water. My
house happens to be the lowest. So the grosok and mud pile up in front of my house. There used indeed to be voluntary
labor. But after a while no-one wanted to do it anymore, they were exhausted. The grosok increased, and one person
could no longer clean it up, it needed many. The village or sub-district authorities alone wouldn’t be capable of overcoming
this matter. If you don’t believe me, tell the local government people to come here. If they can manage to clean it up,
I’ll drink their urine!”
Villager, Sampung, Sampung. Ponorogo (37)
112
In December 2002 the rains were so heavy that the drains could not handle the volume of refuse;
they backed up and the refuse blocked off the main road and peoples’ front yards. The villagers had
reported these problems several times in the past to the village administration, but it had inadequate
resources to fix the problem. The complaints were also being ignored by both the district
administration and the state enterprise, each claiming that the other had responsibility for dealing
with the problem. Thus in January 2003, youths from the area most seriously affected by the
problem blocked off the drains as both a means of protest and to reduce the impacts of the refuse on
their living environment. This shifted the problem back up to the main road in the center of the
village. An agreement was made between hamlets not to clean up the main road as a means of
drawing the attention of the district and other authorities. The problem remains unresolved, and, at
the time the research was conducted, the BPD (village parliament) was trying to negotiate with the
company on behalf of the villagers affected by the problem.
Who Owns the Mbondei Land? Land, Tradition, and Mediation in western Flores
Like so many of the villages throughout Manggarai district in Flores, Tanah Rata village is in the
midst of a controversial land dispute. At the centre of the dispute is the Mbondei land, a 50-hectare
expanse of what was originally clan land administered by the Motu Poso ethnic group. The land’s
status is now unclear. The dispute involves the Kisol Seminary, Himastan (a community group of
farmers), two smaller groups of villagers, as well as the original clan owners. There have been
numerous attempts to resolve the dispute, yet until now these attempts have been fruitless and the
possibility of bloodshed remains.
In 1967 the Motu Poso clan’s land owners transferred the rights to the Mbondei land to the
Kisol Seminary so that they could use the land to herd their cattle. Due to the fact that most
customary (adat) practices involving the transfer of land in Flores are based on oral agreement,
there was no written documentation that could be used to clarify whether the Kisol Seminary
was given ownership or simply usage rights to the land. Conflicting versions of events from
the surviving witnesses were of little help in determining the status of the land. In the resulting
confusion, and contestation, provocative actions could lead to conflict.
By mid-2002 tensions had begun to surface. It was a fresh June morning in Tanah Rata when
around 140 members of Himastan arrived at Mbondei and began dividing the 50 hectare piece of
land among themselves for the cultivation of crops. Not only was it time to draw attention to the
injustice of someone else occupying and claiming the land of their ancestors; they could also see
the benefits of the extra money the crop land would bring them. By the end of the year, after
encouragement from the farmer’s rights group (Himastan), two other groups of villages from
Kelurahan Tanah Rata decided to also lay claim to parts of the land. The once quiet grazing pastures
of Mbondei were suddenly a hive of activity, with people cutting down trees and cultivating crops
on a daily basis.
Not only was the Kisol Seminary shocked and angered by Himastan’s actions; the original clan
land-owners were also outraged. In response to Himastan’s actions, the government (village,
sub-district and finally district) stepped in, as did the adat elders and church representatives. In
general these institutions agreed: according to the government and the adat institution, Himastan
and the two village groups had not only violated adat law, but had also violated government
113
regulations by cultivating the Mbondei land. Even so, there was confusion over who had jurisdiction
to address the problem.
The adat functionaries requested that the cultivators sign a statement declaring that they would
cease activities on the disputed land. However, the Head of Himastan refused: “We don’t need to
write a statement and we will continue to work. We will ignore the warning from the Village Head
and the adat functionaries.” The sub-district government then decided to get involved, drafting
and distributing a warning letter prohibiting the cultivators from working the land. The government
held several meetings with the disputing parties, with no success. Finally, the cultivators signed a
statement declaring that they would cease their activities on the land. Yet just as the case looked
like coming to an end, the cultivators withdrew their statement and went back to working the land.
Tensions rose once again.
The sub-district government had reached a dead end, as had the numerous other mediators who
had made efforts to resolve the case. There was little more they could do except hand the case
over to the district government. Still, the resolution strategy is unclear. The conflict is currently
in the hands of the district government and, with tensions mounting, one wrong move from any
of the parties could result in bloodshed.
Power, Complexity and the Limits of Local Conflict Management Capacity

In the case related to the PD Sari Gunung government mining company, the sheer complexity of
the number of actors involved, and the fact that a powerful company with links to state interests
was one of them, meant that local attempts to solve the problem were unsuccessful. In this
particular case, it was the complexity of the problem, the lack of resources and lack of clear
responsibility and cohesion between different levels of government (and the company), which
meant that the problem could not be solved. In such a case, it is unlikely that ‘local capacity’
would be able to solve the issue, although the local BPD (village parliament) was making a
valiant attempt. In the land case in Manggarai, the host of different parties involved in conflict
resolution also led to a level of complexity that meant that local efforts for resolution were
ineffective.
For many conflict cases such as this, broader power interests are at play. It is not surprising
that, in such cases, local conflict management capacity is in some ways irrelevant. Indeed, the
village in Ponorogo was one of the strongest in our sample in terms of local conflict management
capacity. This should serve as an important warning to those who see community-driven
development (CDD) projects as the answer to problems of local conflict. CDD projects need
to be accompanied by complementary strategies that focus on power relations and incentives at
higher levels.
4.2 Local Factors and the Likelihood of Conflict Incidence and Escalation
Extra-local factors do matter. The extent to which violent conflict occurs at the local level is
dependent on both exogenous and endogenous factors. Exogenous variables may include national
economic policy, population flows, and centralized security institutions. Such forces can overwhelm
local communities. Large shocks can lead to both high levels of social tension and violent conflict.
114
However, external macro forces only partially explain why some conflicts take violent form.
Social tensions exist in many places, yet only lead to violent conflict in some. External forces of
the same type and scale impact in different ways in different places.
Our case studies show that in environments like East Java and NTT, which are characterized by
frequent low intensity conflicts, local factors related to both social structure and dominant norms
matter in terms of determining the likelihood of relatively simply local disputes arising in the first
place and, if they do, of them escalating into more complex and protracted conflicts.162
One of the reasons cases like that of the land conflict in Manggarai get so complex, with so many
actors involved, is that they are allowed to escalate. Comparative analysis of our conflict cases
shows how major conflicts often have their roots in smaller-scale (‘everyday’) forms of conflict.
Using a modified version of the process tracing methodology (George and Bennett 2005; Varshney
2002) we traced the chronologies of 68 local conflict cases with the aim of trying to establish why
conflicts ended up with the outcomes they did (violent or not, escalated or not), and the ‘pathways’
they took to get there. We broke each conflict up to a series of chronological parts; at any particular
given moment, an existing conflict might either (a) stay at the same level of intensity, (b) escalate,
or (c) get resolved, either temporarily or permanently, through negotiation or outside mediation.
Breaking down conflicts into smaller parts allowed us to analyze what factors led to outcomes (a),
(b) or (c) at any given moment.163
We found that three realms of variables helped to shape the pathways along which conflicts traveled
and to determine outcomes.164
The first realm centers on the actions of mediators, leaders or officials in conflict resolution, and
the extent to which they possess the necessary legitimacy, willingness and capacity to both make
and enforce decisions. Comparative analysis of the cases found, unsurprisingly, an inverse
relationship between the ‘efficacy of intermediaries’ at local levels and the likelihood of conflicts
escalating. Direct conflict management interventions, such as strengthening the legitimacy and
capacity of local mediators, and providing incentives such that they are willing to address conflict
issues, can limit the likelihood of given incidents of conflict turning violent.
However, the comparative case analysis also highlighted the extent to which the structure of
inter- and intra-group relations, and the relationship between society and the state, affected conflict
patterns. Two realms of variables relating to what we called earlier the local ‘conflict environment’
mattered in influencing the likelihood of conflicts escalating and/or turning violent.
162
One of the reasons we concentrated on local factors in the study was that most studies, at least in Indonesia, have
tended to prioritize extra-local (often national) explanations of conflict in Indonesia (see Bertrand 2004 and much of
the institutional transitions literature).
163
For a justification of breaking larger events down into smaller ‘episodes’ for analysis, see McAdam, Tarrow, and
Tilly (2001).
164
We outline this model in much greater depth in Barron, Smith and Woolcock (2004). The two provincial papers—
Diprose (forthcoming) on East Java and Satu and Barron (forthcoming) on NTT—apply these frameworks to much of
the case material.
115
The first realm centered on various elements of the ‘rules of the game’: the laws and norms that
shape the immediate context in which disputants, their representatives, or external mediators engage
one another. Conflicts are less likely to escalate and turn violent when people have a collective
sense of what behavior is acceptable—and what is not—and/or how a particular dispute should be
decided (and by whom). When the rules system that is applicable to govern a particular action or
behavior is hazy or disputed, and/or where multiple sets of incoherent norms exist, people may be
more likely to engage in provocative, and ultimately conflictual, behavior (Ross 1993; Rothchild
1997). In our sample, this was particularly true of conflicts over land, where different kinship
groups had different rules and understandings of how land should be allocated. The largest conflict
in our newspaper dataset—resulting in 15 deaths—was of such a nature.
A second realm focuses on the norms and politics pertaining to inter-group (“us-them”) relations—
or what we call the ‘dynamics of difference’— and the ways and means by which differences are
constructed, and are able to be mobilized, re-imagined, and exploited for strategic advantage. Where
group boundaries between groups are marked, with little opportunity for cross-group interaction,
conflicts were more likely to take place. When they did occur, they also escalated more easily,
largely because the appeals to group interests were used as a basis for mobilization.165
These contextual variables are important at multiple levels. Anderson (1991), for example, has
shown in his work on nationalism how identity formation and transformation are conditional on
demographic, political, economic, cultural, policy, and legislative environments at multiple units
of analysis. Likewise, the rules that structure human exchange exist at multiple levels. Laws can
be passed by legislatures at the national, provincial and district levels; Sub-District and Village
Heads also shape local rule systems. Civil society exists at multiple levels and has an impact at
each. The relative youth of Indonesia as a post-colonial nation, and the correspondingly short time
for national legislation to have developed into everyday and effective norms in remote and still
traditional areas, has meant that (locally derived) traditional laws and customs continue to frame
everyday life.
A key finding was thus that local structure does matter. Our research strategy sought to isolate
the importance of local factors by comparing cases over similar issues but with different outcomes
within sub-districts and districts, so as to hold constant for extra-local factors.166
On a case-by-case basis, local factors played a large role in determining both the pathways
conflicts would take and their eventual outcomes. Even most of the conflicts that overwhelmed
communities (like the land case in Manggarai) had their roots in local issues; all the evidence
points to the fact that they would not have escalated in the way they did, and hence become
unmanageable at the local level, if they had been dealt with early on. Further, if local rule
165
This was particularly true in Manggarai district. We discuss this in more detail in Barron, Smith, and
Woolcock (2005). On the ways in which culture can provide a basis for group mobilization by leaders seeking to
consolidate or expand their power, see also Wolf (1964; 1969; 1999) and Brass (1991).
166
See Annex B for a fuller discussion of how case studies were selected and local factors disaggregated from
extra-local ones.
116
systems had been less incompatible, or if group boundaries less marked, it is unlikely the
conflict would have escalated in the way it did. Improving the number and quality of inter-group
interactions in most cases makes group identities less divisive, and it makes it harder for elites to
mobilize community members based on narrowly particularistic senses of ‘us’ and ‘them’.167
Increasing access to decision-making is also likely to make local rule systems more cohesive.
Decision-making that encourages access from all sections of the population is less likely to result in
outcomes that provoke conflict, because (a) when people are involved in processes that they deem
equitable and legitimate, they are less to dispute their outcomes, and (b) institutionalizing open
discussion can help parties find common ground, much of which only becomes clear through the
very process of discussion itself.168
4.3 Strategies for Improving Conflict Management Capacity: the Demand Side of Good
Governance
If both local and extra-local factors matter, a key policy question becomes which should be
prioritized. Should strategies to limit violent conflict in Indonesia and elsewhere focus on trying
to affect local conflict environments, or should they concentrate on national factors (e.g., correcting
inequitable and unclear land distribution, weak justice and security sector, etc.)? From our research
it appears both are necessary. There are real limits to what locally focused approaches can achieve.
In particular, in cases involving the military or large private companies (often in cohorts with
local or national government), local capacity becomes largely irrelevant. This is true for many of
the massive land and natural resource conflicts that occur throughout the archipelago.
Yet, at the same time, there does appear to be an important role for local approaches. In the
medium- to long-run, stimulating demand for reform from the bottom-up may be the best way of
changing the institutional structures that allow conflict to flourish. Lessons from around the
world show that exclusively top-down and/or technocratic approaches to ‘improving justice
systems’ or ‘building government capacity’ often fail, primarily because they do not affect the
incentives at play or take adequate account of the identities, values, and meaning systems in
which such incentives are embedded.169 More recent and alternative approaches, such as KDP,
seek to correct these shortcomings by incrementally building the demand for and capacity of
local governments, doing so by simultaneously increasing the civic skills of everyday citizens,
and by establishing new procedures and precedents for positive engagement between citizens
and their state.
167
It is important to note that such interactions need to be institutionalized. See Ashari and Hergesell (2006) on the case
of Poso, Central Sulawesi.
168
In the short-run, of course, opening up local decision-making can increase tensions and conflict and elites lose their
monopoly. We saw in Chapter 3 how resistance to democratization from elites can make conflict more likely. The
literature has also shown the extent to which transitions to democracy can see surges in levels of conflict (Gurr 2001;
Haggard and Kaufman 1995). This was all too evident in Indonesia, which saw a massive upsurge in violence in the
immediate years following the fall of Soeharto (Varhsney et. al. 2004).
169
On the limits of past approaches to ‘judicial reform’ in developing countries, and the (more hopeful) prospects of
current initiatives, see Sage and Woolcock (2005).
117
“Generally KDP has had a positive impact … it has trained and educated the villagers in musyawarah and mufakat
(discussion to reach an agreement) and the right procedures for reaching an agreement. KDP has taught the villagers
how to discuss issues and come to a democratic agreement based on deliberation and consensus. It has also taught
people to value other people’s opinions.”
Former Section Head of PMD (PjOK), Ruteng, Manggarai (M.1029)
Such approaches invariably take time, and in building up the ‘demand side’ it is also important to
work directly with governments to help strengthen the corresponding ‘supply side’. Yet in contexts
like contemporary Indonesia, a key strategy for facilitating change at higher levels is itself stimulating
demand from below. Building, consolidating and refining such approaches are vital not just for
‘making services work’ in poor communities, but, more fundamentally, for ‘making democracy
work’ in volatile settings in the midst of multiple economic, social, and political transitions, the
unsettling vicissitudes of which are likely to be a feature of life in Indonesia for decades to come.
5. Summary
Local conflict in Indonesia is largely a product of weak institutions. The last chapter assessed
whether KDP had a direct impact on strengthening conflict management institutions, through
introducing people (facilitators) and spaces (decision-making forums). The research found that, in
general, direct impacts are minimal, but that in places where effective local conflict management
institutions do not exist, KDP forums and facilitators at times play a role.
However, whereas the direct impacts of KDP on conflict management are small, the program
has notable (and positive) indirect impacts on the local institutional environment in the areas
in which it operates. Through changing inter-group relations, and the relationship between
citizens and the state, the project affects local conflict environments – the structures and norms
that make conflict more or less likely to arise and/or to escalate. KDP is helping improve
inter-group and state-society relations.
First, the research found that across a range of different identity cleavages, KDP had helped
contribute to improvements in inter-group relations. Ethnic, religious and class relations in NTT
have improved since KDP was introduced, and these changes are greater in treatment than control
areas. There is also some evidence that relationships between those affiliated with competing
martial arts (silat) groups have improved in East Java. Further, improvements in group relations
grow larger over time. Villages that have had KDP for four years show, in general, greater
improvements than those that have had the program for shorter periods. KDP provides a space
for different groups to come together to collectively discuss their needs and priorities, something
increasingly rare (especially at the inter-village level). KDP also facilitates group interaction by
improving transportation networks.
Second, KDP also appears to be effectively reengineering the relationship between citizens and
the state at a local level. KDP brings a set of rules and norms concerning, amongst other things,
who should participate in decision-making, the criteria that should be used for resource allocation,
and what checks and balances should be in place to control local power. In this sense, projects
like KDP are inherently political: they value one system of social organization over another, and
they introduce resources (financial and human) to change incentives to make it more likely that
118
this form of social order will materialize. The evidence shows that KDP is successfully helping
to democratize village life. Marginalized groups (and, in particular, women) are far more likely
to take part in KDP meetings than in other village government meetings. Yet increased participation
in KDP also appears to be spilling over into other domains of village life. 42% of villagers in
areas who had not had KDP reported that more marginalized groups were coming to village
meetings than in the past, compared to 62% in KDP areas. 75% of villagers in four-year KDP
villages reported that more groups came to village meetings than before. Decision-making in
village meetings has also become more democratic, and this effect is greater in KDP areas than in
the control sites.
Third, villagers report that problem solving has improved and conflict resolution has become more
democratic since KDP arrived. The key informant survey shows similar reported rates of
improvement for KDP and non-KDP areas. However, the qualitative fieldwork shows clear links
between the program and such normative changes, with KDP (when it functions well) creating a
positive precedent, in the process helping to stimulate demand for changes in the ways in which
local decision-making and conflict resolution operate.
We find no direct correlation between the levels of indirect change and reduced levels of conflict.
This, we argue, is partly related to the importance of extra-local factors in determining conflict
levels; it is partly related to the fact that it will take time for the democratization impacts of KDP to
lead to increased conflict management capacity. However, analysis of the cases shows that local
factors do matter significantly (in particular, as to whether or not the conflict is likely to escalate to
a level at which communities can no longer deal with it). The cases also suggest that the
democratization of Indonesian village life, and the strengthening of inter-group relations, will make
communities more robust to conflict in the medium-long term and that ‘demand-side’ approaches
are a necessary compliment to strategies that address the ‘supply’ of good governance.
The research shows that KDP alone does not create these profound changes. Rather, when it
works well, and in environments favorable to change, KDP interacts with existing processes of
social and political transformation, acting as a catalyst that legitimizes processes already underway.
Indonesia, at present, is in the middle of a complex and contested transition: from authoritarian
state to democracy; from being one of the most centralized states in the world, to one of the most
devolved. The research shows broadly positive changes from Indonesia’s transition at the local
level. Village-level decision-making is becoming more democratic. Villagers are participating
to a greater extent in local political life, and, increasingly, they are holding their leaders
accountable. Norms are changing. KDP, when it works well, and in conducive environments,
acts as a catalyst to these processes of social and political reform. KDP is a vital resource that
reformers at the local level can use to legitimize their position. Indeed, as the next chapter, will
explore in greater depth, the indirect impacts of KDP on conflict management are greater in
higher capacity areas. Programs like KDP do not operate in a vacuum. The success (or failure)
of programs such as KDP, which aim to promote local level democratization in a post-authoritarian
environment, should be measured not on the sole observable impacts of the program alone, but
on whether or not it effectively support existing processes of change.
119
120
Chapter Seven
Explaining Variation: The Influence of Context on Program Spillovers
Thus far, we have examined the different impacts—positive and negative, direct and indirect—
that KDP has on local conflict and conflict management. We have found that KDP triggers
conflicts and interacts with existing tensions, but that these do not become violent. KDP has
little direct impact on either conflict levels or conflict management; KDP forums and facilitators
are not often used for conflicts not related to the program. However, it does indirectly affect
‘conflict environments’ and, in so doing, helps improve medium- to long-term conflict management
capacity.
These are aggregate findings representing general trends across the different research areas.
The general results, however, mask tremendous variation between different areas. For example,
while KDP triggers low-level conflict in many areas in which it operates, in some villages it
does not. Similarly, KDP facilitators and forums are rarely used for resolving conflicts that are
not related to the program, but in some villages they are. Across the forty-one villages, KDP
was found to lead to positive normative and behavioral change, and improved relations between
different identity groups, yet in some cases these effects were negligible. Why? This chapter
seeks to provide a basis for explaining variation in KDP’s impact on local conflict. The focus
here is on understanding interactions between types of disputes, the efficacy of KDP
implementation, and levels of local and district capacity.
1. Explaining Variation
In this chapter we outline the extent of variation between different areas where KDP is
implemented for each of the dimensions of impact discussed in previous chapters. As yet we
have not examined how the various factors which lead to differential impacts of KDP on conflict
management capacity interact with different levels of pre-existing conflict management capacity.
These are important questions for projects like KDP, which aim to do ‘small development’ on a
large scale. How can projects be designed that allow for local discretion in implementation, yet
which take account of differences in local capacity? In what kinds of areas can projects like KDP
have an impact on local conflict and its management? How, and to what extent, does the way in
which KDP operates affect its impacts? We examine reasons for variation by looking primarily
at three variables: capacity at the district level, capacity at the village level, and program
functionality.
Variation in KDP impact is not surprising for a number of reasons, some relating to the characteristics
of Indonesian villages, others to the nature of the program itself. First, the different villages in
which we conducted research vary greatly. The difference between a village in a prosperous
part of Ponorogo, East Java, and one in a remote corner of Manggarai, Flores, is about as large
as could be imagined in Indonesia. Socio-economic and institutional structures are vastly different
between villages in, say, Madura and those in Sikka, Flores. Different cultures exist locally,
differentiating not only villages in different provinces or districts, but also those within the same
121
district. In such diverse climates, it is not surprising that the impacts of KDP differ from area to
area. Indeed, our research design was explicitly set up to maximize contextual variation.
Second, local capacity varies enormously. ‘Capacity’ is an oft-used but ambiguous, abstract
and contested term, being as it is normatively loaded. There are a number of dimensions to
capacity and it can be measured at different levels. Debates rage as to its most important
determinants, with different people emphasizing, among other things, education/skills, access to
financial resources, institutional quality, and social capital or cohesion.170 There is debate over
the extent to which capacity is an individual or collective attribute, with differential emphases on
the agency and opportunity structures that help shape the freedoms people have.171 Sometimes
it is considered a development input (e.g., local capacity is necessary to reach development
outcomes, such as economic growth); for others, it is a development outcome in itself (note the
explosion of ‘capacity building’ programs from the World Bank and other donors).172 It is beyond
the scope of this chapter to provide a detailed overview or critique of ‘capacity’ as used in the
development studies literature; for our present purposes, it is sufficient to note that different
forms of local capacity clearly have the potential to impact upon how KDP functions in different
areas, and the resulting impacts on local conflict and conflict management. In the analysis below,
we consider what we regard as the two defining elements of local capacity—problem solving
and collective action—at two levels (the district and the village). Given the extent to which KDP
relies on local actors, it is not surprising that it functions in different ways in areas with different
preexisting levels of capacity (see again Box 2.3).
Third, the structure of KDP itself arguably contributes to variation in performance between areas.
KDP deliberately devolves decision-making to the local level. While strict program rules exist—
for example, relating to the meetings that need to be held, decision-making processes, how funds
should be dispersed, and so on—tremendous discretion exists locally for both program staff and
participants. Indeed, KDP is a conscious move towards the notion of development projects as “ways
to trigger and support processes in which villagers exercise discretion in solving self-identified
development problems” (Guggenheim 2006). It is thus not surprising that program performance
and impacts vary considerably at the local level.173
We examine reasons for variation by looking primarily at three variables: capacity at the district
level, capacity at the village level, and program functionality. Each warrants a little explanation.
We define capacity as the ability to engage in collective action activities and to solve group
problems. We derived measures of capacity at the district and village levels.174 We did so for
170
On approaches to understanding ‘capacity building’ and their implications for development policy, see McNeil and
Woolcock (2004). See also Eade (1997).
171
See Alsop and Heinsohn (2005).
172
Fukuyama (2004) makes the case for conceptualizing development as capacity building. He argues that “the prob-
lem of capacity destruction cannot be fixed unless donors make a clear choice that capacity-building is their primary
objective, rather than the services that the capacity is meant to provide” (2004, p. 55; emphasis in original).
173
On the importance of discretionary and face-to-face decision-making for explaining development failure (and,
concomitantly, as a key ingredient in making services work), see Pritchett and Woolcock (2004).
174
In the Indonesian contexts, the district is extremely important because, with decentralization, tremendous resource
allocation and policy-making discretion exists at this level. The quality of local government helps determine both the
ways in which development programs are implemented and the likelihood of violent conflict. The village level is also
important. Decentralization, as we noted earlier, has helped open-up village level politics.
122
the former through conducting interviews at the provincial level with government, international
and national NGOs, regional development experts, universities, and KDP staff. At the village
level, the researchers ranked the villages where they conducted research on problem solving and
collective action. Vignettes were used to ensure comparability across districts and provinces.175
Villages were then separated into those with ‘high capacity’, those with ‘medium capacity’, and
those with ‘low capacity’. Given that each of these is in a district with either ‘high’ or ‘low’
capacity, a given village will be of one of six different capacity ‘types’—e.g., a low capacity
village in a high capacity district, a medium capacity village in a low capacity district, a high
capacity village in a high capacity district, and so on. We use this ranking of villages to tease out
differences in the impacts of KDP between areas with different levels and kinds of capacity; it
also allowed us to determine for a given dimension of impact whether it is the district or village
context that matters more.
Second, we sought to see how program functionality affects program impact. As we outlined in
Chapter 3, KDP, despite vigilant oversight, does not always function as intended. Sometimes
elites control the program; sometimes a facilitator or program participant runs off with the money.
Clearly, both of these situations result in less than optimal program performance. But how does
this affect the program’s impact on conflict? For each dimension of impact, we examine the
extent to which and the ways in which program functionality matters.176
2. KDP-Triggered Conflict and Local Context
In Chapter 3 we argued that KDP forums can trigger three common forms of conflict: in-built,
malfunction, and interaction conflict. These forms of conflict rarely, however, become violent.
Among our survey informants, 45.5% in East Java and 19.4% in NTT reported that village level
forums triggered conflicts, whereas at the kecamatan level, reported rates were even higher:
85.0% and 57.3%, respectively. But almost no respondents said that these conflicts had turned
violent.
2.1 The Effect of Local Capacity on KDP-Triggered Conflict
Levels of KDP-Triggered Conflict: The Impact of District Capacity

Survey respondents in the high conflict management capacity districts (Ponorogo and Sikka)
were far more likely to report that KDP triggered conflict than those in the low capacity districts
(Pamekasan and Manggarai). This was particularly true for KDP meetings held at the village
level, which were reported by more respondents as triggering conflict than were meetings at the
175
See Annex C for the methodology, and the ranking list. Anchoring vignettes (see King et. al. 2004) are short,
familiar anecdotes that are presented to villagers in different settings; the vignettes are drawn from everyday life and
include examples of (say) strong, moderate, and weak program performance; the responses to these vignettes can be
used to ensure greater compatibility and comparability between questions asked in different contexts.
176
Unlike the capacity measures, we do not have a quantitative measure of program performance. At face value, it is
hard to conceive of reliable indicators that would be truly independent of local context effects. For example, one could
use data on the presence or absence of corruption, but the measured presence of corruption could actually be a sign of
the program operating well if it was caught through the program, and then effectively addressed. For our present
purposes, we can only make binary distinctions between strong and weak program performance on the basis of evi-
dence drawn from our case studies documenting instances of program malfunction; for future work, we would
recommend incorporating common ‘anchoring vignettes’ as a way of qualitatively assessing (and then quantita-
tively validating) comparative measures of program performance.
123
sub-district level; the difference was most marked in East Java (Figure 7.1). In East Java, a
clear difference was apparent between Pamekasan (low capacity) and Ponorogo (high capacity)
districts. In NTT, there was less variation between high and low capacity districts, with the
higher reported rate in Sikka a function of one sub-district, Paga.177 This seems to correlate
with particular implementation problems in Paga:
“Delinquent loans have become a big problem, proving to hamper the performance of KDP and making it difficult for the
program to reach its target, namely, improving the livelihood of the poor. According to me, it is a result of poor coordination
at the kabupaten (district) level. From a different perspective, the KDP process in the first year was weakly executed,
perhaps because it used a new pattern … a pattern for empowering the grassroots community. Everyone was shocked,
most of all the beneficiaries. They were still infected with the ways of the previous period whereby infringements were
considered normal … Other weaknesses at the village level included poor management. Just imagine, our Village
Heads are trying to manage just Rp. 10 million in village aid and the administration half kills them. Then all of a sudden
they are faced with tens or even hundreds of millions of rupiah. They are just so confused, even though they are being
helped by a facilitator…”
Sub-district Head, Paga, Sikka (S.1134)
When the responses across both the provinces and the two levels of meetings types are combined,
63% of informants in the two high capacity districts reported that a KDP forum had triggered
conflict, compared to 44% in the low capacity districts.
Figure 7.1: Percentage of Respondents Reporting KDP-Triggered Conflict by District Capacity
% Reporting KDP-Triggered Conflict
Why would conflict be more likely in districts with high capacity? Would we not think that KDP
would be more likely to trigger disputes in areas where they do not have the capacity to deal with
problems?
Two factors explain the higher reported rates in high capacity districts. First, in some of the
lower capacity areas (particularly Pamekasan in East Java), elites consciously prevented conflict
177
In Paga, 47% of informants reported that KDP village forums triggered conflict. Across the five other sub-districts
in Flores, rates were between 10% (Cibal and Talibura) and 21% (Lambaleda).
124
from arising, and/or being reported if it did, by not allowing KDP to operate as it is meant to. In
the fieldwork we found examples of funds being distributed in equal proportions amongst
villages without competition based on the proposals; the program was not socialized properly, so
elites faced little dissent when they captured the program. However, this conflict avoidance,
while limiting tensions in the short run, can provide a basis for future conflict. While the survey
does not capture the knock-on effects of conflict avoidance strategies, the qualitative fieldwork
showed some of the risks.178 As we discussed in Chapter 2, preventing competition can lead to
other forms of conflict in the future, or to the problems reemerging in other forms of grievances.
That is, interaction conflict is more likely. Indeed, the lower level of reported KDP-related
conflict in the low capacity areas appears to be in part an indication of the program not working
as intended.
A second pertinent factor concerns respondents’ perceptions of what constitutes conflict. The
fieldwork found that an important precondition for communities to effectively manage conflict is
for them to first identify it. In areas where local capacity is greater—that is, where people are
generally more able to deal with conflicts as they arise—there was a greater recognition of the
problems. The presence of grievances which are expressed rather than repressed, and which are
more frequently reported, is not necessarily a negative thing; they may be more quickly resolved
than those which fester below the surface. As implied above, we can thus consider high reported
rates of non-violent KDP-related conflict as a positive sign of the program operating well.
Levels of KDP-Triggered Conflict: The Impact of Village Capacity

Importantly, however, the picture changes somewhat if village level capacity is factored in. First,
while KDP-triggered conflict is more likely (according to the respondents) in high capacity districts,
it is less frequently reported in high capacity villages (Figure 7.2) across all forms of district capacity.
Kecamatan-level conflicts are more frequently reported than those at the village level; the former
are captured in the district capacity measures but not those at the village level. Village capacity,
unsurprisingly, has little impact on the levels of conflict at kecamatan (sub-district) meetings.179
When KDP-triggered conflict at the village level does occur, the data from the survey respondents
demonstrates that this is more likely in the low capacity villages across all forms of district capacity.
More people reported that KDP triggered conflict in low capacity villages (40%) than did in medium
capacity villages (39%) or high capacity villages (24%) across all research areas.
Furthermore, when village-level KDP triggered conflict occurs in high conflict management
capacity districts, it most often occurs in low capacity villages. In other words, district and
village level capacity appear to work in opposite directions when it comes to the likelihood of
KDP triggering conflicts. As Figure 7.2 shows, low capacity villages in high capacity districts
are most likely to report that village level KDP meetings triggered conflict (58% of respondents)
across all regions. At the same time it is also the low capacity villages in the low capacity
178
The survey does not capture the seriousness of conflicts in terms of the numbers of people involved, scale of the
impact, nor the length of time it took to resolve the dispute.
179
Very similar level of informants—70% in high capacity villages, 71% in medium capacity villages, and 72% in low
capacity villages—reported that KDP triggered conflict in kecamatan (sub-district) meetings. This is not surprising,
given that a kecamatan is likely made up of many villages with different capacity, and that our villages were not
selected to be representative of the kecamatan in which they lie.
125
districts which report the most KDP-triggered conflict within these districts (23% of respondents
compared to 13% in high capacity villages).
How do we explain this? In line with previous discussion and findings, this demonstrates that
district level capacity contributes to villagers’ knowledge of their rights and KDP processes, and
the likelihood that they will recognize, respond to, and seek solutions for problems. However,
(a) the frequency of these problems is more likely in low capacity villages; (b) the likelihood that
they understand the program is lower (also adding to tensions); and (c) the likelihood that they
have the capacity to manage the more difficult problems is lower—all accounting for the interesting
findings of opposing forces.
District level capacity appears to be more important than village capacity; the differences between
high and low capacity districts are greater across all village capacity types than are the differences
between villages within each district capacity type. This suggests that district and village capacity
interact differently with KDP. We discuss this further below.
Figure 7.2: Local Capacity and Whether KDP Forums Triggered Problems/Conflicts
KDP Forums Triggered Problems/Conflict
Forms of KDP-Triggered Conflict and Forum and Facilitator Use

Previous chapters outline the three different types of KDP-related conflict: in-built conflict
relates to competition at various stages in the project cycle; program malfunction conflict
occurs when project processes and rules are not followed as intended and the program breaks
down; interaction conflicts are those indirectly caused by KDP, where the program interacts with
existing tensions or power structures.180 Using the evidence from the key informant survey presented
below, the following discussion demonstrates that local capacity (at the village or district level) has
little effect on the extent to which KDP forums are used to address in-built conflict—they are
used in all cases. It also has little effect on the likelihood that KDP processes will successfully
180
See discussion in Chapter 3, Section 2.
126
resolve either in-built or malfunction conflicts when they are used. However, local capacity
does appear to have an impact on the likelihood that KDP processes (forums and/or facilitators)
are used to address KDP malfunction conflicts. As with the findings in the previous section,
district and village-level capacity work in opposite directions. However, this time it is the
districts with lower conflict management capacity that are more likely to use the forums.
Forums
Local capacity has little effect on the use of KDP forums for in-built issues. Rates of forum use
for this type of conflict are high across areas with differing capacity; 79% of survey informants in
both high and low capacity districts said that KDP forums were used to address in-built conflicts.
At the village level, rates are also similar, the sole outlier being high capacity villages in high
capacity districts, where use of forums is lower.181 This could suggest that either high capacity
villages are more accustomed to dealing with the type of minor in-built problems generated by KDP,
and hence do not consider them conflicts, or that they resolve their in-built KDP problems before
they even reach the KDP forums, or through other means. Evidence from the qualitative fieldwork,
and other survey evidence, suggests the former. One of the impacts of KDP is that KDP-type
processes, and the social changes they bring about, become more widely accepted. As we
discuss below (Section 3), these changes are likely to be greater in high capacity areas; i.e.,
areas that are conducive to change.
In contrast, local capacity appears to contribute to variations in reported levels of forum use for
program malfunction conflict. In low capacity districts, the use of KDP forums for addressing
program malfunction conflicts is higher: 45% of informants said forums were used for this type of
conflict in low capacity districts in both provinces, compared to 23% in high capacity districts.
The one exception is the use of forums in high capacity villages. Here, rates are similar across
high and low capacity districts, suggesting that village capacity compensates for lack of district
capacity and forums are recognized as a means of resolving problems when district conflict
management capacity in general is low (Figure 7.3).
Figure 7.3: Local Capacity and Whether KDP Forums Were Used
for Program Malfunction Conflict
KDP Forums Used for Program
181
The reported level is 58% compared to a range of 73% to 89% for the other village/district capacity combination
types.
127
Facilitators
KDP facilitators are also more likely to be used to successfully address KDP problems in low
capacity districts than in high capacity ones.182 Among informants in the low capacity districts,
64% said they were successfully used; 43% did so in the high capacity districts. Again, in low
capacity areas they resort to program mechanisms to seek to resolve problems where capacity to
problem solve is generally low. Similar to our previous findings, village level capacity works in
the opposite direction. High capacity villages in low capacity districts, where knowledge of
program is better but where capacity to manage problems is lacking, were the most likely to use
facilitators (68%); low capacity villages in high capacity districts were the least likely (33%).
In general, the higher reported rate of use of forums and facilitators in low capacity areas is not
surprising. The measure in the survey conflates two things: the incidence of conflict of a given
type, and the use of the forum to deal with it. Levels of more serious program malfunction conflict
(which previous chapters have demonstrated is more problematic) could be more frequent in lower
conflict management capacity districts. In all development programs, program malfunction conflicts
are more likely to occur when socialization has been poor, or where local elites have been resistant
to sharing local decision-making responsibilities (see Chapter 3)—all frequent findings in the
fieldwork in low capacity districts. However, it may also be that reported rates are higher because
in low capacity areas people have fewer avenues where they can take their problems to seek forms
of redress, compared to higher capacity areas. The fieldwork shows that both hypotheses have
some explanatory power.
The Success of Forums and Facilitators in Managing KDP-Triggered Conflict

Reported rates of KDP-triggered conflict are higher in high capacity districts. Conflicts related
to in-built competition are equally likely to be addressed in KDP forums in high and low capacity
districts. Conflicts related to malfunctions in the program are more likely to be addressed in
forums in low conflict management capacity areas, and also successfully addressed in these
areas by facilitators. However, when KDP forums are used, regardless of the type of KDP
problem/conflict, they are highly successful across areas with varying capacity. Where in-built
conflicts are addressed in KDP forums, they are slightly more likely to be resolved in high capacity
districts than in low capacity ones, but the difference is small and success rates are uniformly
high in terms of resolution (97.6% versus 92.7%). Conversely, low capacity districts report
higher success rates of resolution for malfunction conflicts (85.7% versus 70.9%); see Figure
7.4. Despite these differences, all areas show extremely high levels of success. There are no
clear patterns when we factor in village capacity, but the fieldwork in these areas presented in
previous chapters demonstrated that in villages with lower conflict management capacity, the
likelihood that facilitators or forums would actually be successful was also dependent on the
extent to which they integrated informal leaders and mechanisms into the resolution process.183
182
The survey question refers to the “successful” use of KDP facilitators “outside of KDP forums”. We thus do not
have quantitative information on when they are used unsuccessfully, and/or their role inside forums.
183
For in-built conflicts, however, the success rate is lower in areas with low capacity at both the village and district
levels (78% success rate) than in any other village-district capacity combination (which range from 93%-100%). Sample
sizes are too low for malfunction conflicts to be disaggregated for both village and district capacity levels.
128
Figure 7.4: Local Capacity and Successful Resolution by KDP Conflict Type
In-built (n=248)
The data thus shows that variation between regions is greater in terms of the levels of use of
forums and facilitators for addressing different types of problems, than for whether or not they
are successful once the problem is introduced to these KDP mechanisms. While levels of success
do vary, they are still high when we aggregate responses at the district level. However, we need
to consider the seriousness and cause of the problems in lower capacity areas to account for the
30% (high capacity districts) and 14% (low capacity districts) of respondents who did not consider
the program malfunction conflicts to be successfully resolved through KDP. The key informant
survey data presented in Chapter 3 outlines how KDP forums and facilitators were often unable
to resolve corruption cases or those involving KDP facilitators as the source of the problem at the
village or sub-district level.
Given that many respondents conveyed that conflicts are successfully resolved once they are
brought to KDP forums or facilitators, attention must focus on why people in some areas chose to
bring malfunction conflicts to KDP forums, and why KDP facilitators get involved outside forums,
whereas they do not in other areas. Understanding better the criteria used by actors in different
places to decide when a particular forum is an appropriate place, or whether a facilitator should
get involved, helps in determining why there are different levels of use.
There are several reasons for this. The results support the conclusion that communities, or
particular elites within communities, assess the likelihood of successful resolution of a given type
of problem through a particular dispute resolution forum or mechanism before opting for its use.
Both the qualitative fieldwork and other empirical research on dispute resolution in villages
indicate that this kind of selection process occurs.184 Community leaders tend only to become
184
Ellikson (1991) uses the concept of “controller-selecting rules” to account for this phenomenon.
129
involved in disputes that they feel they can ably resolve based on their past experience, and/or that
will improve their standing/status within the community.185
Local norms also matter. In understanding how local capacity matters in different ways for different
kinds of KDP problem (in-built and malfunction), and different ways of addressing them (inside
forums and outside by facilitators), it is instructive to distinguish between uses which are perceived
as “normal” and those which are (or could be) considered “extra-ordinary” and “extra-ordinary
plus”. For example, the discussion of in-built tensions in KDP forums is considered by both program
staff and participants as usual and normal, whereas their deployment for particularly sensitive KDP
problems (e.g., allegations of corruption or misappropriation of project funds by the village head or
program staff) can be considered “extra-ordinary”.186 The use of KDP personnel and mechanisms
to address non-KDP problems (e.g., disputes over property boundaries, or problems generated by
other development projects), on the other hand, can be seen as “extra-ordinary plus”.
When survey respondents were asked why in-built problems were resolved in KDP forums, the two
most popular responses were that there was no other place (East Java) or because it was the appropriate
place (NTT). This did not vary across areas with differing local capacities. In contrast, the use of
forums for complex malfunction conflicts (except those which involve unintentional ‘problems of
omission’) such as those relating to corruption, and the active intervention of facilitators in conflicts
outside of KDP forums, is more unusual. Local capacity at both levels appears to influence the
processes through which people identify and select KDP forums and/or facilitators to deal with
program malfunction problems. These “extra-ordinary” uses are more common in low capacity
areas for a number of reasons. First, the need is greater; levels of program malfunction conflict
tend to be higher. Second, few other avenues exist to deal with these problems; other actors, such
as local government, may be less likely to intervene—indeed, in many cases they are implicated in
the conflict. As we discuss below, the different interaction of local capacity with “normal” and
“extra-ordinary plus” uses of KDP forums and facilitators also helps explain their greater use for
non-program conflicts in low capacity areas.
2.2 Program Functionality and KDP-Triggered Conflict
Variation in the extent to which KDP functions as intended did not affect aggregate levels of
conflict in our research areas. However, the ways in which the program functioned did affect the
types of KDP-related conflict that were most likely to emerge. More serious forms of conflict
(related to program malfunctions and interaction with existing tensions) were, unsurprisingly,
more likely in areas where KDP did not work well. The quality of program functionality is thus
of great importance, because the forms of program-related conflict that erupt in areas where KDP
is not working as intended are far more likely to have negative consequences than the kinds that
emerge where KDP works as intended, which can often have positive impacts.
185
The ongoing Village Judicial Autonomy study, within the World Bank in Indonesia, has found much evidence to this
effect. The papers can be found at www.justiceforthepoor.or.id.
186
Just as one might be reluctant to go to the police with allegations of police corruption, so too might one expect
villagers to be unlikely to use KDP forums as a place for raising concerns pertaining to the inappropriate actions of
local power-holders.
130
In Chapter 3, we demonstrated that three types of conflict are associated with participatory
development projects like KDP. The impacts of each form tend to be different. Conflict related to
competition—a consequence of the principles of open decision-making, with not enough resources
available for all ideas and proposals to be funded—was, we argued, generally a positive phenomenon.
It is through such competition that some of the cognitive and behavioral changes that can help
improve conflict management capacity in the long-run (see Chapter 6) are generated. Rather, it
was lack of opportunity for open competition, or competition without a level playing field, that was
more likely to lead to prolonged tensions or destructive conflict.
This was the case in Jenangan sub-district in Ponorogo district. As in Pademawu sub-district in
Pamekasan, it was decided (against program rules) that in Year One of KDP-I that only five villages
were eligible for the program that year, the criteria being those which had previously been funded
as part of the IDT project. This caused tensions, with villagers protesting the lack of competition in
Year One, especially in later years when they learned more about KDP rules.187
Disputes related to competition were thus more likely in areas where the program functioned
well. Where the program was not allowed to function as intended (with program rules subverted,
money corrupted, etc.) disputes of this type were less common. The incidence of micro-conflicts
related to the competition embedded in the KDP system is, we have argued, a direct sign of both
good program functioning and an indirect indicator of positive spillover. While this kind of
development-related conflict is more common in cases where KDP works well, conflicts of this
sort (directly or indirectly) in most cases have a net positive effect.
In contrast, program malfunction conflict—where ‘problems of omission’ (such as poor

socialization and implementation) or ‘problems of commission’ (e.g. corruption) lead to disputes—
and interaction disputes—when projects interact with pre-existing local tensions, power structures
and conflicts—tend to have negative impacts. Program dysfunctionality is at the root of both
forms of conflict. As we explored in depth in Chapter 3, these conflicts tend to have negative
impacts, both in terms of the seriousness of conflict that emerges and the difficulty of solving
such problems.
Figure 7.5 shows schematically the different forms of project-related conflicts that are likely to
occur in places where program functionality is good and where it is poor.
187
Sub-district Head of PMD, Jengangan, Ponorogo (1303); Saving and Loans Group Head, Jenangan, Jenangan,
Ponorogo (1304). This is a program malfunction conflict, in that the intended KDP rules and procedures were not
applied.
131
Figure 7.5: Types of KDP-related Conflict and Program Functionality
FORM OF IMPACT
Direct Indirect
FUNCTIONALITY
Ia. Direct Ib. Indirect
PROGRAM
Good (+) In-built In-built

competition competition
Poor (-) II. Program III. Interaction

3. The Effect of Local Context on the Direct Impacts of KDP on Conflict Management
Capacity
KDP forums and facilitators are infrequently used to deal with conflicts unrelated to the program.
Thirteen percent of survey respondents in East Java, and 20% in NTT, said that non-KDP problems
were addressed in KDP forums. KDP facilitators almost never address non-KDP conflicts outside
of the forums.188 Because facilitator use for non-program conflicts is so low, we are unable to
disaggregate by capacity level.
3.1 The Effect of Local Capacity on Direct Impacts
Use and Success of KDP Forums for Non-KDP Conflicts

Local capacity appears to influence whether KDP processes are utilized for what might be termed
“extra-ordinary” KDP uses. As we discussed above, KDP forum use for program malfunction
issues is higher in low capacity areas, with district capacity exerting a larger influence than village-
level capacity because higher levels of the administrative hierarchy are required to address the
more serious problems. A similar pattern is observable for “extra-ordinary plus” uses, namely
addressing non-KDP conflicts in KDP forums. The most common non-KDP conflicts addressed in
forums relate to other development projects. In East Java, 36% of respondents who reported that
KDP forums were used for non-KDP issues, and 58% in NTT, said they were used for conflicts
related to development. Other kinds of conflict addressed included disputes between villagers,
problems with thieves, and conflicts with other villages.
When the data is disaggregated at the district level, low capacity areas are consistently more likely
to utilize KDP forums for non-KDP problems/conflicts, and the conflicts are also more likely to be
successfully addressed once there. The use of these forums is two and a half times more likely in
low capacity districts than in high capacity ones (see Table 7.1).
188
In NTT, 4% of informants said they did; none in East Java agreed.
132
Table 7.1: Local Capacity and Use of KDP

Forums for non-KDP Conflicts
District Capacity Use Success

High (n=118) 8.5% 84.6%
Low (n=110) 20.9% 92.9%
Closer examination of the data shows what is driving this difference. Whereas the use of KDP
forums for addressing problems with other development projects was similar across districts
with different capacity, non-development issues (e.g., other conflicts within and between villages)
pushed up the higher use of forums for non-KDP issues in the low capacity districts.
This hints at the reason why such usage of KDP forums is greater in lower capacity districts. In
these areas, fewer institutions exist for effectively dealing with conflict; this finding is supported
by the qualitative data presented in other chapters. Indeed, in the low capacity sub-districts, the
main cited reasons for KDP forums being used (and for them being successful) related to a
combination of appropriateness and necessity: the KDP forums/processes were viewed as being
good for problem solving (in Lambaleda sub-district), as were the skills of the sub-district
facilitators (in Ruteng); in Cibal the lack of other options was cited. This is an important finding,
especially when thinking about applying KDP-type development projects into high conflict areas
where capacity is low. It suggests that where the program works well and has a good reputation,
and in areas where conflict mediation capacity is otherwise limited, KDP may be able to fill a
role in addressing certain kinds of conflict. In contrast to the findings for conflicts arising from
program malfunction, village capacity has little effect on the likelihood of conflicts unrelated to
KDP being addressed in program forums.189
The Different Impacts of District and Village Capacity

Table 7.2 summarizes the impacts of capacity at different levels on the variables discussed above.
A number of points are clear. First, district capacity matters more than village capacity in determining
program impacts. Second, for most variables, impacts are greater in low capacity districts. Third,
in many cases, district and village capacity work in opposite directions.
From these results it is clear that district capacity is more important in determining the effects of
KDP than is capacity at the village level. The table shows that for five of the nine variables,
village capacity had no or negligible effect, compared to only two variables for district capacity.
189
The reported rates of forum use for non-KDP problems are: 13% (high capacity villages), 15.4% (medium capacity
villages), and 14.9% (low capacity villages). Sample sizes are 77, 104 and 47, respectively.
133
Further, the difference in impact between low and high capacity districts is generally higher than
the difference between low and high capacity villages.190 This all suggests that KDP’s effect on
numerous dimensions of conflict and conflict management varies more between districts than villages
with different levels of capacity. In understanding why this is so, it is necessary to consider why
greater effects are observed in low capacity districts.
Table 7.2: The Impacts of District and Village Capacity on Change Variables
Form of Capacity
Variable District Capacity Village Capacity
Conflict Levels
(a) Level of KDP-triggered conflict High capacity, larger impact Low capacity, larger impact
Level of use of KDP mechanisms for …
(b) In-built conflict No pattern/effect No pattern/effect
(c) Program malfunction conflict Low capacity, larger impact High capacity, larger impact ∗
(d) Non-KDP conflict Low capacity, larger impact No pattern/effect
(e) Facilitators using outside forums Low capacity, larger impact High capacity, larger impact
When used, level of success of KDP mechanisms for …
(f) In-built conflict No pattern/effect No pattern/effect

(g) Program malfunction conflict Low capacity, larger impact No pattern/effect
(h) Non-KDP conflict Low capacity, larger impact No pattern/effect
(i) Success of facilitators Low capacity, larger impact High capacity, larger impact
*Only in high capacity districts
The table shows that KDP forums and facilitators are used more often for different forms of conflict
(both those related and unrelated to KDP), and that they are more successful, in low capacity
districts.191 The evidence points to the fact that KDP forums and facilitators are used most where
they are needed in situations where problems arise. For problems unrelated to KDP, the reasons are
clear: in areas with low capacity, fewer effective institutions and mechanisms exist to deal with
problems and conflicts (which we saw in the examples provided in Chapter 3).
It is not surprising that, for these types of conflict, it is district-level capacity that matters more
than village capacity. The newspaper data showed that levels and forms of conflict tend to vary
more at the district level than at the provincial, sub-district or village levels. Conflict is concentrated
in particular districts, and not in others.192 The reasons for this are both cultural and structural
190
The gap in reported rates of KDP-triggered conflict is 31.3% between low and high capacity districts, compared
with 16.4% between low and high capacity villages. Other differences are: use for program malfunction conflicts
(21.3% district, 10.4% village); use for non-KDP conflicts (12.4% district, 2.4% village); facilitator use (21.2% dis-
trict, 13.4% village); and successful use for malfunction conflicts (14.8% district, 10.5% village). The sole exception
is for successful use of forums for non-KDP conflicts. Here, village capacity matters more than district capacity: there
is a 16% gap between high and low capacity villages in successful use, compared with a 7.7% gap at the district level.
191
The one exception is for in-built KDP conflict where district capacity has little effect. This, as we argued above, is
because this is a normal and routine use for KDP.
192
This backs up previous evidence from the UN, which showed conflict to be geographically concentrated in a small
number of districts (Varshney, Panggabean and Tadjoeddin 2004).
134
(Barron and Sharpe 2005), each shaped by the existing institutions in an area. Many important
institutions—both formal and informal—at the district level are involved in resolving larger-scale
conflicts. As we noted earlier, the district now has tremendous policy-making discretion, far more
so than at lower levels, or the provincial level (McCarthy 2004).193 Local elites also tend to be most
powerful and cohesive in district capitals (Barron, Nathan and Welsh 2005).
The research also showed that most problems at the village level are able to be addressed effectively
if the scale of the problem is limited to the village. It is only larger problems and conflicts—such as
those between parties from different villages—which tend to escalate, in large part because of a
lack of effective inter-village mediation mechanisms.194 In real ways, it is institutions and actors at
the district level that structure village level norms pertaining to government administration and
decision-making. There is less variation in demand for new mediating institutions between villages
in districts with similar levels of capacity than between villages in districts with differing capacity.
For problems related to KDP, use of forums and facilitators is also greater in low capacity areas.
This, as we argued above, relates to the increased likelihood of program malfunction conflicts in
districts with low capacity and the nature of the problems, which require higher levels of authority
and intervention for resolution.
However, village level capacity is not irrelevant. Indeed, the data shows that while district level
capacity largely determines the need for KDP forums and facilitators to act as conflict resolution
mechanisms, village capacity has an impact on the likelihood of them being used, given a certain
level of need. A certain amount of capacity is required at the village level in order for KDP
mechanisms to be used in non-simple ways, such as for program malfunction conflict. Local
leaders in low capacity villages may be less supportive of the use of institutions which they do
not fully control for problem resolution.
Indeed, as we showed earlier, in places such as Pamekasan in East Java (a low capacity district),
many village elites saw KDP as a threat to their authority, and hence undermined its use. Only in
villages with higher capacity (such as Padellagan or Sumedangan) was KDP able to operate
effectively. In these villages, KDP worked relatively well, with the community involved in the
planning process, in contrast to other programs:
“In my opinion, the KDP process is the only program that involves the community. From the planning process to the
implementation, each component is involved from the beginning. This is different from P3DT [Infrastructure Development
Program for Less-Developed Villages] where the community is not involved in the processes.”
Golkar member, Padellegan, Pademawu, Pamekasan (1204)
It should be noted, however, that this only holds for uses most closely related to the program—e.g.,
forum use for malfunction conflicts, and facilitator use.
The evidence suggests that local capacity does matter in determining the use of KDP for different
types of problem. Use depends in large part on the need, and this in turn depends on the presence
or absence of existing institutions. However, even in areas where need is great, it is necessary
that there is a minimal level of capacity present to allow the program to fill the existing gap.
193
Although there is talk of the Government, fearing decentralization has gone too far, taking measures to strengthen
provincial level governments.
194
The two provincial reports, especially the NTT one, address these issues (Satu and Barron forthcoming; Diprose
forthcoming).
135
Thus, while district capacity may be more important in determining usage, attention should
still focus on building village level capacity.
3.2 Program Functionality and Direct KDP Impacts
KDP forums and facilitators are far less likely to be used to address conflicts not related to the
project in areas where the program is not functioning well. As explored earlier, the use of such
spaces and people for non-project-related conflict resolution is rare and ad hoc. Forums and
facilitators were most likely to be used for this purpose when either (a) facilitators or other program
staff took the initiative to intervene in such disputes, or (b) where the parties to the conflict, or the
community more broadly, took the dispute to KDP staff and/or meetings. For a number of reasons,
both scenarios are much less likely in places were KDP is not implemented well.
First, KDP facilitators and staff are much less likely to agree to the program or forums being
used for conflict resolution performance when program performance is poor. In such cases,
they may (rationally) feel that the program is already overwhelmed with dealing with problems
internal to the program, and hence are not be keen to take on extra tasks that are not prescribed
within the PTO (operational guidelines). It is also true that, in general, poor program
implementation is the result of poor performance from local facilitators and staff. In the research,
we found, time and again, that the best facilitators were those who exercised autonomy and
creativity to deal with the numerous issues that arise during the project cycle. The best facilitators
are able to know how to apply the spirit of KDP rules, rather than just blindly implementing the
letter of the law; they are able to ascertain when “extra-ordinary” and non-prescribed uses of
KDP mechanisms, such as dealing with disputes not directly linked to the program, is suitable.
Weaker facilitators are less likely to be able or willing to exercise such discretion. As a result,
weak program staff may be less keen on using KDP for conflict resolution purposes, even in
lower capacity areas where there may be a real need and demand for this.
Second, we found that people were less likely to take problems to KDP forums or staff when
they did not trust their local facilitators. As we discussed in Section 2.1 of Chapter 4, mediators
must have both institutional and personal legitimacy if they are to be effective. Where the
public views either the program or particular staff as tainted, they are much less likely to bring
non-project-related problems to project forums. In places like Lambaleda in Manggarai, where
a sub-district facilitator (FK) ran off with USD 9,000, trust in KDP diminished quickly.
It is hardly surprising that villagers do not trust the program or see it as an appropriate avenue for
dealing with non-project problems when front-line (or other) KDP staff have proved themselves
incapable of administering the project in a fair way.195
195
Interestingly, in Lambaleda a number of informants still felt that KDP forums were good for problem solving.
This shows the extent to which community members are able to distinguish the actions of particular individuals
within the program (e.g., the FK) from the worth of the program overall. Informants told us that the fact that the case
had been followed-up by the project bureaucracy had restored people’s faith in KDP, although they were frustrated
that at the time of the interviews (February 2005) there had not yet been a conviction, with the case stuck in the
provincial court in West Timor.
136
Third, and related, we have already noted how people tend to bring their problems to venues
where they feel resolution is most likely. This accounts for the high success rate of KDP forums
and facilitators for all types of problems that are brought to them. In areas where project
performance is poor, it is unsurprising that people feel that resolution of a given problem is not
likely. The result is that they are less likely to bring the problem forward in the first place.
KDP is unlikely to have any positive direct effects on the resolution of local conflicts if the
program does not function as intended, and this is true in both high and low capacity areas. If we
want KDP forums or facilitators to be used for conflict resolution, a first order task is thus ensuring
that the program functions well.
4. Indirect Impacts and Local Context
Whereas the direct impacts of KDP on conflict management tend to be more noticeable in low
capacity districts, the opposite holds for indirect impacts. In the last chapter, we examined the
effect of KDP on three dimensions of social change: group relations, behavioral change, and
normative change. The evidence shows that local capacity is not particularly important in determining
changes in relations between groups. However, for most dimensions of behavioral and normative
change, impacts are more noticeable in high capacity areas. This suggests that while KDP stimulates
processes of positive change, it does not cause them alone. Rather, KDP is a catalyst that has a
multiplier effect on ongoing processes of social change; where these are already underway, indirect
KDP impacts are likely to be greater.
4.1 The Effect of Local Capacity on Indirect Impacts
Group Relations
The survey results and qualitative fieldwork indicate that KDP strengthens relations between
different identity groups. Where a given identity cleavage is salient (e.g., ethnicity, religion,
class or political affiliation), KDP contributes to improved relations. The data from Chapter 6
note that group relations were on average 17% more likely to improve in treatment than in control
locations. However, the survey data shows little correlation between the preexisting capacity in
an area and the observed extent of change in group relations. There is considerable variation, yet
no patterns emerge. For example, change in relations between the rich and poor is greater in high
capacity districts, for both treatment and control sites, most likely due to democratization trends
across the board. At the same time, the survey data shows greater change in relations between
religious groups in low capacity districts. For ethnicity in NTT, high capacity districts in KDP
sites report slightly greater change; in control sites, the opposite is true, with low capacity areas
experiencing greater change. Similarly, there are no clear trends when we factor in village
capacity.196
That local capacity has little effect on group relations is at first surprising. One would think that
the presence of capacity as we measured it—i.e., where problem solving and collective action
are strong—would be reflected in positive changes in group relations. Yet, two reasons present
196
The full results are given in the summary of the key informant survey; see Barron, Clark, and Mawardi (2004).
137
themselves as to why this might not be the case. First, the survey question is measuring change
in group relations over a given period rather than the strength of inter-group relations at a given
point. Presumably, capacity is to some extent endogenous—that is, it is deeply embedded in
local social structures and cultural practices—and as such is hard to ‘change’. Thus in high
capacity areas, group relations may have been strong in the first place, and thus reported rates of
change in group relations may be smaller than in areas where such capacity did not exist in the
first place.
Second, for there to be a change in a given type of group relation (e.g., ethnic, religious, etc.),
there needs to be a relevant initial cleavage of that type. In Chapter 6 we discussed changes in
ethnic and religious relations only in Flores; these cleavages are not important markers of identity
in more ethnically and religiously homogenous East Java. Variation in the presence of a given
cleavage exists not only between provinces, but between districts and, indeed, between villages
in the same district. Some of our villages in Flores, for example, have more prominent ethnic
cleavages than others; in East Java, the rich/poor divide is greater in some locations than others.
There is no clear correlation between variation in the presence and strength of identity cleavages
and local capacity.
As Anderson (1991) and others have reasoned, identity formulation and transformation are contingent
on demographic, political, economic, cultural, policy, and legislative environments at multiple
levels and grievances towards changes in these environments.197 Across the cases, we found
three primary factors relevant to the construction, maintenance and salience of group difference:
(a) history as institutional legacy and consolidation; (b) history as the invocation and re-imagining
of the group’s biographical narrative; and (c) appeals to present interests and future aspirations to
sustain group size and vibrancy.198 Local capacity will matter in shaping these factors. For example,
in an area with a history of collective problem solving, it is more unlikely that leaders will appeal to
narrow interests that emphasize the difference between one group and another. These factors are
relevant, but demographic factors—e.g., the heterogeneity of a given population—are equally so.
The variation between districts and villages in the degree of changes in group relations is explained
more by the different demographic characteristics and the different histories of our research locations
than by ‘capacity’ itself.
Behavioral Change
Whereas the data shows no clear link between preexisting local capacity and changes in group
relations, the same is not true for behavioral changes. Earlier, we examined two dimensions of
behavioral change: participation and decision-making. Disaggregated at the district level, levels of
participation improved to a greater extent in low capacity areas. For decision-making, in contrast,
greater improvements were noted in high capacity districts.
197
On this, see also the work of Eric Wolf (1964; 1999) who has shown how structures and power relations shape
cultures.
198
See our discussion of the “dynamics of difference” in Barron, Smith and Woolcock (2004, pp. 21-29).
138
In Chapter 6, we showed that villages which have participated in KDP showed greater increases
in the number of groups coming to other non-KDP village meetings than did those which had not
had the program, and that increasing numbers of groups came as KDP stayed in an area for
longer periods of time (Figure 6.5). If we compare treatment and control sites, it seems that the
impact is concentrated heavily in low capacity districts (Figure 7.6). While the rates of positive
change were almost the same for treatment and control areas in high capacity districts, positive
change was 25 percent larger in villages in low capacity districts that have had KDP, than in
villages in low capacity districts which have not.199
Figure 7.6: Local Capacity and Positive Changes in Participation in non-KDP Meetings
(KDP vs. Control Sites)
Positive Change in non-KDP Meetings
In contrast, the impact of KDP on local decision-making appears to be larger in high capacity
districts. Survey respondents were asked whether decision-making had become more democratic
since KDP had arrived (or for an equivalent amount of time for control locations). As we discussed
in Chapter 6, improvements in democratic decision-making were greater in KDP areas than in non-
KDP areas. Figure 7.7 shows that in KDP areas, decision-making is likely to have become more
democratic in high capacity districts. Further, in low capacity districts, village capacity is important:
only 35% of informants living in low capacity villages in low capacity districts noted positive
change, compared to 72% of respondents in high capacity villages in low capacity districts. This
clearly shows that a certain level of local capacity is necessary for decision making to become more
democratic, but when this capacity exists KDP complements the process.
199
As for many of the direct impact measures, while low capacity districts are more likely to report more groups coming
to meetings, village capacity works in the opposite direction, with high capacity villages more likely to report improve-
ments than medium or low capacity ones (76% versus 68% and 66%, respectively).
139
Figure 7.7: Local Capacity and Positive Changes in Decision-Making
90%
Positive Change in Decision-Making
80%
70%
60%
50%
40%
30%
20%
10%
0%
If we compare control and treatment areas, it is clear that, at least at the margin, KDP can contribute
to decision-making becoming more democratic in low capacity areas. As Figure 7.8 shows, in low
capacity districts, 23% more people in KDP districts report that decision-making is more democratic
than in non-KDP areas.
Figure 7.8: Local Capacity and Positive Changes in Decision-Making (KDP vs. Control Sites)
140
Thus local capacity has differential effects on the likelihood of more groups participating in
village meetings, and on the probability of decision-making becoming more democratic. This
can be explained by the different level of discretion and control that local elites have over the two
variables. The qualitative fieldwork showed that in the low capacity districts (Pamekasan and
Manggarai), local elites tended to protect more closely their monopoly on the exercise of political
power. This was also true at the village level. In low capacity areas, elites tend to prefer
participation to be limited to a closed group of people. Decision-making also tends to be more
authoritarian. In high capacity villages, however, the villages had developed strengths to break
free of district-wide cultures of patronage system, as was the case in Padellegan and Sumedangan
villages in the low capacity district of Pamekasan. KDP is running quite smoothly in these areas,
with higher levels of women’s participation, less complaints of poor socialization, and better
competition compared to the other research villages in the district.
It is relatively easy for elites to control decision-making processes within meetings. Given their
dominant role in village life, they can fairly easily hijack the agenda and dictate the direction of
meetings. It is thus not surprising that it is only in areas with higher capacity (at both the district
and village level) that KDP has an effect. Here, elites tend to be more open to having inclusive
decision-making processes, and thus KDP is allowed to function as intended.
Yet in contrast, local elites have much less control over participation in meetings. The research
showed that the primary factor driving low attendance in village meetings was that people simply
were not invited. A separate study found that distributing formal invitations to KDP meetings more
widely increased participation in KDP accountability meetings by 35% and non-elite attendance by
80% (Olken 2005). If KDP facilitators perform well, it is thus relatively easy to improve participation
in meetings, at least compared to changing decision-making patterns within them. The behavior of
elites thus matters less in determining patterns of participation than methods of decision-making
which ensure that marginalized groups actually secure a serious voice in decisions affecting their
welfare (see Gibson and Woolcock 2005).
Why would increases in the participation of marginalized groups be greater in low capacity
areas? Two primary reasons emerged from the qualitative fieldwork. First, the starting point is
lower; in low capacity areas, fewer groups tend to participate in village-level decision-making,
and hence there is more room for improvement. As such, improvements in participation are
notable to a greater extent in low capacity areas, both those which have had KDP and those
which have not. Second, KDP compensates to some extent for the lack of capacity in low
capacity areas. In high capacity areas, processes of political and social change have already
changed the nature of who participates in village-level meetings (and how). In low capacity
areas, where this has not occurred, KDP—when it functions well—can help stimulate broad-
based participation.
Changes in Norms
We measured two dimensions of normative change: village problem solving and village conflict
resolution. Figures 7.9 and 7.10 show that for both, capacity matters. Levels of improvement
are consistently, and considerably, higher in high conflict capacity districts. A similar, but less
marked, pattern is visible for village conflict capacity, where improvements in villages receiving
141
KDP are greater in high than medium and low capacity villages. However, the effects of village
capacity are almost entirely confined to villages in low capacity districts.
Figure 7.9: Local Capacity and Positive Change in Problem Solving

Positive Change in Problem Solving
Figure 7.10: Local Capacity and Positive Change in Conflict Resolution

Positive Change in Conflict Resolution
142
This picture is strengthened when we compare KDP locations with the control sites. As Figures
7.11 and 7.12 demonstrate, it is only in high capacity areas where KDP appears to be having an
impact with respect to improving positive changes in problem-solving capacities. Indeed, in low
capacity districts, reported improvements in the indicators of changes in norms are lower in KDP
areas than in non-KDP ones.
Figure 7.11: Local Capacity and Positive Change in Problem Solving (KDP vs. Control Sites)
Positive Change in Problem Solving
Figure 7.12 Local Capacity and Positive Change in Conflict Resolution (KDP vs. Control Sites)
143
Why might this be? Several possibilities present themselves. One explanation is that KDP’s
resources fuel existing tensions between village elites, and that in low capacity environments
KDP participants, processes, and facilitators are unable to stem these tensions, thereby frustrating
villagers and leading them to report falling levels of problem-solving capacity. Alternatively, it
may be that newly ‘empowered’ KDP participants in low capacity villages are both presenting
fresh challenges to prevailing power structures and expecting more of their leaders, in the process
eliciting resistance from village elites that manifests itself as heightened tensions and thus
perceptions that problem solving capacity is falling. It is hard to distinguish empirically between
the relative merits of these possibilities, but both find support elsewhere in our data (as reported
in Chapters 3 and 6, respectively). Either way, these findings should not be interpreted as
arguments for not introducing KDP-type development projects into moderate conflict (and/or
low capacity) environments; rather, they should be read as (a) a serious caution against any belief
that community-driven development (CDD) projects, merely by virtue of being ‘participatory’,
will help foster better conflict resolution skills; and (b) evidence for the importance of
understanding local social structures and how development projects interact with them, in order
to assess overall project efficacy.
This clearly indicates that KDP’s ability to bring about positive changes in norms of decision
making is contingent upon existing capacity, thus further strengthening the conclusion that KDP
can tap into existing processes of change but it is less able to generate that momentum single-
handedly.
4.2 Program Functionality and Indirect KDP Impacts
KDP is unlikely to have positive indirect impacts on conflict management where program
performance is poor. Whenever there are significant levels of elite capture of KDP processes,
corruption of the program, or excessive intervention by external non-program actors, this not
only reduces the effectiveness of KDP as a stand alone program, but also reduces its potential
for having positive spillovers on conflict resolution, institutional strengthening, and democratization.
We wrote earlier of how program malfunction conflicts could be caused by either ‘problems of
omission’ (including poor socialization and non-deliberate weak implementation) or ‘problems
of commission’ (intentional deviations from program rules, including acts of corruption). Both
forms of program dysfunctionality affect the likelihood of the program having positive indirect
spillovers.
Poor socialization limits the impacts of KDP on group relations, behavioral and normative change.
The reasons for this are simple. It is impossible for villagers to learn new routines of behavior
(e.g., new opportunities for participation, changes in the way decision-making should take place,
etc.) if they are not properly told that these new opportunities exist, and if they do not see visible
examples of how these new modes of participation and decision-making work in practice. The
New Order state sent out very clear signals on the role that ordinary villagers should play in the
process of economic development—namely, as that of beneficiaries, rather than active participants.
Strong vertical hierarchical structures limited the chances for villagers to substantively engage
144
in decision-making, aside from in a clearly bounded realm of ‘local activities’ where gotong royong
collective action was encouraged (cf. Rao 2005). In both the ‘treatment’ and ‘control’ villages in
our study, we found that the new discourse of participation that has accompanied the fall of
Soeharto and the electoral democratization of Indonesia was strong. However, this did not always
translate into changes in behavior. Norms are sticky; changes in patterns of behavior normally
lag behind changes in rule systems (North 1990; Ellickson 1991). Many Indonesian villages
remain quasi-feudal, with strong hierarchies as to who can appropriately participate in different
realms of village life.
KDP’s power lies in providing a readily visible example of how local decision-making can be done
differently. Where people do not receive a clear signal that KDP really is different, the program
will not have any impact on modes of participation or decision-making. Without a strong and
active socialization campaign, it is unlikely that the program will change existing norms. The
indirect impacts of KDP do not come naturally and automatically; they require ongoing efforts
from program facilitators and staff to show that the new post-New Order discourse can actually
translate into doing things differently (and ‘better’) at the village level. Such efforts are the very
essence of democratization.
If ‘problems of omission’ limit the indirect impacts of KDP, ‘problems of commission’ are likely
to have an even bigger effect. Impacts are limited when elites capture the project; clearly in such
cases there are fewer opportunities for participation from ordinary villagers, and hence changes
in participation and decision-making are less likely. Likewise, where there are inadequate
opportunities for real participation, group relations are less likely to improve, as KDP forums
may not be places where people from different (conflicting) identity groups can come together
for joint decision-making. Poor targeting—which can stem from biases in decision-making
processes—will likely reduce the incentive for active involvement from groups that lose out,
making improvements in participation and group relations less likely:
“The KDP funds were meant to help the poor, but it was the rich people who received them. This is because they are
revolving funds and they weren’t prepared to take the chance of funds not being repaid. The poor people were afraid.”
FGD with female group, Satar Punda, Lambaleda, Manggarai (M.515)
Corruption is also likely to have a significantly negative effect on degrees of indirect impact. We
cannot state strongly enough that KDP’s impact comes largely from the visual and visceral
precedent it sets. In villages where KDP was working well, where meetings really were open to
all and where decision-making processes were fully democratic, indirect spillover effects on
broader village decision-making were observed. When people see how democratic decision-
making can function, almost without exception they embrace it and push for these changes in
other aspects of village life. However, in cases where “nothing seems to have changed”, such as
where project money is corrupted, cynicism only increases. This, in turn, neuters any opportunities
for significant change, because radical reform is always predicated upon not only demand from
the bottom, but also belief that such reform is attainable (cf. Appadurai 2004).
145
5. Summary
While it is important to recognize the institutional imperatives for statements regarding the
aggregate impact of development projects—at the end of the day, governments need to know
why they should invest finite resources in particular strategies to reduce poverty—it is also
important to “look beyond averages” and explore the sources of variation in project performance.
Such variation is inevitably associated with development initiatives, and especially those with a
strong social/participatory component, since such initiatives rely not on a single homogenous
‘technology’ (e.g., textbooks, immunization shots) but the far more heterogeneous skills and
motivation of people, individually and collectively. Understanding this variation in the
performance of social development projects is vital not only for presenting an honest and accurate
account of project impact, but for identifying key organizational (in KDP’s case, the forums and
facilitators) and contextual (district and village capacities) junctures in the project life cycle that
may help managers and participants alike improve the project’s overall quality.
Figure 7.13 summarizes the results from this Chapter, and indeed many of the key findings of the
study as a whole.200 It shows how both the strength and direction of each of the different forms
of impact (direct and indirect) of KDP on conflict management are contingent upon the extent
to which the program functions well, and the contexts in which it works. Each of the boxes is
filled in with a relative impact score: the negatives (-) demonstrate a negative program effect;
the positives (+) demonstrate a positive effect; and a zero (0) demonstrates no impact. For
both the negatives and positives, the number of plus or minus marks demonstrates the strength
of impact. Hence +++ shows a greater positive impact than +, and so on. The figure
demonstrates the extent to which the impacts of KDP are contingent on the ways in which it
functions and the areas in which it works.
Figure 7.13: The Impacts of KDP on Conflict Management in Different Environments
Context Capacity
Low High
Type of Impact Program Functionality Program Functionality
Low High Low High
Direct
Forums (places) -- ++ -* 0
Facilitators (people) 0 0 0 0
Group Relations 0 ++ + +++
Indirect
Behavioral (participation) 0 +++ 0 +

Behavioral (decision-making) 0 + + +++
Normative 0 + 0 +++
* While we noted higher rates of KDP-triggered conflict in high capacity areas, such conflict is much less likely to escalate and/or turn violent. Hence negative
impacts are greater in low capacity areas, where program functionality is poor.
200
We have split behavioral change into two dimensions: participation and decision-making. The reason is that context
has a different impact on each. KDP is likely to have a greater impact on levels of participation in low capacity areas.
Conversely, impacts on decision-making are greater in high capacity areas.
146
A number of key messages emerge. First, and most broadly, it is hard to discern a “pure” KDP
impact on local conflict; that is, KDP’s programmatic impact is inseparable from the nature of
the social context in which it operates, and the extent to which it is effectively implemented. In
low capacity environments where KDP is poorly implemented (as a result of, say, inadequate
socialization of participants or weak enforcement of program rules), KDP can exacerbate local
conflict by providing a new resource over which elites compete. A well-implemented program
in a low capacity environment can, however, produce positive development outcomes and enhance
conflict management capacity.
Second, ‘capacity’ operates at multiple levels, and can work in multiple—sometimes opposing—
directions, depending on the form of impact we are assessing. The use and success of KDP
mechanisms for addressing conflicts stemming from KDP, for example, varies considerably
depending on specific interactions between implementation effectiveness and ‘district’ and ‘village’
capacity, with district capacity often playing a primary role. Where conflict stems from KDP
malfunction, for example, lower district capacity increases the use of KDP forums and facilitators,
while the forum has little strength in lower capacity villages. Conversely, local capacity has little
effect on the use of KDP forums for in-built issues.
Third, whether KDP is working well or not is more important than the context in which it operates
in determining the level and direction of impact (although, of course, the two are often related).
Where program functionality is poor, hardly any positive spillover effects are observed, with
only marginal positive impacts on group relations and in decision making. In both low and high
capacity areas, getting the program working greatly enhances the positive impacts the program
can have on both conflict management and development outcomes.
Fourth, and related, where the program does not work as intended, positive effects are not only
limited, but there is also a chance of the program having negative impacts. As we discussed
earlier in the chapter (and in Chapter 3), KDP forums can trigger conflicts, or exacerbate existing
ones, when program functionality is poor. Given that program functionality is more likely to be
poor in low capacity areas, attention and resources should be directed towards improving KDP
performance in areas with low capacity.
Fifth, as we have seen, the impacts of KDP on constructive conflict management are primarily
indirect. KDP forums and facilitators are rarely called upon to address non-KDP problems except
as a last resort; KDP itself generates the positive impacts it does largely indirectly, i.e., through
changing group relations (bringing diverse groups together), enhancing negotiation skills
(enhancing the capacities of marginalized groups to ‘engage’ in public debate), and altering
perceptions and expectations (enhancing these same groups’ ‘capacity to aspire’ and the new
standards they require of both public officials and local leaders). There is nonetheless important
variation in these outcomes. Local capacity (at the village or district level) seems to have little
effect on whether or not KDP forums are used to address in-built conflict, and little effect on the
likelihood that KDP processes will successfully resolve either in-built or malfunction conflicts
when they are used. However, local capacity does appear to have an impact on the likelihood of
KDP processes (forums and/or facilitators) being used to address KDP malfunction conflicts.
147
Sixth, KDP does not displace existing forums for local dispute resolution; in high capacity
environments it can serve as a valuable complement, strengthening already well-functioning
institutions, while in low capacity environments it can provide a positive alternative to (substitute
for) absent, captured, or dysfunctional forums. The marginal impacts of a well-functioning KDP
are higher in low capacity areas (because KDP forums need to take on a wider range of tasks),
though a minimal level of capacity is needed to provide a basic foundation on which to build which
is not yet present in some villages. On the other hand, indirect impacts are greater in high capacity
areas, where KDP can enhance, and act as a catalyst to, ongoing processes of political and
social change.
148
Chapter Eight
Conclusions and Recommendations
1. Introduction
This study has examined the links between KDP and local conflict and conflict management.
Drawing on a range of primary data sources from two very different provinces—East Java and
NTT—we find little evidence that KDP per se has either a positive impact on conflict at an
aggregate level or a direct positive impact on conflict at the local level. We find stronger evidence,
however, that KDP generates positive indirect impacts on factors closely associated with a
community’s capacity to manage conflict, namely social relations between groups, transparency
and widespread participation in decision-making, and norms and attitudes regarding how to address
disputes. We find that development projects can cause, or interact with existing, conflict. However,
KDP conflicts are less likely to turn violent than those associated with other development projects
in our research areas. This, in part, is because KDP projects emerge from a community-led
process and hence are less likely to clash with local priorities or needs. It is also because KDP
has a number of in-built conflict mediation and feedback mechanisms that allow tensions to be
addressed as they arise. Perhaps most important, we find that KDP’s impact more generally is
highly contingent on implementation effectiveness (i.e., the extent to which KDP functions as
intended) and local context (in particular, the capacity of both the wider district and the more
immediate social structures to govern everyday village life). KDP can minimally compensate
for poor local governance, but is most effective when it complements broader reform initiatives
in well-governed districts and villages.
These findings suggest a number of areas to which attention should be focused by policymakers
and practitioners, especially those concerned with implementing development projects that aim
to minimize prevailing social tensions and conflicts, to enhance the participation and civic skills
of marginalized groups, and/or that are likely to work in ‘moderate’ conflict environments. They
also draw attention to the specific evaluation challenges posed by community-driven development
(CDD) projects such as KDP, with their emphasis on people—i.e., frontline field staff deploying
both professional discretion and engaging in intensive face-to-face interactions over several
months—as crucial mediators between formal program rules and the idiosyncrasies of local
communities. KDP is not a standardized ‘technology’ or product, but a process for reconfiguring
inclusive and accountable decision-making in village life. As such, it requires a correspondingly
innovative package of assessment tools for determining the nature and extent of its impact.
This chapter summarizes the main findings from the study related to five areas: (a) the development-
conflict nexus; (b) ensuring development projects ‘do no harm’; (c) direct conflict management
interventions; (d) indirect interventions to strengthen such capacity; and (e) improving the
assessment and evaluation of community development projects. For each, we sketch out the
broad implications and make a number of recommendations for those involved in developing
and implementing development programs and projects. We then provide specific recommendations
for KDP.
149
2. The Development-Conflict Nexus
Development and conflict go hand in hand. Development programs inevitably shape local conflict
dynamics (in both high and low capacity environments) because they introduce finite resources
into communities and thereby either alter power balances or consolidate the grip of prevailing
elites. Competition over these resources can lead either directly to conflict, or can interact with
existing tensions, thereby causing them to escalate. Programs such as KDP, which aim to reconfigure
inter-group and state-community relations, are also likely to shape local power relations and hence
conflict dynamics. Projects operate in an arena in which power relations are constantly being
worked and reworked. By supporting various groups, and/or particular ways in which local decision-
making should be conducted, they are inevitably political. In order to ensure that projects ‘do no
harm’, it is necessary to assess the potential impacts of the program on peace or violence in the
project preparation stage (and thereafter).
Acknowledging the intrinsic linkages between development and conflict results in a number of
implications for how we conceive of the development process and how we prepare development
projects.
(a) Conceptualizing conflict and development: the political dimensions of development

Conflict is not only something that occurs in ‘conflict-ridden’ places. Rather, it is an inevitable
feature of social life the world over, varying in intensity from fully blown international and civil
war to more local tensions. Processes of contention take place at multiple levels, from national
politics down to the village; this is true even in geographic areas that are not seen as being
particularly prone to conflict. Development projects and policies operate in arenas where power
relations are constantly being (re)negotiated. Indeed, they are a political resource: development
activities (whether they involve building schools or infrastructure, providing micro-credit,
supporting the formation of women’s groups, etc.) constitute external injections of resources and
systems of rules that legitimize particular actions and discourses and thus strengthen particular
individuals, groups or ideas, often at the expense of others. This is true for all development
interventions, including those not specifically targeted at influencing local forms of social or
political organization or conflict. They are thus likely to impact—positively or negatively—on
conflict dynamics. The political bases of development processes and interventions must be
acknowledged.
Recommendations201
Existing conflict dynamics need to be taken into account when planning development
projects. Project preparation work in areas that have experienced (or are prone to) conflict
should include a mapping of local conflict dynamics and political economy. This should
201
Clearly we cannot provide recommendations that apply always and everywhere to all development projects. In this
instance, we seek to outline a general set of recommendations that individual task managers and project staff should
interpret in the light of the specific characteristics and requirements of their projects and the contexts in which they
operate. Many of the recommendations are made with the World Bank in mind, but most should be applicable to other
development organizations (other multi-laterals, bi-laterals, NGOs, etc.), too.
150
include consideration of: key problems/tensions in program areas; salient identity

cleavages, and how boundaries between them are defined and upheld; leadership of groups,
including how leaders are selected and their actions legitimized; local power structures;
sources of and procedures for problem solving and conflict resolution; and dominant
cultural norms.202
Attention must be paid in the design stage to the ways in which projects may interact
with existing tensions. This involves analyzing:
o The distributional impact of development interventions (who benefits, who does not,
and how this maps onto existing power dynamics and cleavages).
o Identifying ways in which projects may be co-opted by particular groups (and designing
strategies to deal with this).
Research and analytical tools should be revised to allow for better consideration of how
the development-conflict nexus plays out in particular contexts. At present, tools such as
the Social Analysis sourcebook focus generally on issues such as livelihoods, resettlement,
and the like.203 These tools can be adapted to allow for a better understanding of the
interplay between development and conflict, including the areas of focus outlined above.
(b) Understanding local context: prioritizing and investing in local knowledge

Project interventions cannot be treated in isolation from the contexts (political, cultural,
institutional, and economic) in which they operate. Projects only have the impacts they do because
of their interaction with phenomena external to the project. What works in one context will not
necessarily work in another. It is partly for this reason that KDP has varying impacts—at times
positive, at other times less so—in different areas. It is thus necessary for development
organizations to understand in detail the contexts in which projects will operate, and to factor
this knowledge into project design, implementation and evaluation. This has implications for
both staffing (short and long term) and project preparation.
Recommendations
More staff should be based in country offices and these staff should spend more time in
the villages in which projects operate than in national capitals. Too often projects are
designed in isolation from the areas in which they will eventually be operational.
Strategic decision-making should also be devolved to the extent possible, to ensure that
decisions are compatible with local realities.
More staff with area-specific expertise, fieldwork abilities, and local language knowledge
should be hired. In institutions like the World Bank, these skills are still in short supply.
Recruiting locally, and keeping staff in particular countries (and regions) for longer
periods, would help build up the local knowledge stock of country teams, and lead to
the development of better projects.
202
We recognize that such social mapping exercises can be quite time and labor intensive (and hence expensive), but
their justification rests largely on the premise that the information derived from them will minimize (even more
costly) problems down the line (i.e., ‘pay now or pay later’). Much of this work can be done locally, but will require
considerable up-front investments in training. (We would also stress that quantitative surveys are just as—if not
more—expensive and time consuming, but seem to be approved almost routinely in development planning.)
203
The Social Analysis Sourcebook is a product of the Social Development Unit of the World Bank and aims to give
guidance for country teams on how to factor such analysis into the development of programs.
151
Invest in analytical work at multiple levels. It is necessary to invest in studies and

assessments in the project preparation phase in order to understand better the local context.
For national programs (like KDP), it is not feasible to do a detailed local mapping in
every area. Comprehensive mapping in selected areas should be accompanied by broader
social research and analysis that helps map the broad institutional environment (and
how this plays out locally), the basic local political economy, and some of the risks. In
this instance, social research is an essential development tool, generating data and
analytical frameworks that can be drawn upon to ensure that projects accurately and
appropriately ‘map’ onto local contexts.
Box 8.1: The Development-Conflict Nexus – Recommendations for KDP
Training
• As part of their initial training, facilitators (at the district, sub-district and village levels) should be given basic instruction
on (a) the general relationship between conflict and development, and (b) their likely role in mediating conflict as part of
the KDP implementation process. Facilitators are not merely interchangeable conduits for transferring resources and
information to villagers, but frontline representatives of a development program overtly seeking to give marginalized
groups a greater stake in decision-making and resource allocation via a competitive bidding process. As such, they
should be fully aware that they are in the conflict management business.
• Specifically, facilitators and program staff should be given enhanced training in how to handle program-related (and
other) conflicts, including problem-solving skills.
• Facilitators should receive a short refresher course once per year.
Monitoring and Support

• KDP should build partnerships with local NGOs or training institutes, who have expertise in conflict management.
These institutions should be contracted to monitor the conflict mediation and problem solving performance of KDP
facilitators, to design/amend training materials based on ongoing findings, and to provide quick ‘booster’ training for
facilitators with particular needs – e.g. those in problematic areas.
Project Preparation
• The continued importance of social analysis in program preparation and ongoing assessment should be emphasized
and improved. This should be done along the lines of the procedures envisioned for SPADA, which emphasize
understanding of local institutions and power relations, as well as poverty incidence.
3. Ensuring Projects ‘Do No Harm’
Development programs inevitably shape local conflict dynamics, not only in areas that have
experienced high levels of violent conflict but elsewhere, too. Our research found numerous
examples of cases where projects had negative impacts. In our research areas, 36 violent conflicts
related to development programming (all but one of them from non-KDP projects) took place
between 2001 and 2003. Comparative case analysis points to a number of reasons why some
development conflicts become violent while others do not, and suggests a number of
recommendations to ensure that development interventions ‘do no harm’ (or at least minimize
their potentially harmful impacts).
(a) Involve communities at every stage

One of the reasons that KDP was less likely to trigger conflicts than other development projects
in our research areas was that it involves communities in a way that other projects do not. In
152
areas where KDP is working well, communities put forward proposals, are involved in deciding
which should go forward from the village, which should then be funded and how the projects
should be implemented. They are also involved in accountability meetings throughout to ensure
that work is proceeding as intended. We found that this constant community involvement helped to
limit tensions relating to the program; individuals and communities were less likely to protest
outcomes, because they felt they had a genuine stake in deciding how these outcomes and decisions
were reached.
Recommendations
All projects should have avenues for beneficiaries to participate in decision-making. This
is particularly true when the project entails providing services or resources that require
information largely unknowable to external agents, and/or where monitoring compliance
with agreements is difficult. This applies not only to ‘social’ and ‘community’ development
projects, but also to projects such as the construction of large-scale infrastructure (roads,
dams) and the provision of technical services (health, education). Beneficiaries need to
be able to define their own priorities, and choose projects, procurement and implementation
strategies in accordance with their values, aspirations, and the idiosyncrasies of their
circumstances. If any of these stages are by-passed, programs are (a) less likely to work
well, and (b) more likely to cause destructive conflict.
Particular groups may be excluded from participation in projects. Careful upfront social
analysis can help identify marginalized groups, so that facilitators can work with them (and/
or those that want to exclude them) to ensure their effective participation.
Information on other development project resources (e.g., from the government, NGOs)
should be made available to everyone, but especially to those groups whose proposals were
unsuccessful in securing KDP funds. Knowing where and how to access alternative sources
of funding can help to lessen the disappointment, and perhaps deep frustration, of those
villagers who spent considerable time and energy preparing a proposal, but nonetheless
‘lost out’. (It can also, potentially, help improve linkages between local communities and
higher levels of government.)
(b) Ensure programs work as intended: socialization, monitoring and follow-up

All forms of development-related conflict are more likely where program processes do not work
as intended. Where program malfunctions occur, the risk of fresh conflict, or of the program
interacting with existing tensions, increases. Where people understand decision-making processes,
they are more likely to accept outcomes, even when these are contentious. Program implementers
have three main tools for ensuring that development programs work as intended. First,
socialization, the process of sharing information about the project and how it is meant to work,
must be comprehensive and ongoing. Where communities understand program rules and
objectives, they are less likely to dispute program outcomes and are more likely to hold local
elites, who may try to co-opt or bend program rules, to account. Second, monitoring the program
is extremely important. This involves not only monitoring program outcomes, but also, more
importantly, processes. Third, following-up on cases of malfeasance, such as corruption, can
send out strong signals that make it less likely that others will try to break program rules.
153
Recommendations
Project participants (actual and potential) need to be fully ‘socialized’ to understand the
processes in and through which the project makes decisions (and what kind of project it is).
Having the most effective decision-making system in place is one thing; ensuring that clients
know how and where to avail themselves of it is another. More and better information (and
fewer information asymmetries between providers and clients) ensures that the ‘rules of the
game’ are clear to everyone, and that outcomes—whatever they may be—have been reached
through coherent, equitable, and legitimate processes. How project decisions are made, by
whom, in what forum, using what criteria, and with what result, need to be as transparent
(and as accessible) as possible. (This is likely to be a very political and highly contentious
process, and should be approached accordingly.)
Socialization must be an ongoing process and a number of different media should be used.
In many cases, there is an opportunity to try innovative approaches. Such approaches may
include the use of community radio, theatre, or other alternative methods. Those that are
grounded in local cultures are more likely to be successful. Experimenting with different
models, and then evaluating what works, is essential.
Socialization should also be a two-way process, allowing information to be fed up the system
as well as downwards. Decision-making systems need to be not only internally coherent
and externally accessible, but amenable to feedback from both providers and users so that
necessary corrections and adaptations can be made on regular basis. Clients, in particular,
need to be able to monitor (at relatively low cost) how well each phase of the system is
working; this is essential for ensuring that both people and processes and held accountable
for the outcomes they (separately and/or collectively) produce.
Regular external monitoring of programs is also necessary. For projects that are large in
scope, this could involve establishing partnerships with local partners, including local
parliaments, NGOs, media and religious groups.204
Where program staff are found to be deliberately breaking program rules, e.g. in cases of
corruption, there should be prompt and thorough follow-up. This goes beyond sacking or
disciplining staff; in serious cases, project resources should be spent ensuring that cases go
through the court system.
(c) Enhance internal complaints mechanisms and avenues of redress

Many conflicts triggered by development programs are the result of program malfunctions and
frustrations with processes, problems that are compounded when there are no or weak avenues
for redress. Where there are clear and accessible mechanisms within development programs
to deal with these concerns early and often, the escalation of problems is much less likely.
Frustrations are also a product of the failure of putative feedback mechanisms to report back
on actions taken.
204
It should not be assumed, of course, that local NGOs, the media, and religious groups are inherently ‘better’ (e.g.,
‘more independent’) than other formal institutions. The point is to identify several credible mechanisms that can
provide an adequate monitoring and accountability role.
154
Recommendation
There is a vital role for complaints mechanisms within all development projects. These
need to be sufficiently independent from parallel program and elite structures. They must
also be:
o Well-advertised, publicizing their responsibilities and that of others to act;
o Accessible to beneficiaries, geographically (i.e., close to where they live), economically
(i.e., affordable), and administratively (i.e., providing clear procedures for engagement);
o Willing and able to take responsibility for program decisions, and to communicate
outcomes to villagers.
(d) Promote greater understanding of the project cycle, focusing on the points within it—
competition for funding, opportunities for ‘elite capture’—at which conflict is most likely
In programs which have a three or four year cycle and yearly phases of competition for and
allocation of funding (such as KDP), intra-village competition for prioritizing funding proposals
is most likely to stimulate tensions as the socialization process begins; this decreases in later
years as villagers have a clearer understanding of competitive processes, and as the aggregate
number of projects being put forward increases over time. Inter-village competition for allocation
of funds to villages, however, is most likely to increase in the last two years of the program,
particularly when certain villages miss out on funding repeatedly as the aggregate number of
proposals being funded increases. In other cases, some tensions arise when particular villages
have been sanctioned for previous poor performance. Given that KDP overtly seeks to establish
more inclusive forms of decision-making, it should be expected (and it was clear from the evidence
in this study) that certain entrenched elites will resist such efforts, thereby heightening the
likelihood of conflict.
Recommendations
Project staff should receive training in how conflict may arise at various points in the
project cycle. This training should aim to give them practical tools to deal with the
different kinds of tensions that may emerge at different stages, including guidelines on
who to turn to if their own immediate managers are implicated as part of the problem.
Those designing programs should not retreat from promoting competition over funding.
This is so even in areas where tensions are high, providing the processes are clear, transparent,
and enforced.
When using competitive mechanisms to allocate resources, the most ‘pro-poor’ components
should focus on securing (and sustaining) access to a fair process of deliberation, not
pre-determining allocations for particular groups. Where one group does not understand
‘the rules of the game’, and hence is at a disadvantage, problems are more likely to emerge.
The absence of fair competition is actually more likely to lead to conflict than competition
per se. Affirmative action initiatives for marginalized groups (e.g., women, widows) should
focus on effective facilitation and ex ante inclusion rather than ex-post resource allocation.
Sustaining the engagement of marginalized groups in competitive processes entails helping
them forge a viable basis of countervailing power; being present in key meetings is a
necessary but insufficient condition for effecting durable, progressive change. Among
155
other things, this means setting (and enforcing) rules and deploying terms of debate that
accord with the everyday experience of marginalized groups.
(e) Where appropriate, include both formal and informal leaders in development programs
Problems were best resolved when redress involved program staff, state representatives, and
informal leaders. State and informal leaders need to be embedded in the process. In conflict
situations, actors involved should be included in ways appropriate to local customs (insofar as
that is possible).
Recommendations
Create spaces, resources, and incentives in development programs for the substantive
involvement of the state, informal leaders and, where appropriate, former combatants in the
planning and implementation of the program. This is important for sustaining both the
‘supply’ of and ‘demand’ for broader transparency and accountability mechanisms. Forging
greater synergies between state actors, ‘civil society’ representatives, and other non-state
leaders should be regarded as an important programmatic objective.
Establish flexibility within the rules system that allows for the creation of ad-hoc
involvement of external parties. In certain cases, it may be more appropriate to involve
only particular parties. Allowing rule flexibility, and creating a financing mechanism, to
permit extra meetings to be held, or temporary working groups to be established, can
provide alternative spaces and forums for project-related problems to be resolved, without
putting too much pressure on normal program procedures and activities.
156
Box 8.2: Ensuring projects ‘Do No Harm’ – Recommendations for KDP
Training
• Facilitators and project staff should be trained in understanding potential flash points in the long-term project cycle.
Discussions of practical responses at different stages in the cycle can draw on the concrete experiences of facilitators.
Socialization
• Pilots to test alternative ways to socialize project procedures and activities should be expanded. In addition to the
current work on using community radio, pilot initiatives using theatre, print media, art and other cultural mediums should
be tried. Many such media already exist; engaging with existing outlets is more effective and efficient than establishing
new mechanisms. As such, this work can be contracted out to local NGOs.
• A common evaluation framework should be used, and the results should lead to the development of a new, comprehensive
communications and socialization strategy for the next iteration of KDP and SPADA.
• A consultant should be hired to gather best practice examples of communication strategies that have worked within
KDP, and this information should be widely publicized within the KDP network (e.g., at trainings).
• In Aceh, information facilitators (FKI) have helped narrow the information gap between the province and local level. The
FKI network should be expanded into other provinces.
Participation
• Extra resources should be employed to increase the intensiveness of socialization and facilitation for minority groups who
often miss out on funding through KDP. This requires the building of ‘countervailing power’ through empowerment activities.
• To ensure such groups take part in KDP, make diversity of participation in meetings part of the performance criteria for
village (FD) and sub-district (FK) facilitators. Emphasize the importance of: gender mix; non-elites; including poor and/
or isolated dusun (infrastructure, incomes, etc).
• Beyond the current lists of marginalized villages, extend this to dusun (sub-village). FK should prepare rankings of
dusun (through the FD), with the list assessed by the district facilitator (KMKab).
• If the proposals of marginalized groups fail to secure funding, continuing efforts should be made to build links between
KDP and other funding sources. This can be done through better coordination with the government’s musrembang
budgetary planning system at the district level.
Handling Complaints and Follow-Up

• The Handling Complaints system could be strengthened. This would include: advertising the visiting schedule of the
KMKab (district consultant) to sub-districts, to increase accessibility; use of local notice boards (and community radio) to
advertise these times (and ensure their regular maintenance); and community meetings (oral communication). Such
tasks can be assigned to local NGOs.
• Encourage greater use of local media for (a) advertising complaints mechanisms, and (b) reporting back on follow-up
measures when an issue is of public concern. In sensitive cases, care must be taken to ensure confidentiality.
• The Regional Management Unit (RMU) should have a budget to reimburse expenses for people who travel to meet
them (at the sub-district level) if they cannot visit the village themselves. This facility should be widely advertised.
• Encourage better follow-up on cases that go through the system. Many cases get stuck in courts; a mechanism should
be developed for pushing these cases forward (e.g., through ‘Justice for the Poor’).
Promoting Creative Problem Solving

• Encourage flexibility and innovation on the part of facilitators with respect to having additional meetings to deal with
problems as they arise. Doing so will require the availability of extra funds that can be used for holding ad-hoc meetings,
or establishing temporary working groups, to deal with particular project-related problems.
• Staff should also be reminded that knowing how (and when) to distinguish between the letter and the spirit of KDP rules
is a key component of solving problems that emerge during KDP implementation. It should be stressed that exercising
discretion is a defining feature of professional life, and that innovation on the part of field staff—not unwavering adherence
to the precise wording of each principle, or inflexibility in the face of novel circumstances—will be needed in order to
ensure that problems (which will inevitably emerge) are addressed effectively. Encouraging discretion (and corresponding
accountability mechanisms) will require providing incentives within the program structure for those who exercise creative
problem solving in their local context. This should be built into the assessment of facilitators.
157
4. Direct Conflict Management Interventions
KDP has little direct effect on aggregate levels of violence in the research areas. It is clear that
in areas that have experienced significant outbursts of violent conflict, programs like KDP
must be complemented by other initiatives aimed at preventing or managing conflict. At the
same time, there is scope to extend the role of KDP to address particular sorts of conflict, most
notably other development-related disputes.
(a) Design complementary programs/strategies to constructively address the drivers of conflict

Increasingly Community-Driven Development (CDD) projects, and other bottom-up approaches,
have been viewed as having the potential to address local conflicts and to increase local conflict
management capacity. The evidence from our study suggests that a more measured conclusion is
necessary. At times project forums and facilitators are used for conflict management; but, more
often, they are not. It is necessary to resist the temptation to believe that a single development
intervention, even (or especially) one with a strong participatory component, can single-handedly
reduce local conflict. At best, effective community development projects can be part of a solution
if they are able to function as effective complements to (and not substitutes for) the state, other
formal civic institutions, and related democratization initiatives.
Recommendations
Program-related forums should not automatically be transformed into more general conflict
resolution forums, as this can undermine their ability to function well within the program.
In particular, facilitators should avoid meddling in conflicts that relate to political or elite
competition at higher levels. This has particular implications in higher conflict areas. Here,
project forums should not be used for addressing large-scale social problems related to the
conflict. Rather, other (parallel) structures should be developed, to allow for bottom-up
input into conflict resolution processes at higher levels of government (and/or society more
generally). Adequate channels for advertising the presence of such measures, and reporting
on their efficacy for addressing conflict, need to be provided.
In areas where building local conflict management capacity is a priority, complementary
approaches to CDD will need to be developed. These should focus on strategies that: (a)
strengthen institutional capacity to deal with conflict; and (b) address the drivers of conflict.
The nature and content of these strategies will depend on the particular conflict and context.
In the Indonesian case, such initiatives will include (but are not limited to):205
o Progressive land (re)distribution and the development of effective land dispute resolution
mechanisms;
o Security sector reform and the strengthening of community policing;
205
For more on approaches to limiting violent conflict in Indonesia, drawing on the fieldwork conducted as part of this
study, see: Barron, Kaiser and Pradhan (2004), Barron, Nathan and Welsh (2005), Barron and Sharpe (2005), Barron,
Smith and Woolcock (2004), Clark (ed.) (2005), Diprose (forthcoming), and Satu and Barron (forthcoming). These
papers and others on conflict in Indonesia are available on-line at: www.conflictanddevelopment.org
158
o Strengthening of the judicial system and linkages with alternative dispute resolution
(ADR) procedures;
o Strengthening of the decentralization processes, particularly budgetary and decision-
making transparency and accountability;
o Employment generation schemes, in particular aimed at young men.
(b) There are opportunities for facilitators and forums to play a larger role in some conflicts,
particularly those related to development
KDP facilitators infrequently deal with non-KDP conflicts and KDP forums are rarely used for this
purpose. This is largely because for most forms of conflict, they are not seen as the appropriate
actors/arenas, with other systems of recourse being in place. However, for conflicts related to other
development projects, and in cases related to other forms of resource distribution, there is room for
KDP to play a larger role. Most development projects do not have adequate complaints systems or
means of redress for beneficiaries who feel that processes or outcomes have been unfair. In such
cases, KDP is a suitable mechanism for directly addressing conflict.
Recommendations
Training should be given to facilitators about how to effectively intervene in other

development-related conflicts. Clear guidelines should be given on when to intervene and
when not to, as well as local context-relevant approaches to solve particular problems.
Discernment is vital: intervention in non-project conflicts can be useful, but sometimes
projects work well precisely because they are perceived as being apolitical. Program
facilitators need to balance carefully their responsibilities to the beneficiaries and maintaining
the integrity of the program, with the potential role they can play as independent mediators
of conflicts that might otherwise turn violent.
It is necessary to have strong vertical and horizontal support mechanisms for facilitators
who are directly engaging in conflict mediation or who are working in high conflict (and/
or high corruption) areas. This should include ways to bring in outside help if the
situation worsens.
159
Box 8.3: Direct Conflict Management Interventions – Recommendations for KDP
Defining the Limits of When KDP Should Get Involved

• Be wary of necessarily expecting KDP (or SPADA) facilitators to be primary mediators in non-program conflicts. It is
important to clearly define the kinds of conflicts and situations in which facilitators and program staff should get involved,
and when they should not.
• Guidelines should be given on what forms of conflict facilitators/forums may address, with emphasis on local flexibility.
• Training on problem solving methods relevant for particular types of problems in different local contexts should be
provided. This could include using examples from previous research in the training to:
o Demonstrate how conflicts escalate if they ignore problems;
o Identify local mechanisms that work in limiting violent conflict;
o Learn from other regions;
o Discuss ways to improve local conflict management capacity.
• Developing a core curriculum on conflict analysis and mediation is of utmost importance for both KDP and SPADA.
Ideally, the institution contracted to develop this should also conduct ongoing research activities to update materials and
to ensure they are relevant to local contexts.
Building Networks
• There is a need for linkages to other mechanisms to deal with problems that arise in the forums but which cannot be
effectively addressed, or in which solutions cannot be enforced. This may include both state-based and alternative
dispute resolution mechanisms.
• Often KDP facilitators can provide support by plugging disputes (and disputants) into parallel conflict resolution systems.
Encourage facilitators to work with and actively support other mediators, including informal leaders, and have program
facilitator presence on peace forums that exist in higher conflict areas to ensure coordination and learning.
• Use the initial social analysis process to procure information on the landscape of formal and ‘alternative’ dispute resolution
procedures in the local area.
• Improve coordination between KDP and other development programs (governmental and non-governmental), as well
as any conflict prevention and/or resolution initiatives.
• Keep KDP facilitators informed with respect to relevant decisions made and regulations passed at district level.
Creating New Mechanisms and Spaces

• If KDP is not the correct space, and there do not appear to be other mechanisms available, KDP facilitators should
encourage the creation of other ad-hoc forums/musyawarah to deal with problems.
• Discretionary funds should be made available for ad-hoc non-violent conflict initiatives (e.g., internal and external to KDP).
Providing Support
• There is a need for improved vertical linkages, whereby problems can go up through the KDP structure until they find
effective resolution.
• The successful management of local conflict, especially those stemming from KDP itself, is a function not only of the
skills and commitments of individual frontline staff, but the support they receive from their superiors. In particular,
facilitators in isolated and/or poorly governed areas need to know that they are not alone in their efforts to implement
KDP and to address the conflicts it invariably generates. Local and district staff are on the front lines in conflict situations;
as such, these staff should have a clear strategy in place for supporting facilitators (FKs and FDs) as and when they
encounter local conflict (for ensuring both their physical safety and effective project implementation).
• Reporting structures must be strong and responsive; early warning information systems should also be designed to help
when things go wrong.
160
5. The Indirect Impacts of Development Projects on Conflict Management
(a) Development programs can indirectly have a positive impact on local conflict and conflict
management capacity
Projects such as KDP can have important impacts on and interactions with processes of social,
political and economic change. Often the result of such interactions may be as important and
significant as the direct impacts of the project. KDP is leading indirectly to changes in social
structures, institutions and norms that appear to be making communities more robust to conflict in
the medium- to longer-term. There are opportunities to build upon this by thinking more strategically
about how KDP (and other programs) can positively affect these variables in conflict situations.
For long-term conflict prevention and management, equal value should be placed on the impact on
these intermediate variables as on the desired outcome variable of ‘reduced conflict’.
Recommendation
Development organizations and governments should think about indirect pathways to

conflict reduction as well as direct. It is important to take into account (and document)
impacts on a range of variables that, over time, relate to conflict outcomes. There is a
need to consider how programs impact on social structure (between groups, and between
groups and the state). Programs should give specific attention to the ways and means by
which they—intentionally or otherwise—include certain groups and exclude others, and
how the dynamics between these groups are likely to change over the course of the program.
‘Conflict’ programs that aim to prevent or directly manage conflict should pay more
attention to the local political economies in which they are implemented.
(b) Improving inter-group relations: the impact of KDP on social structure

Conflict is less likely to ensue, and is more likely to be addressed early when it does emerge,
where the social fabric of village life is characterized by diverse ties between different identity
(and ‘stakeholder’) groups (Varshney 2002). This hypothesis holds up at the micro-level in the
context of the Indonesian villages in our sample. On-going participation by diverse groups in
civic spaces (such as village councils, recreational gatherings, business associations, water-users
groups) can serve to nurture, expand, and reward such ties. Those designing projects—both
those explicitly aiming to address conflict, and those with other development goals—should
think about ways in which project designs create or limit incentives for inter-group participation
and the development of sustainable inter-group ties.
Recommendations
Projects that expand the number of groups with a clear stake in the outcome of
decision-making are more likely to be supported, implemented, and maintained.
Projects should think about innovative ways to promote and sustain the involvement of
different groups, especially so when such involvement meets resistance (as it is likely to)
from local elites.
The spaces that participatory development projects create for inter-group discussion are
often as important as the more tangible outputs they produce. In environments where
161
there are few opportunities for inter-group mingling, development projects may be a suitable
medium for creating such spaces.
(c) Normative and behavioral changes

KDP has an impact on people’s participation, which is spilling over into demands to participate
in other development programs; this, in turn, helps to reform local governance more generally.
This process is occurring to a greater extent in areas where democratization processes are already
underway; in such instances, the program and other forces of social and political change reinforce
each other. Even where democratization is not occurring to the same extent, there appears to be
program effects, although these occur more slowly. Further, when well implemented, KDP can
improve not only the diversity and equity of decision-making spaces, and the civic skills of
participants, but alter the normative and cultural expectations regarding procedures for resource
allocation and conflict management. This is an important accomplishment in environments
heretofore characterized by pervasive elite capture (including corruption), norms of violent
retribution, and the systematic exclusion of marginalized groups.
Recommendations
When well implemented, participatory projects can effectively stimulate demand for local
democratization and improvements in governance. They can provide powerful examples
of how democratic decision-making can take place; and these precedents can create demand
for broader democratization that spills over into other avenues of local political life. Further,
projects can provide examples of alternative non-violent means for solving problems;
precedents have powerful ‘path dependence’, and can be harnessed for constructive (and
more progressive) ends.
Senior project staff should give consideration as to how programs interact with processes
of social and political change (especially at the district level). The spillover effects of
strategies to create demand through projects will be strongest if they are accompanied by
coherent and complementary strategies—at the local, regional, and national level—to
improve the ‘supply’ of good governance.
162
Box 8.4: Indirect Impacts of Development Projects on KDP – Recommendations for KDP
Positive Precedents
• Upholding KDP principles is especially important for the early establishment of ‘positive precedents’. Such precedents
should be publicized through the KDP network and used as a basis for training. Likewise, ‘negative precedents’, such
as cases where money is corrupted, can ruin the effectiveness of the program. These should be punished, and the
outcome of measures taken to correct it widely disseminated to communities.
Improving Social Ties

• The inclusion of marginalized groups is not merely an administrative requirement of KDP, but a key principle and first
step towards it helping to improve the fabric of social life at a local level. In areas where identity cleavages exist at a
higher level, mechanisms should be designed to bring people together people at higher levels of geographic specification.
This is planned for the SPADA project. However, consideration should be given to extending the KDP structure vertically
(through district-level forums), in areas where SPADA will not work.
Linking to Supply-Side Approaches

• KDP appears to be an effective way to stimulate demand for democratization and good governance, factors that in the
long-run will make communities more robust to conflict (though conflicts may arise as a result in the short-run). Attention
should now be paid to linking this ‘demand’ approach to strategies that improve the ‘supply’ of good governance. This
will require improving the linkages between KDP and the state at the district level. It will also require better coordination
between the KDP team and those within the World Bank working on the governance agenda.
External Socialization
• Indirect approaches to promoting conflict management and institutional reform are still relatively rare in the development
community or amongst developing world countries. The experience (positives and negatives) of the approach followed
by the team in Indonesia should be ‘socialized’ externally—e.g., to the World Bank outside of Indonesia, to other
development actors, and to the Government of Indonesia.
6. Implications for Evaluating Community Development Projects
Assessing the efficacy of complex social development projects to negotiate new procedures for
decision-making in village life requires a diverse set of methodological and analytical tools. Project
evaluations traditionally measure progress against a set number of outcome variables, specified
before the project begins. While it is important to strive for clearly defined metrics of project
success (e.g., Alatas [2005] on the economic rate of return achieved by projects delivered through
community mechanisms versus those provided by external contractors), the very nature of projects
such as KDP and the key social objectives (participation, empowerment) towards which they
strive belie stringent assessment by a single research tool, or of a set of preordained outcome
variables. In this context, the orthodox (economic) criteria for what constitutes a ‘rigorous’
project assessment need to be considerably expanded.
(a) Studying the interaction between projects and ongoing processes of change
A major conclusion of this study has been that there is no such thing as a ‘pure’ project impact.
Development programs do not have impacts on their own; rather, their impacts arise from the
ways in which the resources and rule systems they introduce interact with extant phenomena
such as the interests and incentives of actors on the ground, the forms of social structure present,
163
and so on.206 It is thus necessary to evaluate how the program interacts with elements of the
context. The complexity of this is increased by the fact that contexts themselves are not static,
being driven by processes of social, political and economic change that, in turn, also powerfully
shape project outcomes.
The approach above suggests that a fundamentally different—and in many ways, more difficult—
method of program evaluation is called for in assessing participatory projects (such as KDP).
While it is important to determine a net program ‘effect’ (in econometric terms, an effect when
all other variables are held constant), it is also the case that such impacts are simultaneously and
inseparably stem, at least in part, from other elements of the context. Documenting the nature
and extent of these interaction effects can only imperfectly (at best) be derived by formal
quantitative methods; as such, vital empirical space is opened for incorporating qualitative methods
and the unique insights they generate pertaining to processes and mechanisms. Moreover, it is
perhaps equally (if not more) important for program designers to understand why a given project
had the effect it did, or, more precisely, how exactly it contributed to a given outcome, than to
empirically disentangle the differential impacts stemming ‘independently’ from project, selection,
and context effects. It is this knowledge that is most important when thinking about whether a
program can be scaled-up, whether it should be replicated elsewhere, whether design changes
are necessary, etc.207
Recommendations
Projects cannot be evaluated in isolation from the contexts in which they operate. Because
these contexts vary, program evaluations should look more at variations, trying to understand
what caused different outcomes in different places, than at trying to establish a pure and
uniform program impact.
Evaluations should seek to tease out the implicit hypotheses in project designs, and then
seek to test these. Understanding how these ‘chains of effect’, or sequences of mechanisms,
actually work is perhaps more important to future program design considerations than trying
to discern a ‘pure’ (net) project impact. Put another way, establishing how and why a
development project worked—most likely in different ways to different degrees in different
places—is likely to be more instructive (certainly to front-line project managers) than deriving
an aggregate measure of whether it worked.
(b) Use mixed methods approaches

The difficulty and complexity of conducting useful and comprehensive program evaluations should
not be used as an excuse to limit evaluation activities. Rather, multi-faceted and mixed methods
206
Our thanks to David Mosse for stressing this point in his review of this study. He explores this argument in much
greater depth in Mosse (2005).
207
One way to understand why programs like KDP are successful (or not) in achieving desired outcomes is to treat them
as ‘laboratories’ though which processes of social change can be studied. That is, projects can be treated as sets of
hypotheses about what works in certain contexts. These hypotheses can then be tested, in part by finding appropriate
proxy intermediate and dependant variables, in part by deconstructing them and inductively trying to work out why
social phenomena occur using case study and ethnographic methods. (Our thanks to Scott Guggenheim and Lant
Pritchett for many interesting conversations on this point.)
164
approaches can provide important insights into the strengths and limitations of development
interventions, and why they are successful or not; in so doing, they can usefully contribute to the
design of better, more effective projects and policies.
Recommendations
Encourage monitoring and evaluation from the outset; resist the (strong) pressures to ignore
or short-change evaluation opportunities. One obvious limitation of this study is the lack of
comprehensive baseline data to allow for before and after comparisons. To the extent possible,
program managers should invest in baseline and longitudinal studies to allow for more precise
evaluations of project impact.
Adopt methods and procedures for assessing impact that comport with the project’s distinctive
design characteristics and context idiosyncrasies. As noted earlier, the nature of the problem
being studied should drive the choice of instruments, and vice versa.
Box 8.5: Implications for Evaluating Community Development Projects – Recommendations for KDP
• Continue to encourage a diverse range of strategies for assessing the various aspects of KDP’s design, implementation
protocols, and impact.
• Continue to be a standard-bearer for disciplining debate regarding the efficacy of CDD projects generally, and KDP in
particular—whether excessively supportive or critical—to the evidence, even as it highlights the circumscribed role that
“evidence” actually plays in the political economy of decision-making pertaining to the design, placement, implementation,
and efficacy (perceived or actual) of development projects.
7. Concluding Thoughts and Future Directions
Growing awareness of the pervasiveness and consequences—local, regional, and global—of

conflict in developing countries creates strong imperatives to identify programmatic solutions,
and to expand or replicate putative ‘success stories’ and ‘best practices’. Through the high national
and international profile that KDP now (rightly) enjoys, the pressures to portray it—and CDD
projects more generally—as a solution to conflict are considerable. The results of this study
suggest a more measured response. At an aggregate level, we find little evidence that KDP, in
and of itself, reduces levels of violent conflict. Rather, at the community level (at which it
operates), we find that KDP, unlike other development projects, is largely successful at resolving
the conflicts it necessarily generates, and that its indirect impacts—on group relations, civic
skills, and norms of cooperation—while significant, are heavily dependent on how well the project
is implemented, and the characteristics of the broader institutional context (specifically, the
capacity of local governance mechanisms).
In short, KDP is not, and CDD projects more generally are not, a silver bullet solution to local
conflict. At worst, they can inflame existing tensions and thereby become part of ‘the problem’;
at best, they can be part of a solution when they work to complement ongoing governance reforms.
As such, projects such as KDP should be regarded as components of, not substitutes for, coherent
development strategies for enhancing local conflict mediation capacities and improving the
transparency, accountability, and inclusiveness of decision-making mechanisms.
165
Much remains to be done to explore the ways and means by which development projects shape
the dynamics of social and political change at the local level. As a basic first requirement for
project design, our findings highlight the central importance of having accessible and effective
feedback mechanisms in place, both for improving project performance and for addressing the
distributional conflicts that even (or especially) successful projects generate. As is true elsewhere
in life, prevention is far better than cure. A second key issue for both project designers and
evaluators is the challenge of taking seriously the impact of ‘context’. Much lip service is given
to such development bromides as “one size doesn’t fit all”, but it is much rarer to see explicit
attention given to crafting projects that are at once large enough to achieve broad impact yet also
able to adapt themselves to the idiosyncrasies of particular settings.
Moreover, the prevailing imperatives of (large) development agencies for neat metrics of project
performance, and the corresponding skills sets they tend to require and reward, makes it difficult to
assess the efficacy of CDD projects like KDP, whose mechanisms (‘participation’, ‘inclusion’) and
outcomes (‘empowerment’, ‘capacity for collective action’) are inherently hard to define and measure.
Taking context seriously should mean that those perspectives and methodologies best suited to
engaging with and responding to it are given their due.
Indonesia is currently undergoing a momentous ‘triple transition’ in its political, social, and
economic life. Power relations are in flux, identities are being renegotiated, and institutions are
transforming. Changes in incentives, and in the role of formal and informal institutions at various
levels, have altered the ways in which individuals and groups relate to each other, and the nature
and extent of their relationship to the state. This is the contested and uncertain arena into which
development programs such as KDP, which aim to promote progressive social change and more
accountable government, enter. Change is invariably slow and uneven, and the mechanisms
through which it occurs are diffuse and complex. Yet our research found that KDP does contribute
in important ways towards making local democracy work. Significant lessons can be learned
from the program regarding ways to design and promote constructive development transitions,
in Indonesia and beyond.
166
ANNEXES
Annex A: Details of Research Areas
1. East Java
Throughout history, East Java has had extensive political, strategic, administrative and economic
influence in Indonesian society. Its wealth and proximity to the center of national government on
the island of Java have ensured that it has one of the highest levels of development and infrastructure
(albeit unequally distributed) in the country as well as considerable political influence. The transition
between the kingdoms which dominated from the tenth to thirteenth centuries (the Kediri, Singosari,
and Majapahit kingdoms) ultimately saw Islam dominate parts of the province and provide religious
and political links between East Java and many of the surrounding islands.
When Islam began to be integrated into parts of East Javanese society during the time of Majapahit,
tensions between the rituals of Javanese mysticism and the practices of devout Islam emerged. The
north-eastern part of the province, known as the horse-shoe area (daerah tapal kuda),208 eventually
embraced Islam, and provides one of the bases of support for the national Muslim organizations
and parties. One such group associated with the region is NU (Nahdlatul Ulama), which represents
40 million Indonesians nationwide and has political links with PKB (the National Awakening Party),
the dominant political party in the province (Hefner 2000). NU is a massive sprawling organization
with a network of endogenous and exogenous religious, educational, small and medium enterprises,
trade unions and other bodies. Nonetheless, traditional Javanese practices still dominate other
parts of the province, in particular the more secular nationalist areas in the south-west. People in
these areas, however, still identify themselves as Muslims, and, in the words of several respondents,
as “Identity Card (KTP) Muslims”, they also continue to believe in the spirit life, carrying out
rituals to worship their ancestors and celebrate key days in the Javanese calendar.
1.1 Research Districts and Sub-Districts

The qualitative research and key informant survey was conducted in eight sub-districts in two East
Javanese districts. The two districts are at opposite ends of East Java: Ponorogo is in the south-
west of the province, bordering Central Java; Pamekasan is on the island of Madura, off the north
coast of the Java mainland (Maps A1 and A2). While poverty rates are similar between districts
(Table A1), the districts vary in many other ways.
208
The regions belonging to the horse-shoe area include all of the districts on the island of Madura (Bangkalan,
Sampang, Pamekasan, and Sumenep), as well as Banyuwangi, Situbondo, Pasuruan, Jember, Probolinggo, Lumajang,
and Bondowoso.
167
Table A1: East Java Research Sub-Districts

(Population and Poverty)209
Population Poverty Rate
Pamekasan 696,932 0.41
Palengaan 65,790 0.54
Proppo 67,619 0.64
Pasaen 48,619 0.38
Pademawu 66,867 0.24
Ponorogo 842,211 0.44
Sampung 39,888 0.52
Badegan 30,307 0.64
Jenangan 54,094 0.33
Slahung 53,080 0.47
Pamekasan
Pamekasan, on the island of Madura, is a district located in the horse-shoe area of the province,
where Islam forms a part of daily life and where political affiliations with Islamic parties, religious
leaders, and Islamic boarding schools are the norm. A major religious identity division separates
the santre (santri, or the pious) from the banne’ santri (non-santri, or non-pious). Community
members educated in Islamic boarding schools and/or with strong allegiances to local Muslim
clerics are defined as santri and are granted elite standing by the community. Santri often go on to
occupy both formal and informal leadership positions in the village and the wider region, both in
government and non-government institutions. Other cultural groupings exist along class lines,
with peasants/commoners, bureaucrats, and people of noble descent being distinguished by the
different levels of language used in everyday conversation.
The island’s soil surface is dominated by limestone and lime precipitate; despite plentiful rainfall,
such geological factors make it difficult for the soil to absorb and store water, leaving a harsh, dry
and infertile terrain (Kuntowijoyo 2002). Such an environment is conducive to growing tobacco,
the mainstay of Madurese agriculture and the primary source of income. However, water is fiercely
sought after by island’s peasant farmers (accounting for 70-80% of the population) to sustain their
livelihood (De Jonge 1998). Such conditions also affect the livelihoods of other occupation groups
in the region such as the fishermen and traders, whose sales also depend on a successful agricultural
harvest. Hence, when water is in scarce supply during the dry season, the informants argued that
tempers rise and conflict becomes more prevalent. The harsh agricultural conditions give rise to
the stereotype of the Madurese as harsh, hot-blooded, and frank individuals who adhere to cultures
of violence for problem solving.
Legends from Madura indicate that violence can be traced back to the year 929 AD when the island
was founded.210 In addition, there is a long history of nyikep (the custom of carrying weapons),
209
The following data sources are used in this Annex. The poverty data is from the SMERU poverty map (2004), which
uses data from the 1999 Susenas and Podes surveys. The population data at the district and sub-district level is that
reported by the local offices of the Bureau of Statistics (BPS 2002a, 2002b). The population data at the village level is
from the 2000 census (SMERU 2004).
210
For example the legend of Prince Sagoro describes his struggle to defeat two giant snakes coming from the ocean.
With the aid of the Muslim cleric, Kyai Poleng, Prince Sagoro finally succeeded, and the snakes turned into a pair of
spears. Prince Sagoro is considered to be the first occupant of the island of Madura and the pair of spears to be
Madura’s first weapons.
168
Annexes
which many of the respondents argued was inherited from their Madurese ancestors, and is necessary
for their daily agricultural practices. De Jonge (1998) also writes in detail of the violent culture
which ensued following the rebellion of the Madurese and their separation from the Mataram kingdom
during the Dutch colonial era around 1700.211 When repressed and poor citizens no longer saw the
institutions in the kingdom as legitimate centers of power, crime rates spiked and as a result the
community began to develop means of problem solving which did not involve government
interference (Wiyata 2002, pp.67-68). The most well known custom of using violence for problem
solving is carok.
Carok is the custom of defending one’s honor and seeking justice through a sickle duel, which can
occur between individuals, as well as between both small and large groups. The duel will more
often than not be carried through to the death, even if this contravenes state law and local regulations.
Wiyata (2002) relates the development of the custom back to the infertile and harsh environment
endured by the Madurese and the history of feudal politics outlined above. However, over time the
custom of carok has become intertwined with changing local values. With religious beliefs
dominating the social landscape, defending slurs against Islam has more recently become one of
the key causal factors for engaging in carok. This was encountered several times during the study.
Similarly, the patriarchal nature of Madurese society together with the local teachings of Islam,
have resulted in women being considered a part of ‘man’s being’. Hence, slurs against the name of
a wife, sister, or loved one, as well as (perceived) extra-marital affairs, are considered a slur against
the name of the husband, brother or father and worthy of honor killings.
The custom of carok has also been used to defend property, occurring as a result of heated arguments
between men, and has been used as an intentional and calculated means of political and economic
gain. While the respondents argued that the custom is dying out in the more urban areas in Madura,
in part because of education and modernization,212 Madurese values instilled from birth continue to
promote a sense of courage in the Madurese people, best captured by the local expression “ango’a
pote tolang etembang pote mata” (better white bones than white eyes). This saying expresses the
value that it is better to die than to live in fear, haunted by feelings of shame.
The Madurese culture in general differs from the Javanese culture which dominates the majority of
the East Java province. While Javanese farmers cultivate wet fields, Madurese farmers use a variety
of different local practices to cultivate dry fields. Such agricultural practices and ways of living
have led to a specific housing pattern which clusters houses together across villages, often along
kinship lines. Such a settlement pattern builds strong internal family solidarity, with insults to
individuals regarded as insults to the whole extended family. However, at the same time, this
211
The Madura kings were given political authority and rights by the Dutch to arrange their own government in all areas
but foreign affairs. The kings used forced labor and collected taxes in the form of crops or money. These were used to
support the needs of the royal elite, bringing them unearned wealth. The Madurese people, in contrast, who already
lived humble lives due to the geographic conditions, became poorer and dispossessed, with many of them migrating to
Java to seek employment. Such migration patterns continue today (De Jonge 1998, pp. 22-23).
212
Police data in Pamekasan demonstrates that there were 202 carok incidents in five years (1990–1994). Of these 202
cases, 40.6% participants were uneducated or had never attended school, 53.1% had SD (primary school) education,
and 6.3% had SLTP (Junior High School) education. Not one carok actor in this data had studied at an Islamic boarding
school (Wiyata 2002, p. 51).
169
residential pattern weakens the sense of community across the whole village, particularly in more
rural isolated villages which do not frequently engage with newcomers, traders, or people from
other parts of the village (Wiyata 2002).
Socially, different groups of people have come to occupy each of three large regions in Pamekasan.
The first is the area in the North where many peasant farmers, fisherman and traders live.213 Second,
the Muslim clerics (Kyai) and their boarding schools of Koranic studies (pondok pesantren) occupy
the areas in the west of the district, where the community possesses higher levels of religious
education.214 The final region is the area in the south-west where many of the members of the
government administration and parliament live. Many members of the elite groups in the community
from the aristocratic classes (priyayi) live in this region.215 This historical and geographic division
of the community has implications for the social structure of the community. For example, the
regions in the north and to the west of the district have higher levels of crime than the areas to the
south east where the elite resides. Furthermore, crime networks known as bajingan216 dominate the
north and parts of the religious areas of Pamekasan.
Map A1: Location of Pamekasan Research Sub-Districts
213
Sub-districts Batu Marmar, Pasean, Waru, and parts of Pegantenan and Pakong.
214
Sub-districts Propo, Palengaan, and parts of Pegantenan and Pamekasan.
215
Sub-districts Pademawu, Tlanakan, Pamekasan, and parts of Galis and Larangan.
216
Bajingan are people involved in organized crime networks, local mafia-like figures. This term is usually translated
as ‘thug’ or something similar, but in Madura bajingan can be an informal security mechanism and have leadership and
problem solving qualities. Bajingan is used here in the singular as well as the plural.
170
Annexes
Ponorogo
Ponorogo is situated at the far end of the province, bordering Central Java. In contrast to
Madura, the dominant ethnic group is Javanese. In terms of key social identity groups in the
current era, the priyayi tend to be associated with the bureaucrats and gatekeepers of Javanese
morality in Ponorogo, the abangan are loosely associated with the grassroots non-Muslims (or
non-devout Muslims), and the santri are the devout followers of Islam.217 Of course, both
Geertz and other writers recognize that these groupings are not definitive, with many shades of
grey in between.218
Similar to historical developments in Pamekasan, the introduction of Islam in Ponorogo was not
a smooth transition, facing challenges from both the grassroots and local leaders. In protest at
the growing dominance of Islam, one local leader, Ki Ageng Kutu, exiled himself to Kutu village
and founded the Surukebeng martial arts school (the so-called “padepokan”). He consequently
gathered together the youth in his padepokan to create a support base, training them in martial-
arts techniques, especially in wrestling, stick fighting, and sword fighting. Training was carried
out to the gamelan (Javanese orchestra) music, a tradition that continues today.
Performances of these martial arts are encapsulated in reyog (traditional Javanese dance)
performances, which were used as a means of public satire against the infiltration of Islam in the
region. By the time of the Demak kingdom in the 16th century, reyog embodied a synthesis
between Islamic teachings and traditional Javanese spiritual teachings (kejawen), being allowed
to remain in cultural consciousness as long as it was altered to encompass the teachings of Islam.219
Reyog performances continue to be very popular in Ponorogo today, being organized to celebrate
special days on the Javanese calendar—births, weddings, and other ceremonies—and as a form
of popular entertainment scheduled throughout the year, with local citizens, young and old,
attending en masse.
Furthermore, martial-arts groups and practices (known as silat) continue to dominate the region.
There are two dominant martial arts groups with their headquarters in the neighboring Madiun
district: SH Terate (the Terate Setia Hati220 Martial Arts Organization), and SH Winongo (the
Winongo Setia Hati Martial Arts Organization). The two groups stemmed from the same
organization, Setia Hati, which was formed in Surabaya in 1903.221 With tens of thousands of
members and a hierarchical organizational structure, these groups are a key identity marker for
217
Geertz argues that Javanese society is made up of three categories: abangan, santri, and priyayi. Priyayi are the
Javanese elite connected with the ancient Javanese aristocracy, who adhere to the practices of the traditional Hindu and
Javanese faith, and who use high Javanese language. The santri are members of the community originally associated
with sea-faring merchants, who adhere to the teachings of Islam. Finally, the abangan are the Javanese peasant farmers
who have a diverse spiritual faith, often continuing to practice their traditional rituals as well as those encouraged by
Islam (see Geertz 1987, 1992).
218
See, for example, Fauzanafi (2002).
219
When Reyog was first created, it was performed using a tiger mask, called singabarong, with peacock feathers on
top. The mask was a symbol of the reign of King Brawijaya V (Majapahit) who was dominated by his wife who
encouraged him to embrace Islam. Male dancers with horses were dressed like women, mocking the Majapahit armed
forces and portraying them as a bunch of sick chickens. In contrast, Pujang Ganong, a symbol of Ki Ageng Kutu,
danced freely ridiculing Singobarong and defeated him (Fauzanafi 2002, pp. 64-67).
220
Setia Hati means ‘loyal’.
221
From “A Compilation of Materials from the SH Terate Martial Arts School, Distributed by the Ponorogo Manunggal
Terate Cooperative”, p. 4 (see Probo 2003b).
171
male (and sometimes female) youth in Ponorogo, and encounters between them—often at reyog
and other cultural performances—can have violent consequences.
The Ponorogo region is geographically divided into two sub-areas: highlands and lowlands.222 It
has 14 rivers which irrigate the rice growing areas in the lowlands, whereas the mountainous
areas are used for dry-land agricultural activities (BPS 2002c). The majority of people work as
farmers, farmhands, and laborers, although in the more urban areas employment in government
and in service provision is more common.223 The use of stereotypes is prevalent in the rural local
communities, being signified by the Sekayu River. The wong wetan kali community (the
community living to the east of the river) view the wong kulon kali community (the community
living to the west of the river) as rough, lawless and kejawen (Javanese people who mix cultural
practices with that of Islam).224 Meanwhile, the wong kulon kali community regards the wong
wetan kali community as newcomers who do not fully understand Ponorogo culture.225 The
kejawan and reyog groups are closely associated, with the district government at present
referring to the reyog community as having the personality of thieves. 226 However, for those
involved in reyog, they are traditional heroes who walk the fine line between being warriors and
outlaws, with the warok (the Ponorogo experts on martial arts and mysticism) as their leaders
(Simatupang 2002, p. 40).
The political map of Ponorogo can also be considered from this cultural viewpoint. The current
Bupati (District Head), Markum Singodimedjo, is from the cadre of the Golkar party.227 During
the 2000 election race, he lobbied the reyog leaders to pressure PDI-P’s preferred candidate to
quit the election race. In addition, Markum lobbied the Islamic leaders in the community and
other reyog leaders to pressure the “Islamic” and PDI-P party members of the parliament to
support his candidature. This is one example of how cultural knowledge can be utilized for
political gain in the region.
222
The highlands include Kecamatan Ngrayun, Sooko, Pulung, and Ngebel. The lowlands include Kecamatan Slahung,
Bungkal, Sambit, Sawoo, Mlarak, Siman, Jetis, Balong, Kauman, Jambon, Badegan, Sampung, Sukorejo, Ponorogo,
Babadan, and Jenangan.
223
In 2002, Ponorogo also sent the highest number of female migrant workers overseas from East Java.
224
Another term used to describe the kejawen in both anthropological studies and by the community is abangan.
225
This is not a strict dichotomy because in reality members of each community may possess the personality described
of the other community.
226
The reyog community is now a part of the Insan Taqwq Ilahi (INTI), an informal organization formed by Golkar (the
New Order government party) in the 1980s.
227
According to the analysis of many respondents, he should not have been chosen in the year 2000 because the district
legislature who appointed him was politically dominated by the Indonesian Democratic Party of Struggle (PDI-P).
172
Annexes
Map A2: Location of Ponorogo Research Sub-Districts
1.2 Research Villages

Within East Java, research was conducted in 21 villages (Table A2). The villages were not selected
to be representative of the districts or sub-districts in which they lay. Rather, villages were chosen
based on the location of ‘interesting’ conflict cases identified in the qualitative research, using
detailed criteria aimed at making it easier to isolate program impact (Barron, Diprose, Madden,
Smith and Woolcock 2004). As can be seen from Table A2, the villages vary extensively in terms
of the years that they have had KDP, their population and their poverty rates.
Table A2: East Java Research Villages
No. District Sub-district Village Phase Treatment? Years KDP* Population Poverty Rate
1 Pamekasan Palengaan Banyupelle 2A Control 0 6,423 0.12
2 Pamekasan Palengaan Palengaan Laok 2A Control 0 7,579 0.54
3 Pamekasan Palengaan Palengaan Daja 2A Control 0 7,245 0.38
4 Pamekasan Proppo Panagguan 2B Treatment 2 4,121 0.66
5 Pamekasan Proppo Proppo 2B Treatment 2 1,669 0.46
6 Pamekasan Proppo Maper 2B Treatment 2** 1,734 0.46
7 Pamekasan Proppo Tattangoh 2B Treatment 2 1,722 0.68
8 Pamekasan Pasaen Sana Daya 3 Treatment 3 10,705 0.22
9 Pamekasan Pasaen Tlontohraja 3 Treatment 3 5,079 0.65
10 Pamekasan Pademawu Padellegan 3 Treatment 2 2,861 0.26
11 Pamekasan Pademawu Sumedangan 3 Treatment 2 2,846 0.22
12 Ponorogo Sampung Sampung 2A Control 1 4969 0.47
13 Ponorogo Sampung Gelangkulon 2A Control 1 3469 0.51
14 Ponorogo Sampung Pagerukir 2A Control 1 2125 0.77
15 Ponorogo Badegan Badegan 2B Treatment 2 2642 0.41
16 Ponorogo Badegan Dayakan 2B Treatment 2 3409 0.75
17 Ponorogo Badegan Biting 2B Treatment 2 1884 0.67
18 Ponorogo Jenangan Panjeng 3 Treatment 2 1604 0.28
19 Ponorogo Jenangan Kemiri 3 Treatment 2 3660 0.46
20 Ponorogo Slahung Slahung 3 Treatment 3 7626 0.46
21 Ponorogo Slahung Wates 3 Treatment 2 4108 0.63
* at the time key informant survey was implemented
** key informant survey not implemented
173
2. NTT
Nusa Tenggara Timur (NTT) province differs in almost every respect from East Java. It is one of
the poorest provinces, with a regional GDP per capita of only Rp. 2,201,100, or just over USD 200
(BPS 2002b). Seventy percent of people make their living from the soil, often through subsistence
agriculture. This is despite the fact that, with few exceptions, the mountainous and dry geographical
terrain is not conducive for farming. Most islands in NTT—there are 556 in all, although only 42
are inhabited—face physical limitations of geography, such as low rainfall, lack of irrigation and
drinking water, and susceptibility to natural disasters (Sayogo 1994, p. 4). During the period in
which we conducted research, two of our research kecamatan had to be changed: one because of a
volcanic eruption; a second because of flash flooding.
Infrastructure is extremely limited and most villagers live some distance from markets. In addition
to the effect of the geographic conditions, poverty in NTT is arguably compounded by the inequitable
budget distribution between western and eastern Indonesia, and by corruption and poor governance.
The richest and middle class are normally the ethnic Chinese, traders from Java, Padang (West
Sumatra), the Bugis (South Sulawesi), or the elites of local government.
Ethnic variety in NTT is tremendously rich. One survey estimated there to be 61 linguistic
ethnic groups in the province alone (Grimes et. al. 1997), the result not only of migration over
the centuries, but of the continuing strength of clan and kinship systems, resulting in ethnic
differentiation taking place at a very localized level. In the 2003 census, 29.5% of people in
the province reported themselves as being of an ethnicity with less than 109 persons in the
province.228
The island of Flores, where the research took place, is particularly diverse, with the Flores-Lembata
area being home to 28 different ethnic-linguistic groups. However, within each linguistic group
there are a number of dialects that can make communication very difficult. Some informants we
spoke with put the number of distinct languages on Flores at more than 50. What is more, the
languages are often so different that people in neighboring kabupaten (districts) and even kecamatan
(sub-districts) cannot understand each other at all. Given the linguistic variation, and the
corresponding absence of an overwhelmingly dominant ethnic group (unlike in Java), Indonesian
is the only communal language for communication between people from different ethnic groups in
NTT, and Bahasa Indonesia penetrates perhaps deeper than in many other regions of Indonesia.
Throughout the last century and during the state-endorsed 1980s transmigration program, a small
number of immigrants from other Indonesian islands settled on Flores, mainly Savunese, Bimanese,
Buginese, Javanese and Makasarese.
With a population of 1,553,344, the island of Flores is the largest and most populous island in
NTT. Located between the famous Komodo island (to the west), Alor (to the east), Sulawesi
(to the north) and Timor (to the south), the island was named after the Portuguese name for the
island’s eastern cape, Cabo de Flores (Cape of Flowers). In contrast to East Java, the majority
228
Only in West Kalimantan, Central Sulawesi, Maluku, North Maluku and Papua was the percentage of ‘other’ ethnicities
(i.e., ethnicities behind the 17 largest) higher (Suryadinata et. al. 2003).
174
Annexes
of the Florenese are Catholic (84%), with only 13% of the populace adhering to Muslim beliefs,
thus almost inverting the national Indonesian average. Around 11% of the population of Flores is
recorded as being illiterate.
The number of ethnic groups, and the fact that difference is constructed at such a local level,
highlights the potential for competition. This potential is reinforced by the fact that most
administrative districts in NTT, and in Flores, are based upon a particular ethnic group. Evidence
from the research, and from observation of the island and province’s recent political past,
demonstrates the extent to which these ethnic groups (which are in fact made up of many smaller
suku) favor their own in the distribution of power (see Barron, Nathan and Welsh 2005).
2.1 Research Districts and Sub-Districts

The two research districts are at the opposite ends of Flores. Poverty is slightly higher in
Manggarai, but both districts are much poorer than any of their counterparts in East Java (see
Table A3).
Table A3: NTT Research Sub-Districts

(Population and Poverty)
Population Poverty Rate

Manggarai 591,222 0.68
Kota Komba 38,491 0.72
Lambaleda 26,158 0.77
Cibal 33,715 0.72
Ruteng 60,045 0.72
Sikka 264,560 0.61
Maumere 23,163 0.56
Nita 29,090 0.74
Paga 12,759 0.73
Talibura 23,474 0.63
Manggarai
Manggarai is by far the largest district in Flores and, indeed, NTT. In June 2003, the kabupaten
was split into two, with the formation of West Manggarai.229 Manggarai is relatively densely
populated compared to other districts in Flores (BPS 2002b). Despite the higher rate of poverty
in Manggarai, it arguably has greater development potential than the other districts on Flores. It
has good soil, wet rice irrigation, coffee plants, vanilla, pearls, gas power, and tourism.
Manggarai is the only district in Flores that was influenced and colonized by the Goa kingdom of
South Sulawesi as well as the Bima kingdom of the island of Sumbawa. Verheijen (2001) reports
229
Although all our selected sub-districts still fall within the borders of Manggarai district, we use the old boundaries
for the district-level analysis.
175
that the Goa influence in Manggarai came much earlier, before Goa became an Islamic kingdom.
The influence of the Bima kingdom was characterized by Muslim tradition, which still remains in
the western part of Manggarai, as well as on the northern coast of Flores.
Manggarai adopted the former Bima system of government system (the Kedaluan system), which
continues to influence everyday life and leadership. The Adak (kingdom) of Manggarai was
made up of 38 kedaluan. The Dutch used the kedaluan as the main unit of administration for
the area. The Dalu headed the kedaluan, governing the different gelarang (inter-village
governments) below. These gelarang, in turn, governed the villages (beo or golo) within their
area, each administered by a tua beo or tua golo (Toda 1999, p. 248). The Dutch appointed the
Dalu as King of Manggarai. This system of government was practiced in Manggarai until
Indonesian independence, although the Bima system formally left Manggarai in 1929 (Erb 1999).
During the pre-colonial and colonial period, the kedaluan were often at war with each other.
Verheijen (1991, pp. 16-21) has reported that Manggarai culture encompasses five cultural groups:
Komodo, Waerana, Rajong/Razong, Rembong, and one group in Ngada district (on the northern
coast). However, in reality, group identity still has its roots in the old colonial system outlined
above. The kedaluan in Manggarai still form the basis for group identity and (in some cases) inter-
group tensions. The fact that the sub-districts (kecamatan) that formally divide Manggarai today
were built upon the historical borders of the kedaluan (with one or more kedaluan absorbed into
each kecamatan), has only served to reinforce them as a basis of identity to the present day.
At the micro level, group identity is based on panga/mbaru (lineage) and the beo/golo (village).
Generally, these fit well together because people of the same lineage form one unit (sharing risk
and creating wealth) and live together in a village. Lineage and village identities can be extended
through marriage, as well as though alliances between lineages formed because of common
interests. Several lineages which have the same historical background will join together to fight
against other groups.
At the same time, there is also a unity between beo/golo and lingko (the traditional round
garden). A golo has its own communal garden as a center for the farming system; this garden
is meant to unite all the different lineages in a village. Because of this unity through communal
land, problem regarding land will usually involve the whole lineage or several lineages and
villages. Indeed, a large percentage of land in Manggarai is held communally and governed by
the traditional (adat) system. In such cases, boundaries are not formally titled or codified, but
rather are usually defined by geographic features such as field boundaries, rivers, roads, avenues
of trees and other (often changeable) features. These boundaries can be altered over the years
in response to changing needs and power balances.231 Traditionally, adat elders were the
repositories of knowledge for traditional land boundaries. Maps (kar) were drawn showing
borders, names and measurements of each piece of communal land, and were held by the King
of Tamur in the early twentieth century. However, during the Bupati Gaspar Ehok’s rule, this
map was burnt, some allege in an act of arson. As such, no written record exists for much of
the land in Manggarai.
231
See discussion in Barron and Sharpe (2005).
176
Annexes
In addition, adat systems, knowledge and authority have been under attack since the Dutch era.
Traditional methods of intervention and conflict resolution have been undermined. As the state—
and accompanying ideas of ‘modernity’—has penetrated to the level of the village, an ontologically
diverse market place has developed wherein different groups make appeals based on different
normative systems of justice (Bowen 2003). This can result in competing claims for land, leading
to conflict (Prior 2003; 2004; Clark 2004).
The adat (traditional) land system, where land belongs to the clan rather than the individual
member of the clan, also has the result of elevating individual claims over land into group claims.
This means that land conflict occurs more frequently, because a dispute over any part of the land,
however marginal, is likely to result in conflict. It also means that when conflict does occur, it is
likely to take an escalated form. Further, by conflating identity with claims for resources,
participants are less likely to see the outcomes as negotiable.
Map A3: Location of Manggarai Research Sub-Districts
Sikka
Sikka is smaller in size than Manggarai, and the difference in geographic size has an impact on
the transportation and information network. It is easier to get to villages in Sikka than in
Manggarai; the transportation infrastructure is better in the former, and villages are in general
better connected to local markets. Improved transportation links in Sikka are undoubtedly
contributing to growing modernization within the district.231
231
Over the course of the research, the district was connected to a mobile phone network, and then got its first
internet café
177
Sikka also has some degree of development potential, including a limited amount of tourism.
Crops grown include copra, cacao, and corn, but many areas face water shortage problems and
very dry land. Maumere and Nita, two of the kecamatan (sub-districts) where the research
took place, face serious water problems. During the dry season, people have to buy water for
family needs or they have to compete in order to get limited public water resources. The price
of cash crops is very low, making life difficult for many peasants.
Map A4: Location of Sikka Research Sub-Districts
Culturally, Sikka consists of two main different ethnic groups: Sikka Krowe and Lio. Sikka
Krowe is the biggest cultural group, but in the past it was divided into three swapraja (kingdoms):
Sikka, Nita and Kangae. Sikka was traditionally the most powerful group in the government
bureaucracy, largely because the Dutch viewed the group’s leadership as easy to work with.
During the Dutch administration, many different areas were combined to form the Sikka region,
under the control of an indigenous Sikka person. The Dutch implemented a strategy that
endeavored to create this region as the center for government, the economy, and the development
of religion. Such a strategy led to social jealousy amongst the indigenous groups as many felt
colonized by the Sikka people. Some consider the Sikka people to continue to be favored in
the post-colonial environment, and it has been argued that in the past the Catholic Church has
provided greater support for the indigenous Sikka people.232 However, the relative dominance
of the ethnic Sikka group has declined as education levels—and with them access to political
positions—have increased for other ethnic groups in the district. This has resulted in increased
competition between groups.
232
Interview with Oscar Maggdelini, an expert on adat law and local historian, Maumere, Sikka. While Catholi-
cism did not enter Manggarai until the first missionaries arrived in the 1912-1918 period, Sikka has been Catholic
since 16th or 17th century, when the religion was brought in by the Portuguese from the easternmost part of Flores
(the island of Solor).
178
Annexes
At the micro level, group identity in Sikka can be formed through kinship/lineage, through the
unit area (village/sub-village administrative area), or through common interests. Those who are
not from the same village or lineage are generally considered as orang luar (outsiders). Group
identity is often a gateway for nepotism and corruption. Whereas in Manggarai, communal land
forms a basis for group identity, in Sikka most land is held individually, with the exception of
some land belonging to the ethnic Lio group.
2.2 Research Villages

Within NTT, research was conducted in 20 villages (Table A4). Again, these vary greatly in terms
of years in which they have had KDP, population and poverty rates. Compared to East Java, villages
are in general smaller and have much higher rates of poverty.
Table A4: NTT Research Villages
No. District Sub-district Village Phase Treatment? Years KDP* Population Poverty Rate
1 Manggarai Kota Komba Watu Nggene 2A Control 1 3,060 0.76

2 Manggarai Kota Komba Golo Meni 2A Control 1 3,143 0.69
3 Manggarai Kota Komba Tanah Rata 2A Control 1 1,970 0.81
4 Manggarai Lambaleda Satar Punda 2B Treatment 2 2,793 0.85
5 Manggarai Lambaleda Golo Mangung 2B Treatment 3 2,337 0.86
6 Manggarai Lambaleda Tengku Leda 2B Treatment 3 1,399 0.81
7 Manggarai Cibal Pagal 3 Treatment 3 1,790 0.30
8 Manggarai Cibal Lando 3 Treatment 3 767 0.93
9 Manggarai Ruteng Wae Belang 3 Treatment 3 2,692 0.61
10 Manggarai Ruteng Cumbi 3 Treatment 2 2,765 0.75
11 Sikka Maumere Watu Gong 2A Control 1 2,148 0.46
12 Sikka Maumere Koting A 2A Control 1 1,285 0.40
13 Sikka Maumere Nele Wutung 2A Control 1 1,396 0.27
14 Sikka Nita Nita 2B Treatment 2 2,746 0.56
15 Sikka Nita Bloro 2B Treatment 3 1,387 0.66
16 Sikka Nita Magepanda 2B Treatment 2 3,372 0.87
17 Sikka Paga Loke 3 Treatment 2 1,650 0.91
18 Sikka Paga Pagal 3 Treatment 2 3,874 0.47
19 Sikka Talibura Werang 3 Treatment 4 1,040 0.80
20 Sikka Talibura Nebe 3 Treatment 4 3,194 0.54
* at the time key informant survey was implemented
179
Annex B: Evaluation Methodology
In recent years, the development community has taken welcome, if belated, steps to improve the
quality and frequency of evaluations, the better to improve the evidence base on which policy and
project decisions are made.234 Powerful incentives continue to conspire against making evaluation
a routine element of the project cycle (Pritchett 2002),234 but progress of sorts has been made in
terms of improving the rigor of the techniques used to identify the impact of project interventions
net of other contending factors. These advances, however, have largely been made in areas where
the intervention itself is relatively standardized (e.g., textbooks, cash transfers, infant immunizations)
and where project objectives (improving test scores, lowering infant mortality rates) are clear and
readily measurable. In such cases, calls have been extended to adopt the standards and techniques
of bio-medical research, where the canonical protocol requires (among other things235) random
assignment to ‘treatment’ and ‘control’ groups. “Just as randomized trials revolutionized medicine
in the twentieth century”, argue Duflo and Kremer (2005, p. 228), “they have the potential to
revolutionize social policy during the twenty-first.”
While entirely sympathetic to the need for more and better evaluation, the approach taken in this
study stems from the conviction that ‘participatory’ development projects such as KDP present
qualitatively distinct evaluation challenges, in that the intervention itself is a variable (not a constant,
like a tax cut).236 Adapting to the idiosyncrasies of the context within which the project is operating
is, by design, a central feature of CDD projects, even as the objectives being sought—enhancing
empowerment, participation, collective action, conflict management—are inherently hard to define
and measure. The core evaluation challenge remains one of disentangling project effects from
selection and context effects (see Baker 2000), but the inherent complexity of trying to assess
the efficacy of a large participatory project on local conflict—which is a product of numerous
simultaneously interacting variables—means that a correspondingly eclectic (though no less
‘rigorous’) methodology must be employed (cf. Mosse, Farrington and Rew 1998). Put another
way, the standards and goals remain the same, but the means used to realize them, on both
ethical and methodological grounds, must necessarily be those best suited to the distinctive
challenges the problem presents.
233
The full methodological details underpinning this study are outlined in Barron, Diprose, Madden, Smith, and Woolcock
(2004). A short version, drawing out the main methodological principles, is Barron, Diprose, Smith, Whiteside, and
Woolcock (2004).
234
Primarily because (a) the political and career stakes are such that the risk of finding out that a project does not work
persistently trump the value of any design lessons that could be learned by subjecting a project to a stringent evaluation,
and (b) the size of the portfolio one manages continues to be more prestigious than the demonstrated impact that one’s
projects might achieve (and which may only be realized long after the manager has moved on—e.g., in the case of
education).
235
See Whiteside, Woolcock, and Briggs (2005) for a full discussion of the actual standards of bio-medical evaluation,
and their limits when applied to assessing ‘participatory’ development projects.
236
The important (and complementary to our study) work of Olken (2005) shows that standard quantitative evaluation
techniques can be adapted to assess the efficacy of particular programmatic aspects of participatory projects such as
KDP, but even here this type of innovative research is only possible because of deep engagement with the particulari-
ties of Indonesian village life (i.e., prior and simultaneous qualitative work makes possible the more public stand-alone
quantitative findings). See also Alatas (2005).
180
Annexes
To answer the range of research questions and investigate the causal chain of events, we designed
a mixed method assessment strategy that incorporates qualitative and quantitative components in a
complementary and mutually reinforcing way.
1. Qualitative Methods
1.1 Contribution of Qualitative Methods

The distinctive empirical contributions of qualitative methods stem from their comparative advantage
with respect to assessing the complexities surrounding context-specific subject matter, and the
various processes shaping their historical evolution. As such, qualitative approaches are best suited
to assessing the dynamics and trajectories of local level conflict, as these phenomena are difficult to
quantify or reduce to static numbers, and are intricately tied to (and embedded in) local context,
areas with which quantitative approaches have difficulty adequately accounting for. Moreover,
since the dynamics of (local) conflict are far from being exhaustively researched, qualitative methods
helped the research team remain open to unexpected findings.
The study is especially concerned with issues of process, an area that, for two primary reasons, can
be more comprehensively addressed through qualitative exploration. First, qualitative approaches
are needed to identify the key mechanisms that trigger, sustain, or resolve conflict. To this end, the
research team conducted seven months of qualitative quasi-anthropological fieldwork to develop
case studies of how different actors—villagers, facilitators, local leaders—together negotiate (or
fail to negotiate) different types of conflicts in different settings. Using a modified version of the
process tracing method (George and Bennett 2005; Varshney 2002), researchers investigated ‘conflict
pathways’, seeking to understand the discrete stages in the evolution of conflicts. By doing this,
researchers were better able to identify the factors that transform underlying social tensions into
different outcomes (namely, violence or peace). Second, as an evaluation of KDP, the study sought
to dissect the various components of and processes within KDP, the better to understand how each
interacts with conflict (and the attempts by various actors to manage it). The identification of
positive and negative areas of KDP influence is information that the project managers can use to
improve the program’s effectiveness. For both areas, qualitative methods are critical to understand
(a) the mechanisms by which, and the conditions under which, particular variables become salient,
and (b) the importance of local contexts.
1.2 Types of Qualitative Data Collected

A research team of twelve researchers and supervisors conducted seven months of qualitative
quasi-anthropological fieldwork in 41 villages (see research areas and sampling strategy below).
When in the villages, the researchers had a number of tasks.
First, they developed case studies (68 in all) of how different actors—villagers, facilitators, local
leaders—together negotiate, or fail to negotiate, different types of conflicts in different settings.
Some of these cases were violent; some are not. Collectively, they covered a wide range of
dispute types, including land and natural resource conflicts (which range from large-scale
inter-ethnic conflicts to smaller private conflicts over the inheritance of land), cases of vigilante
justice (against thieves, witch doctors, etc.), gang fights, political disputes (e.g., over local elections
181
and administrative boundaries), conflict over development resources, and domestic and sexual
violence. Throughout the study, these case studies are used to illustrate particular points, as well as
for comparative analysis.
Second, the researchers collected information on 14 topic areas—ranging from how local government
functions, to local socio-economic conditions, to the role of traditional and religious leader—to
allow for cross-village comparison.237 The general data helped illuminate the cases we followed,
gave us a broader picture of the ‘conflict map’ of the research sites, and allowed us to test the
generality of the hypotheses that emerged from the conflict pathway case studies.238
1.3 Qualitative Data Collection Tools

Researchers used four qualitative instruments—in-depth key informant interviews, informal
interviews, focus group discussions (FGDs), and participant observation. The primary research
tool for the qualitative research was the in-depth interview, accounting for the majority of
researchers’ field time and the greater part of the data we report in this study. In all, the research
team conducted almost 800 in-depth interviews. They used an open approach, in which they
were given a range of topics to cover, but were free to choose which to address in a particular
interview; similarly, they were given a list of guiding questions, but were free to adapt them in
order to get at the information needed. Importantly, all the field staff were given extensive
training in research ethics—especially important when addressing sensitive issues such as conflict,
in which respondents themselves may well be implicated as victims or perpetrators—and specific
strategies for both conducting these interviews and writing them up afterwards in a manner that
would facilitate coherent data management (over ten thousand pages of text were generated by
the researchers) and subsequent analysis.
Researchers also used a secondary research tool, the Focus Group Discussion (FGD), for three
main purposes. First, they used FGDs to gather information on specific conflict cases from
groups that might be hard to access in other settings. These groups included the ‘marginalized’
(e.g., poor women), as well as those that one might have difficulty locating or accessing
individually (e.g., victims of domestic abuse). Second, they used FGDs to collect more general
background information on the villages being studied, including data on community life, groups
and networks that existed in the village, economic conditions, etc. Third, they used FGDs to
gather diverse perspectives on conflict and, more broadly, security. Women in particular often
have different perspectives than authority figures on what constitutes the major problems and
issues in a village, and how (and how effectively) they are being addressed. In all, over 100
FGDs were conducted.
237
Researchers used an open approach: they were given a range of topics to cover, but were free to choose which to
address in a particular interview. They were given a list of guiding questions, but were free to adapt them in order to get
at the information needed. See the project field guides and training manuals (Barron, Diprose, and Smith 2004) for a
full list of topics covered, instructions given to researchers, and research areas.
238
Full analysis of the qualitative data is given in the two provincial reports: Diprose (forthcoming) and Satu and
Barron (forthcoming). Other papers from the study also draw heavily on the qualitative findings (e.g. Gibson and
Woolcock 2005). A volume of case studies on land conflict collected during the study has been published (Clark 2004).
Volumes of case studies on development conflicts, vigilantism and local politics are forthcoming.
182
Annexes
For both in-depth interviews and FGDs, researchers followed specific guidelines regarding the
sampling of respondents. In effect, they covered a cross-section of the population, balancing authority
and non-authority figures, and men and women. They interviewed members of a variety of
organizations, professions and village groups. For the conflict cases, researchers used snowball
sampling to identify “experts” on a particular case—i.e., those involved (to varying degrees) and
observers.
The third and fourth research tools we used consisted of, respectively, the informal interview and
participant observation. The study used these ‘informal’ methods for two main purposes: first, to
gather key ‘unspoken’ information from the way people act, their relationships and so on; and
second, to gather spoken information people provided in informal environments. These techniques
further help establish relationships, build trust, and pave the way for formal in-depth interviews,
all of which take substantial time. Unlike the other tools, there was no formal sampling of
respondents. Rather, anyone the researchers met, and everyone they saw, was a potential source
of information. However, these techniques were especially useful at getting information from
‘marginalized’ and ‘silenced’ groups. For example, in some cases women were uncomfortable
being formally interviewed, but were happy to talk when the researchers were helping them in
the house.239
2. Quantitative Methods
2.1 Contribution of Quantitative Methods

The main contributions of quantitative methods revolved around establishing a basis for generality
and causality, a task difficult for the qualitative work because of its small samples and the
(deliberately) non-random selection of research sites and respondents. On the broadest level, the
research design used a quantitative sampling frame to capture the major dimensions of heterogeneity
within the population and to increase the reliability of results. Because we are looking for common
patterns of project impact, our qualitative findings would be strengthened if they held up in a
variety of different settings. Thus we chose two very different provinces and both high and low
capacity districts for fieldwork, and then tracked similar conflict cases in matched KDP and non-
KDP sub-districts (see next section). We used qualitative investigation to verify the accuracy of
matches identified through quantitative techniques to ensure that they reflected realities on the
ground. (Put more formally, qualitative methods can help compensate for a key weakness of
quantitative matching techniques, namely that they can match only on ‘observable’ variables, such
as age, income, land size, etc. It is well known that ‘unobserved’ and/or ‘unobservable’ factors,
such as motivation and political connections, influence project placement and participation; as
such, the deployment of qualitative approaches as a complement to quantitative data helped us to
‘observe unobservables’, thereby generating a more accurate empirical match between the KDP
and non-KDP settings).
239
The field researchers often lived in the houses (spare rooms) of such interviewees during their time in the village,
since there were of course no hotels nearby in which to stay. The researchers were there for periods of several weeks,
and as such these living arrangements afforded them a unique opportunity to engage in quite detailed and personal
conversation in a ‘safe space’ (e.g., the kitchen) with those who would otherwise be very hard to reach. It also meant,
importantly, that they had multiple occasions on which to cross-check the reliability and validity of both respondents’
statements and their own observations.
183
Where the qualitative data was susceptible to the subjectivity of individual researchers collecting
it, we used less contextual methods to elicit more objective numerical data. Surveys, by definition,
collect general non-case-specific data. By putting together a large-N dataset consisting of indicators
and variables that capture the wide range of factors we think may affect levels of violent conflict,
we could more readily test the validity and generality of a range of hypotheses. With this kind of
data, we endeavored to isolate correlations between multiple contributing factors (local
characteristics, civic institutions, KDP implementation quality), and separate the impacts of KDP
from the influence of particular characteristics of the research sites. The larger samples allow us to
test the extent to which our findings hold across a broader range of research sites (villages, sub-
districts, districts, and provinces).
2.2 Types of Quantitative Data Collected and Data Collection Instruments

Whereas the qualitative instruments collected both case and general data, the quantitative survey
instruments collected only the less contextual general data. We used three quantitative instruments:
a focused key informant survey; a very large but less focused key informant survey (PODES); and
a nationwide household survey (GDS).240
Key Informant Survey

In the third phase of the qualitative research, the research team administered surveys to a range
of key informants at the sub-district and village levels, which was limited to the research sites
where the qualitative research was conducted.241 The survey gathered comparable responses to
perception questions relating to the outcomes and processes of KDP, and to broader dimensions
of social change, in the same locations as the qualitative research. The questionnaires focused
largely on the role of KDP in the locality, the extent to which it was used to solve KDP and non-
KDP problems, and the extent to which ‘spillover’ effects could be determined. A shorter survey
was implemented in control sites (shorter, because obviously questions could not be asked about
KDP in locations were it had not been operational). In all, the survey was implemented in 40
villages (28 treatment and 12 control villages). These villages vary along a number of dimensions:
presence of KDP (and years of KDP), governance and problem solving ‘capacity’, functioning of
the program, etc. As such, the data gave us information not just on project effects (comparing
control and treatment sites), but also allowed us to test a number of explanations for variation
within KDP sites. The surveys were administered to three informants at the sub-district level,
and eight at the village level, thus giving a sample of 352 respondents (kecamatan respondents
were not interviewed in control locations).242
PODES
The Government of Indonesia’s Central Bureau of Statistics’ (BPS) Village Potential series
(PODES) is a long-standing survey that collects data at the lowest administrative tier of local
government. It collects detailed information on a range of characteristics—from infrastructure
to village finance—for Indonesia’s current 69,000 villages and neighborhoods. PODES was
fielded at the end of 2002 as part of the 2003 Agricultural Census. For the first time, the 2003
240
To be precise, we designed and fielded the key informant survey; contributed one module to the GDS survey but did
not participate in the fielding; and borrowed data generated by the government PODES survey.
241
Full analysis of the key informant survey data is given in Barron, Clark, and Mawardi (2004).
242
See Annex E for a full break-down of respondents.
184
Annexes
PODES included a section on politics, conflict and crime. The PODES data helped to map out the
incidence of conflict and violence across Indonesia. It also allowed us to test a range of basic
hypotheses about factors correlated with higher levels of conflict (see Barron, Kaiser and Pradhan
2004). The limitations of this data source are that the only respondents interviewed at the village
level were village heads, who tend to underreport conflict in their villages and have a vested interest
in presenting a picture of peace and harmony in order to access development funding.
Governance and Decentralization Survey (GDS)

A second large-N dataset is the World Bank’s Governance and Decentralization Survey (GDS),
the first (baseline) iteration of which was conducted in 2002. The GDS is part of the Indonesian
Decentralization Empirical Analysis (IDEA) project, which is being conducted by the World
Bank together with CPPS-GMU (Centre of Public Policy Studies-Gajah Mada University). The
survey aims to ascertain the effects and impacts of the rapid decentralization that took place in
Indonesia in 2001 on a number of factors ranging from the performance of local governments, to
service delivery, to the functioning of the justice system. A follow-up GDS-2 survey will be
fielded in mid-2006, and will provide information on the linkages between conflict and various
aspects of governance and public administration. While the data is obviously not used in this
paper, it will inform future outputs from this research project. The initial (baseline) enumeration
took place in 150 (i.e., almost half) of Indonesia’s districts in 2001. As part of the follow-up
survey, a module on conflict and problem solving will be inserted into the survey, and the sample
will be extended. The GDS will thus provide a rich data-source on levels of conflict (as reported
at the household level) and data on a range of other ‘institutional’ and ‘governance’ factors that
can be regressed against conflict/violence levels. The formulation of this module was based in
part on the emerging hypotheses from the qualitative fieldwork, newspaper analysis and the
results of the PODES survey. Methodological insights from the latter also affected the survey
design itself.
3. Secondary Data
Given the fact that violence is a non-random and relatively scarce event, and given the reality of
limited budgets and hence sample sizes, it is difficult to use household survey instruments in
order to accurately capture levels of violence in a locale. Key informant surveys overcome some
of these problems but can encounter others, such as massive underreporting of conflict and biases
about the types of conflicts reported.243 There are also weaknesses to the PODES survey (see
above). As such, we used secondary data collection to understand better violence incidence.
Secondary data was also used to provide better understanding of the contexts, and the historical
factors that helped shaped them, in our research areas (see Annex A).
243
See Barron, Kaiser and Pradhan (2004) and Barron and Sharpe (2005) for more on the shortcomings of survey
instruments in measuring conflict.
185
3.1 Newspaper Data Collection

In order to assess patterns and forms of conflict, and variations between areas, we needed to
create a conflict map showing levels of conflict across geographic areas. However, satisfactory
data did not exist to allow for this. As Barron, Kaiser, and Pradhan (2004) note, household and
key informant surveys are weak at accurately measuring conflict. Formal government statistics
are also poor.244 Because of this, we decided to create datasets of all reported conflicts in the
research areas (and surrounding districts) for the 2001-2003 period using three local news sources
for each area. This allowed us to broadly map conflict in our research areas, to estimate aggregate
levels of violence, to characterize the incidents (by conflict type, actors involved, impacts, etc.),
and to identify how representative our qualitative case studies were.245 Analysis of the newspaper
data, and an outline of the methodology, is presented in Barron and Sharpe (2005).
3.2 Other Secondary Data

Other secondary data included information on conflict levels and impacts (police criminal data
and information from health care providers), primary documents in villages (e.g. written adat
laws), and KDP documentation (e.g. on complaints and how they are handled). An extensive
review of the literature—on development, on conflict, on Indonesia and on our specific research
areas—was conducted. Background papers on a number of issues relating to KDP, local conflict,
and the areas where the study took place were also written throughout the study.
4. Integration of Methods
The research design uses an iterative strategy to integrate qualitative and quantitative methods,
establishing an ongoing dialogue between the two approaches. By examining the subject matter
from multiple angles, we used triangulation to verify that our research findings reflected the ‘true’
relationship between KDP and local conflict. Figure A1 (below) illustrates how the combination
of methods and types of data provide a more holistic picture of conflict and KDP impacts, than
any single source or methodological approach alone. As such, each item provides a distinct
contribution toward capturing the larger ‘truths’ our research seeks to uncover.
Preliminary research (e.g., statistics, academic writings, newspaper articles, etc.) on the field
sites was gathered before the intensive qualitative fieldwork began, and informed the design and
implementation strategies for the qualitative and quantitative work that followed. This initial
work began the process of mapping conflict (its spatial distribution and its characteristics) to
help in the sampling of districts and sub-district sites, and it provided input into the development
of the research hypotheses, questions and the data collection instruments.
244
As noted by Varshney, Panggabean, and Tadjoeddin. (2004, p. 10), the New Order did not publish any figures on
the impacts (death, injuries, other destruction) of conflicts. At present, a number of different ministries and depart-
ments have launched programs that aim to assess the potential for conflict. Kesbang (within the Ministry of Home
Affairs), and the Centre for Religious Harmony (within Department Agama) have started such programs, with the
Coordinating Ministry of Politics and Security (Menko Polkam) coordinating the information. However, these
programs are still at embryonic stages, and do not provide comparable information across different geographic areas
(see UNDP 2004, especially Technical Annex A on ‘Knowledge Management for Conflict Risks, Impact and Needs
Assessment’). Police data is also weak, and analysis is hampered by the lack of common definition and categories.
245
The methodology utilized for the newspaper mapping, and full analysis of the data, is outlined in Barron and
Sharpe (2005).
186
Annexes
Figure A1: Research Methods Used in the Study
Once underway, the in-depth qualitative research identified the right kinds of questions (and
their wording) for inclusion in the more general quantitative surveys. Importantly, the quantitative
instruments were designed and developed while the qualitative work was being conducted; as
such, their design reflected the ongoing findings of—and methodological lessons learned from—
the qualitative fieldwork. In return, the quantitative fieldwork provided a test of the generality of
the hypotheses emerging from qualitative investigation. Analysis of newspaper and other
secondary data sources also helped to estimate aggregate levels (and the impact) of violent conflict
in our research areas.
Many of the areas of information on which we wanted to collect specific survey data mirrored
the general data categories the qualitative researchers used. This was deliberate: looking at a
problem from more than one methodological angle gives us increased confidence that what we
are finding is accurate, and that the inevitable, although variable, bias built into any research
method is not unduly distorting findings. In short, if we get the same answers to the same questions
twice, but using different methods and taking a different approach, we can be reasonably confident
that the answers are as reliable and objective as possible.
5. Sampling Frame
One of the most important aspects of the research design is the use of a quantitative sampling
frame to select the sites for qualitative investigation. As indicated above, accounting for
heterogeneity within the population gives us greater confidence that our research findings do not
turn on the particular (idiosyncratic?) characteristics of our research sites.
187
5.1 Provincial Level Variation

The objective of the provincial selection for the qualitative work was to pick two very different
provinces in which to work: as we seek common patterns of project impact, our findings will be
strengthened if they hold up in different settings. We focused on a range of variables to help
determine the nature of a province, including (a) population size and density, (b) ethnic homogeneity,
(c) religious homogeneity (and dominant religious group), and (d) overall level of provincial
development, including provision of and access to public services and infrastructure.
We excluded provinces with the highest and lowest levels of conflict. Since KDP operates at the
sub-district level and below, any positive externalities it may produce are likely to be directed at
managing conflicts that exist at those levels. Given the nature of conflict in many high-conflict
provinces, where cleavages exist on a provincial or at least district level, if we had selected such
provinces we would have biased our research against observing any project impact. Moreover, in
areas of high-conflict—where levels of violence are significantly affected by external actors and
exogenous factors (such as military action)—it would be much harder to separate out the potential
impact of a local level project from all the other causal variables in the research site.
5.2 District Level Variation

In each province, we selected two districts—one with a ‘high capacity’ to manage conflict, the
other with a ‘low capacity’ to manage conflict (see discussion of capacity, below). We selected
these after extensive consultation at the provincial level with government, international and local
NGOs, regional development experts, research institutes and religious institutes, universities, and
KDP staff. Picking both ‘high’ and ‘low’ capacity districts for each province allows us to defend
our claims regarding the nature and extent of KDP’s impact on local conflict resolution by showing
that they take place irrespective of whether the broader environment is ‘conducive’ to conflict
resolution or not.246
5.3 Sub-district Matching

For the first two phases of qualitative fieldwork, we chose two sub-districts within each of the
research districts—a KDP site and a non-KDP site—that were as similar as possible. The
former had already had KDP for three years (our ‘treatment’ sites); the latter had not yet had
KDP (our ‘control’ sites). Selecting similar matched comparisons controls as much as possible
for non-program effects that may stem from socio-economic, institutional, or other differences.
For reasons outlined above, we used mixed methods to identify these matches. We first used
the propensity score matching technique247 to select similar KDP and non-KDP sites on the
basis of pre-intervention characteristics. The actual propensity score was derived using
explanatory variables from the PODES 1996 dataset that could serve as proxies for the economic
level of the kecamatan. Among the PODES variables used were: population, access to markets,
department stores, health and education resources, income, and perceptions of poverty level.248
These are ‘observable variables’, however, so to control for ‘unobservable’ variables (e.g.
motivation, cohesion, leadership, political connections), we used the propensity score to select
246
Ideally, we would have used statistical data on the incidence of conflict to help make these selections, however such
evidence did not exist in Indonesia at the time.
247
A statistical technique used to identify otherwise comparable ‘treatment’ and ‘non-treatment’ groups on the basis of
the probability of being selected. See Rosenbaum and Rubin (1983; 1985), or, for a general introduction, Baker (2000).
248
Our thanks go to Vivi Alatas for doing the propensity score matching.
188
Annexes
three statistically comparable non-KDP (‘control’) kecamatan, and then asked our field
research team to identify which of these was the most appropriate match for the KDP (‘treatment’)
kecamatan. They did this by conducting qualitative interviews at the district level with government
officials and other experts to chose amongst the identified pairs and incorporate other sources of
difference or similarity.249 This generated the overall sample frame in Figure A2 below.
Figure A2: Overview of Site Selection
East Java NTT
High Capacity Low Capacity High Capacity Low Capacity

Ponorogo Pamekasan Sikka Manggarai
KDP Non-KDP
KDP Non-KDP KDP Non-KDP KDP Non-KDP Lamba Kota
Badegan Sampung Proppo Palengaan Nita Maumere Leda Kumba
The third phase of qualitative research extended to an additional two sub-districts in each
research district. We did this to verify that KDP’s performance in the research sites was
representative of how it works in other sub-districts within the same district, and to include
new ‘treatment’ sites to replace areas where KDP was not working as intended.250 We selected
additional sub-districts that had had KDP for at least three years, had passed a minimum
threshold of KDP performance acceptability relating to transparency and accountability, and
were different from the other KDP sites we were studying - culturally, geographically, and
demographically. We added two KDP sub-districts (Slahung and Jenangan) in Ponorogo; two
(Pasaen and Pademawu) in Pamekasan; two (Talibura and Paga) in Sikka; and two (Cibal and
Ruteng) in Manggarai.
5.4 Case and Village Selection

We determined specific qualitative research locations by conflict case, rather than by village.
Our initial maps of major social tensions (constructed in Phase 1 of the research) helped us select
the following primary cases in each sub-district to investigate in non-KDP sites (Phase 2A) and
matched KDP sites (Phase 2B):
249
We learned through this selection process that ensuring the accuracy of propensity score matches involves a substan-
tial amount of field investigation. In particular, cultural differences influence conflict management and are difficult to
measure and match statistically.
250
Where the program was not working properly, KDP impacts were clearly difficult to find. To compensate, we
purposively extended the sample into sub-districts where KDP was functioning properly. In effect, this means that our
research investigates KDP impacts where KDP is functioning relatively well, and not KDP program impacts more
generally, although as we discuss throughout the paper, there is considerable variation in KDP performance and func-
tionality throughout our sample.
189
Cases One and Two: Similar cases, different outcomes (within the same sub-district)
We selected two similar cases of conflict within each sub-district, but with different outcomes: one
violent, one peaceful. We evaluated similarity by type and scale of conflict, underlying tensions,
and types of actors. By examining similar cases we gained a better idea of what factors are present,
or missing, that result in different outcomes (violence or peace).
Case Three: Similar cases in two sub-districts (KDP and non-KDP)
We chose this case to match a conflict case in the non-KDP sub-district with a similar case in the
KDP sub-district. We used the same similarity criteria above, but compared cases in the two different
sub-districts (KDP and non-KDP) within the same district. Not only were the conflicts similar in
type and scale, but the general characteristics of the villages were similar as well.
Cases Four and Five: Peaceful resolution in violent area; violent case in peaceful area
Here we used a qualitative version of a difference in difference strategy, identifying instances of
peace in otherwise relatively high-conflict villages, and instances of high conflict in otherwise
relatively peaceful villages.251 This yielded insights on local mechanisms for peace and conflict
that were as independent as possible of the broader institutional environment (thereby replicating at
the local level the broader selection strategy adopted at the district level). As such, we compared
similar cases in matched (by pre-intervention characteristics) KDP and non-KDP locations to evaluate
any possible program impacts. Moreover, we selected similar villages and cases with similar conflict
dynamics.
Case Six: KDP-related case(s)
For the research in the treatment sites (Phase 2B), we also chose KDP-related cases. These cases may
have been an example where KDP had—directly or indirectly—caused conflict, or where we had
preliminary evidence that KDP had been used to help resolve conflict. In Phase 3, we encouraged the
research team to find more KDP-related cases, as well as those related to other development programs.
Table A5 summarizes the comparative analysis strategy.
Table A5: Overview of Units of Analysis/Primary Data Sources

# Form/Unit of Analysis Source(s) Variables comparing* Sample
1 Comparative pathways Conflict case studies O, M 68
2 Comparative villages Fieldwork/demographic forms C 41
Key informant survey C, M 41252
3 Comparative sub-district Fieldwork/demographic forms C 16
Key informant survey C, M 12 253
Newspaper dataset O 208 254
* Types of variables: C = Contexts; O = Outcomes; M = Mechanisms
251
‘Peaceful’ resolution of conflict is a relative concept, which includes ‘less violent’ and ‘nonviolent but not harmoni-
ous’ resolution. In Indonesia, ‘peaceful’ resolution at times means suppression of a particular incident, with the under-
lying conflict continuing to fester below the surface. See discussion in the next annex.
252
In some places in the paper, we use the key informant survey data only from the treatment areas (12 sub-districts and
30 villages). In other places we use data from treatment and control sites (41 villages).
253
The key informant survey was not conducted at the kecamatan level in control sites.
254
We used data from PODES for 2001, updated by data from the electoral commission (KPU) from 2003, for a full list
of sub-district (kecamatan). Since then some kecamatan have split. However, we code new kecamatan as still remain-
ing within previous boundaries, as of 2003.
190
Annexes
Annex C: Analytical Tools
1. Assessing Local Capacity: Use of Vignettes
For the analysis, we sought to develop measures of ‘local capacity’. This is difficult, because
‘capacity’ is an abstract concept and is difficult to define without reference to concrete examples. It
is also difficult to develop a capacity metric because conceptions of what constitutes ‘capacity’,
and what to measure it against, will necessarily vary across places, people, and cultures. As such,
we sought to use anchoring vignettes to try to ensure a common metric, and hence to allow for
comparison across areas.255 As shown in Box A1, the anchoring vignettes method attempts to
measure the (in)comparability of responses by getting a better understanding of how respondents
comprehend the question that they are posed.
Box A1: Political Voice in Mexico and China
In an attempt to understand “political efficacy” across countries, the following question was asked in both China and Mexico:
How much say do you have in getting the government to address issues that interest you?
(1) No say, (2) Little say, (3) Some say, (4) A lot of say, (5) Unlimited say
Interestingly, respondents in Mexico reported substantially less “political voice” than those in China. For example, more
than 50% of Mexican citizens reported having no say in government, whereas less than 30% of Chinese citizens replied that
they had no say in government. Given the striking different political situations of the two countries, these responses suggest
a potential problem with the survey question or how it is understood. There are a number of reasons that Mexican and
Chinese citizens may have very different perceptions of what “having a say” in government means. Additional questions
can be used to anchor their perceptions of what they understand. For example:
[Moses] lacks clean drinking water. He would like to change this, but he can’t vote, and feels that no one in the
government cares about this issue. So he suffers in silence, hoping something will be done in the future. How much
say does [Moses] have in getting the government to address issues that interest him?
This allows respondents to be broken into categories based on their understanding of Moses’ rights. With additional vignettes,
respondents can be placed on a spectrum of understanding of political efficacy that allows responses about their own
political voice to be interpreted more accurately.
Source: King and Wand (2004)
The anchoring vignettes methodology was developed primarily for survey use. However, we
used the method in a somewhat different way. Researchers were asked to categorize the capacity
of the different villages they had worked in along two dimensions: problem solving and capacity
for collective action. Collective action was defined as the positive act of coming together to
create or address something; problem solving could relate to either formal or informal responses
to a problem. We developed a series of vignettes that they used to categorize the villages, through
deciding which vignettes best described the situation in a given village (see Box A2).
255
King, Murray, Salomon, and Tandon (2004) provide a detailed discussion of the underlying principles and rationale.
For additional application of anchoring vignettes to development issues, see Das and Hammer (2005).
191
Box A2: Anchoring Vignettes Used to Assess Capacities for Problem Solving and Collective Action
Which of these scenarios best characterizes life in your village?
Problem Solving
1. [X] had a disagreement with [Y]. Though it was initially a small argument, [X] and [Y] have called on their respective
groups to intimidate the other person and seek retribution. As a result, tension increased in their village, as more
people became involved. There was no person or group that is willing to assert their authority to solve the
problem and no one trusts the formal legal system to settle the issue. The problem continues to grow.
2. [X] had a disagreement with [Y]. Though it was initially a small argument, [X] and [Y] have called on their
respective groups to intimidate the other person and seek retribution. As a result, tension increased in their
village, as more people became involved. Authority figures try to step in if the problem becomes violent, but there
is very little that they can do to solve the problem. The problem does not grow, but continues to fester.
3. [X] had a disagreement with [Y]. They took their problem to the village head, who was able to provide a solution.
Though they both agreed to a solution when they were in the meeting, [X] does not carry through with his part of
the agreement. He plans on taking the issue to another authority figure or even to court for another solution that
will be closer to her/his goals.
4. [X] had a disagreement with [Y]. They are able to reach a solution with the assistance of {tokoh masyarakat}/
village head. He is able to resolve the problem. Even though both [Y] and [X] are still upset, they will abide by the
agreement because it is important to uphold the rules of the community.
Capacity for Collective Action
1. The main road in village [M] was severely damaged in a recent flood. Though everyone in the village needs the
road to be repaired as soon as possible there are many disagreements about the building process. [X] is losing
a great deal of money from the road damage, but he/she does not feel he/she has enough power within the village
to help craft a solution. Most community members want the road repaired, but are afraid to get involved because
it will cause more trouble than the gains it might bring. The road remains unrepaired.
2. The main road in village [M] was severely damaged in a recent flood. Though everyone in the village needs the
road to be repaired as soon as possible, disagreements about the building process have developed. There was
initial support for plans to rebuild the road: materials were purchased, and many people planned to help with
construction. Now, some of the materials have been stolen and community members are less willing to assist, as
they no longer think that the project will succeed. [X] is very angry since he/she made a contribution, but it was
wasted, and now there is little that he/she can do.
3. The main road in village [M] was severely damaged in a recent flood. Everyone in the village agreed that the road
needed to be repaired as soon as possible and they were able to work together to fix it, despite many problems
along the way. Some of the money and materials for the road were stolen, so the road is not as good as it could
have been. Many people are angry that some of their money was stolen, and they are worried that the road will
soon wash out again.
4. The main road in village [M] was severely damaged in a recent flood. Everyone in the village agreed that the road
needed to be repaired as soon as possible and they were able to work together to fix it. Most of the villagers were
involved in some part of the rebuilding process and the community as a whole is satisfied with quality of the road.
There were, of course, some disagreements along the way, but they were able to be resolved in a way that kept
most people content.
192
Annexes
The vignettes were developed through an iterative process with the help of the researchers themselves.
Based on a review of case studies and village demographics forms, draft vignettes were written.
They were then checked with a number of researchers, and revised to more accurately reflect village
situations that would be applicable across research locations. The final vignettes were sent to all
researchers, who used them to rank the villages where they conducted research. After villages were
ranked, some of the rankings were checked with groups of researchers to see if the rankings (and
vignettes) were accurate.
The two dimensions of capacity were aggregated to provide an overall indicator for the village.
This indicator was broken from a five point scale to a three point one (High, Medium and Low)
based on the overall distribution of villages. These categories, in turn, were used for further analysis
in understanding conflict dynamics.
Table A6: Village-level Capacity
Phase Kabupaten Kecamatan Village Problem Solving Collective Action Score Ranking
3* Manggarai Cibal Pagal 4.00 4.00 4.00 High

3* Manggarai Ruteng Wae Belang 4.00 4.00 4.00 High
3* Pamekesan Pasaen Sana Daya 4.00 4.00 4.00 High
3* Ponorogo Jenangan Kemiri 4.00 4.00 4.00 High
3* Sikka Paga Loke 4.00 4.00 4.00 High
3* Manggarai Ruteng Cumbi 4.00 3.50 3.75 High
3* Pamekesan Pademawu Padellegan 3.50 4.00 3.75 High
3* Pamekesan Pademawu Sumedangan 3.50 4.00 3.75 High
3* Sikka Talibura Wereng 3.50 4.00 3.75 High
3* Sikka Talibura Nebe 3.50 4.00 3.75 High
3* Manggarai Cibal Lando 3.00 4.00 3.50 Medium
2A Manggarai Kota Komba Watu Nggene 3.33 3.67 3.50 Medium
2B Pamekesan Proppo Tattangoh 3.33 3.67 3.50 Medium
2A Sikka Maumere Watugong 3.33 3.67 3.50 Medium
2B Pamekesan Proppo Maper 3.67 3.00 3.33 Medium
2B Ponorogo Badegan Biting 3.00 3.33 3.17 Medium
2B Sikka Nita Nita 3.00 3.33 3.17 Medium
2B Manggarai Lamba Leda Satar Punda 3.00 3.00 3.00 Medium
2B Pamekesan Proppo Proppo 3.67 2.33 3.00 Medium
2B Ponorogo Badegan Dayakan 2.67 3.33 3.00 Medium
3* Ponorogo Slahung Wates 3.00 3.00 3.00 Medium
3* Ponorogo Slahung Slahung 3.50 2.50 3.00 Medium
2A Sikka Maumere Koting A 2.67 3.33 3.00 Medium
3* Sikka Paga Paga 2.00 4.00 3.00 Medium
2A Manggarai Kota Komba Golo Meni 2.33 3.33 2.83 Medium
2A Manggarai Kota Komba Tanah Rata 2.33 3.33 2.83 Medium
2B Sikka Nita Bloro 3.33 2.33 2.83 Medium
2B Manggarai Lamba Leda Golo Mangu 2.67 3.00 2.83 Medium
2A Pamekesan Palengaan Palengaan Laok 2.67 3.00 2.83 Medium
2A Ponorogo Sampung Pagerukir 2.00 3.67 2.83 Medium
3* Pamekesan Pasaen Tlontohraja 2.00 3.50 2.75 Low
2B Ponorogo Badegan Badegan 2.33 3.00 2.67 Low
2B Manggarai Lamba Leda Tengku Leda 2.67 2.33 2.50 Low
2A Sikka Maumere Nele Wutung 2.33 2.33 2.33 Low
3* Ponorogo Jenangan Panjeng 2.50 2.00 2.25 Low
2A Pamekesan Palengaan Palengaan Daja 2.67 1.67 2.17 Low
2B Sikka Nita Magepanda 2.67 1.67 2.17 Low
2A Pamekesan Palengaan Banyupelle 2.67 1.33 2.00 Low
2B Pamekesan Proppo Panagguan 2.33 1.33 1.83 Low
2A Ponorogo Sampung Gelangkulon 1.67 2.00 1.83 Low
2A Ponorogo Sampung Sampung 2.00 1.00 1.50 Low
193
2. Classifying Conflict Outcomes
A key part of the analysis strategy was comparative analysis of conflict cases. Analyzing seemingly
similar cases, over similar issues and in similar structural environments but with different
outcomes, can help illuminate the particular factors that led to given conflict outcomes. As such,
for each of the 68 case studies tracked, we sought to categorize its seriousness and impacts. We
used two variables for our categorization: the extent to which a conflict had violent impacts; and
the extent to which resolution was successful and lasting. Each conflict was coded along these
two dimensions, on a scale of one to three:
Impacts
1. Deaths (where conflicts involved at least one death)
2. Serious damage (injuries, property damage, kidnapping, rape, serious intimidation)
3. Little or no damage (where the negative impacts are indirect and small, or where there were
no negative impacts)
Resolution
1. Long-term resolution
2. Short-term resolution
3. None
As such, each of the conflict cases will fit into one of the nine boxes in the grid below (Figure A3).
Those cases in the top left box are most serious: they have had the highest impact and appear to
have the greatest chance of continuing or reigniting in the future. At the opposite extreme cases on
the bottom right box are those “successful” cases where there was no or limited damage and where
resolution is deep and lasting. Our cases can be arrayed across the table; they are coded along the
two dimensions in Annex D.
Figure A3: Seriousness of Conflicts
VIOLENCE IMPACTS
Deaths Serious Limited or
Damage No Damage
None
DURABILITY
OF Short-term
RESOLUTION
Long-term
3. Patterns of Conflict in KDP and non-KDP Areas
3.1 Creating the Dataset

In order to assess patterns of conflict in areas that had received KDP and areas that had not, we
merged the data in the KDP & Community Conflict Negotiation (KDP & CCN) dataset with
information on which sub-districts KDP had been operational in, and when. The dataset contains
information on conflicts in East Java and NTT for the years 2001, 2002 and 2003. We used the
194
Annexes
resulting dataset to examine KDP’s impact on conflict outcomes in two ways: first, we looked to
see if there was an impact on violent conflict due to a kecamatan ever having KDP; second, we
disaggregated the data to see if there were different effects on the number of violent conflicts as
locations spent more years in KDP.
For both types of analysis, we broke the observations out by year, as shown in Table A7 below:
Table A7: Overview of KDP and Conflict Data

Information on KDP Possible number of years of Matched with conflict data
in years: KDP: from year:
1999-2000 0, 1, 2 2001
2001-2002 0, 2, 3 2002
2003-2004 0, 1, 2, 3, 4 2003
*note: we dropped locations with 4 years, since the number of observations was too small.
We combined all of this information, creating a new dataset that contained the number of years of
KDP for each of the three years of available data. That means that each location appears three times
in our data set. For example, for two different villages: Village A that started in the second year of
KDP 1 and Village B which started in the first year of KDP 2, observations would be as depicted in
Table A8 below:
Table A8: Overview of Comparative KDP Data in Different Villages

Location Number of Conflict Location Number of Conflict
years in KDP variables years in KDP variables
Village A 0 2001 data Village B 0 2001 data
Together, they would provide six observations in the data. Counting this way (rather than just
looking at the total data in 2003) allows us to be more precise in looking for KDP effects and rids
(or at least minimizes) our analysis of any bias that might occur by counting conflicts in a particular
location that happened before the arrival of KDP.
3.2 The Impact of Exposure to KDP

To look at the impact of exposure to KDP, we broke the data into those locations that had never
received KDP and those that had. Based on the examples of Villages A and B above, we would
have had three observations of places that never had KDP (Village A, 2001; Village B, 2001 and
Village B, 2002) and three observations that had KDP (Village A 2002; Village A, 2003 and
Village B, 2003).
Once the data was broken down in this way, we looked at the differences in impact on violent
conflict, and conflicts that resulted specifically in deaths, injuries or property damage. Since we
would expect KDP to be effective in preventing conflict, not in reducing the number of deaths or
injuries, etc., we counted cases that contained violence, not the number of people that were impacted.
195
3.3 KDP Impacts over Time

We expect the impact of KDP to increase over time, improving as people (a) gain familiarity with
the procedures, (b) trust the process, (c) see its benefits, and (d) adapt it to their own needs. To
try to evaluate the impacts over time, we broke the village observations down by number of years
that they had KDP. Again, drawing from the example above, the data would be broken down as
outlined in Table A9:
Table A9: Overview of KDP Impact Data over Time

Years of KDP Data
0 Village A, 2001
Village B, 2001
Village B, 2002
1 Village A, 2002
Village B, 2003
2 Village A, 2003
With the data broken this way, we conducted the same analysis regarding violent conflict, deaths,
injuries and property damage. This allowed us to look for trends over time. Again, we used the
number of cases in our analysis, rather than the number of people affected or the amount of total
property damage.
196
Table A10: List of Conflict Pathway Case Studies
No. Province District Sub-District Case Name Violent Impacts Resolution

1 East Java Ponorogo Sampung The Grosok Case: A Portrait of the Little or No Damage Short Term
Complexity of Conflict at the Local
Level
2 East Java Ponorogo Sampung No Compensation Paid: Conflict Little or No Damage None
Over the Division of Profits Between
Limestone Miners and PD Sari
Gunung
3 East Java Ponorogo Sampung Workers Wages: Conflict over the Little or No Damage Long-term
Division of Profits between
Limestone Miners and PD Sari
Gunung
4 East Java Ponorogo Sampung Fire at the Balai Desa: The Complex Serious Damage Short Term
Conflict over the Village Head
Elections
5 East Java Ponorogo Sampung I’ve Bought Your Votes, So I Win: Serious Damage Long-term
The Role of Botoh in the Village
Head Elections
The cases are available on-line at www.conflictanddevelopment.org

6 East Java Ponorogo Sampung When Brothers are No Longer Serious Damage Short Term
Annex D: List of Conflict Pathways Case Studies
Friendly: A Fight Between the SH

Terate and SH Winongo Martial Arts
Schools in Dusun Sampung Kidul,
Desa Sampung
7 East Java Ponorogo Sampung Mob Violence: A House is Serious Damage Short Term
Vandalized During a Campursari
performance to the West of
Sampung Market
8 East Java Ponorogo Sampung Night of Rhythm, Night of Serious Damage Short Term
Destruction: Even the General
Community is caught up in the
current of the Silat Group Conflict
Annexes
197
9 East Java Ponorogo Sampung Traditional Shadow Puppet Show Serious Damage Short Term
198
Leads to Violence; the fighting
between the SH Terate and Ikatan
Kera Sakti Martial Arts groups
10 East Java Ponorogo Badegan The Cracked Social Edifice: Conflict Little or No Damage Short - Long Term
between the Government,
Contractors, and the Community in
the Development of the
Sumorobangun Dam
11 East Java Ponorogo Badegan When the Nanny No Longer Serious Damage None
Protects her Charge: Then Even
Civil Disobedience is Manifested
through Burning Down the Forest
12 East Java Ponorogo Badegan Stuck Between a Rock and a Hard Little or No Damage None
Place: The Dilemma for KDP
Facilitators in the Kucur Market for
Tourists Case
13 East Java Ponorogo Badegan Bom, Democracy a la Dayakan: Little or No Damage Short - Long Term
Tensions in the Village Head
Election Process
14 East Java Ponorogo Badegan The Pays Lousy yet Others are Little or No Damage Long-term
Never Satisfied: Conflict in the
Village Head Elections in Desa
Biting
15 East Java Ponorogo Badegan Joget: It’s Blissful but can Provoke Serious Damage Short Term
Fury: Conflict between Jiujitsu and
SH Terate Martial Arts Groups
16 East Java Ponorogo Badegan Be Careful Playing With Electricity: Little or No Damage Short - Long Term
Horizontal and Vertical Conflict in
the Electricity Case in Desa Biting
17 East Java Ponorogo Jenangan The Debt is No Longer a Problem: Little or No Damage Long-term
Problem Resolution by the Women’s
Savings and Borrowing Group Desa
Kemiri
18 East Java Ponorogo Jenangan Overdue Cattle: KDP Saving- Little or No Damage None
Borrowing Groups Fail To Run
Smoothly in Desa Panjeng
19 East Java Ponorogo Slahung Joso Macadamisation: Competition Little or No Damage Short Term
Results Leave a Debt of Envy
20 East Java Ponorogo Slahung The Eviction of Plan International Serious Damage Long-term
from Desa Wates: Portrait of the
Domination of the Old Leaders
21 East Java Ponorogo Slahung A Dividing Bridge: KDP Conflict Little or No Damage Long-term
within the Realm of Local Conflict in
Slahung Village
22 East Java Ponorogo Slahung Lucky Start, Unlucky End: The Little or No Damage Short-term
Difficult Settlement of KDP Cattle
Debts in Desa Ngloning
23 East Java Pamekasan Palengaan The High Cost of Sensitivity: A Case Deaths Long-term
Study about Mass Carol Combat at
Aengnyonok Market, Desa
Banyupelle
24 East Java Pamekasan Palengaan An Engagement is Broken Off, a Serious Damage Short - Long Term
Sickle Speaks Out: Carok Due to
Jealousy in Palengaan Laok Village
25 East Java Pamekasan Palengaan Black Magic - A Business Serious Damage Short - Long Term
Opportunity for Bajingan
26 East Java Pamekasan Palengaan Brudin ’The Razor Blade’ Burnt Deaths None
Alive: A Case Study about a Land
Dispute in Dusun Tenggina Dua,
and Desa Palengaan Daja
27 East Java Pamekasan Palengaan An Inheritance Brings Misfortune: A Little or No Damage Short - Long Term
Case Study about a Land Dispute in
Dusun Angsoka Timur, Desa
Palengaan Daja
Annexes
199
28 East Java Pamekasan Palengaan A Rift between Neighbors over a Little or No Damage Long-term
200
Hand-span of Land:Land Dispute
Case Study in Dusun Angsoka
Timur, Desa Palengaan Daja
29 East Java Pamekasan Palengaan My Old Friend has Forgotten Serious Damage Short Term
Himself. When the Kyai Call
Something Red, Then All Madura is
Red!: A case study of the
Jagalan/Abattoir Conflict in Desa
Banyupelle
30 East Java Pamekasan Palengaan Even Agile Squirrels Fall, Let Alone Deaths Long-term
Humans: A Case Study of Mob
Rage Towards a Thief in Dusun
Nagasari, Desa Palengaan Laok
31 East Java Pamekasan Proppo Fresh Water These Days is No Little or No Damage None - Short Term
Longer Refreshing: A Case Study
about a Struggle for Management

Rights of a Bor in Desa Tattangoh,
Kecamatan Proppo
32 East Java Pamekasan Proppo Rumors, Women, and Carok: A Deaths Long-term
Case Study of Carok in the Rice
Fields, Desa Campor, Kec. Proppo
33 East Java Pamekasan Proppo Invulnerability Dissipates with Serious Damage Long-term
Arrogance: A Case Study of Carok
between Families in Desa Mapper,
Kecamatan Proppo
34 East Java Pamekasan Proppo The Fate of KDP at the Centre of a Little or No Damage None
Power Struggle: KDP in Desa
Panagguan, Kec. Proppo
35 East Java Pamekasan Proppo When Ants Fight over the KDP Little or No Damage Long-term
Sugar: A Case Study of Conflict in
the Implementation of KDP in Desa
Tattangoh
36 East Java Pamekasan Proppo The Dang Lebar Land Dispute: A Little or No Damage Long-term
Case Study of a Land Conflict in
Dusun Dang Lebar, Desa
Panagguan
37 East Java Pamekasan Proppo Sacrifice Everything for Victory: A Serious Damage None
Case Study of Conflict in the Village
Head Election in Desa Panagguan,
Kecamatan Proppo
38 East Java Pamekasan Proppo Mending A Shattered Bottle: Is it Little or No Damage Short-term
Possible? A Case Study of the
Conflict in the Village Head Election
in Desa Proppo
39 East Java Pamekasan Proppo Disposed of Due to Black Magic: A Serious Damage Short-term
Case Study about Conflict due to
Accusations of Black Magic in Desa
Mapper, Kecamatan Proppo
40 East Java Pamekasan Pademawu Apparently the Estate Had Been Little or No Damage Short-term
Sold
41 East Java Pamekasan Pademawu Even the NMC is Unable to Resolve it Little or No Damage None
(KDP Case Study: Refusal of UPK
Authority over Revolving Funds
Saving-borrowing in Kel. Lawangan
Daya, Kec. Pademawu,
Pamekasan)
42 East Java Pamekasan Pasean When the Village Head is Little or No Damage Long-term
Undermined. (KDP Case Study: The
Village Head Refuses to Release
the KDP funds in Tlontoraja village,
Kec. Pademawu, Pamekasan).
Annexes
201
202
Annexes
203
204
Annexes
Annex E: Summary Statistics from Key Informant Survey
Table A11: Key Informant Survey Respondents (treatment sites)

Level KDP non-KDP
Kecamatan Kecamatan Facilitator (FK) Sub-district Chief (Camat)
Head of implementation team (UPK)
Village Village facilitator, male (FD) Village Head ( Kepala Desa)
Village facilitator, female (FD) Head of Village Parliament (BPD)
Member of borrowers group (female) Community Leader
Member of borrowers group (male) Midwife
Table A12: Key Informant Survey Respondents (control sites)

Level non-KDP
Village Village Head (Kepala Desa)
Head of Village Parliament (BPD)
Community Leader
Midwife
Activist
Woman’s Association (PKK) Head
Entrepreneur
Female Teacher
Table A13: Characteristics of Key Informants

East Java NTT
Treatment Sites Control Sites Treatment Sites Control Sites
n = 132 % n = 48 % n = 134 % n = 48 %
Age
- <= 30 years old 17 13.1% 7 14.6% 18 13.6% 7 14.6%
- 31-40 years old 61 46.9% 20 41.7 47 35.6% 20 41.7%
- 41-50 years old 27 20.8% 11 22.9 40 30.3% 12 24.0%
- >= 51 years old 25 19.2% 10 20.8% 27 20.50% 9 19.8%
Gender
- Male 86 65.2% 33 68.8% 92 68.7% 29 60.4%
- Female 46 34.8% 15 31.3% 42 31.3% 19 39.6%
Level
- Village 111 84.1% 48 100% 115 85.8% 48 100%
- Kecamatan 21 15.9% - - 19 14.2% - -
Informant Type
- KDP 69 52.3% 15 31.3% 64 47.8% 15 31.3%
- Non-KDP 63 47.7% 33 68.8% 70 52.2% 33 68.8%
Migrant?
- No 90 69.8% 47 97.9% 86 65.6% 48 100.0%
- Yes 39 30.2% 1 2.1% 45 34.4% 0 0.0%
Notes: (1) ‘Level’ is the level at which the informant was interviewed—i.e., in the kecamatan capital, or in the village; (2) ‘Migrant?’: Non-
migrants are those who report having lived all their life in the village in which they now reside; (3) There are four missing c ases for informant’s
age, and six missing cases for whether or not they are a migrant in the treatment site data.
205
Annex F: Summary Statistics from KDP & CCN Dataset
A dataset was put together of all reported incidents of conflict between 2001 and 2003 in the research
areas and surrounding districts. The dataset allows us to look at patterns of conflict by geographic
locality.
1. Actor Combinations in Conflicts in East Java and NTT
Table A14 shows that, in East Java, the largest proportion of conflicts resulting in a death are from
disputes involving two individuals. However in NTT, fatal conflicts between two groups are more
common.
Table A14: Actor Combinations in Conflicts

Resulting in Death (2001-2003)
# deaths %
EAST JAVA
Individual vs. individual 75 47.5
Individual vs. group 48 30.4
Group vs. group 19 12.0
Government vs. government 3 1.9
Individual vs. government 1 0.6
Group vs. government 1 0.6
Other 11 7.0
Total 158 100
NTT
Individual vs. individual 44 37.6
Individual vs. group 15 12.8
Group vs. group 45 38.5
Government vs. government 1 0.9
Individual vs. government 2 1.7
Group vs. government 10 8.5
Other 0 0.0
Total 117 100
206
Annexes
2. Impacts of Conflict in East Java and NTT
Tables A15 and A16 report the physical impacts by district for East Java and NTT, respectively.
Table A15: Conflict Impacts by District, East Java (2001-2003)

District # deaths # injuries # properties damaged
Bangkalan 45 80 3
Pamekasan 17 18 3
Sampang 38 33 5
Sumenep 34 19 19
Pamekasan Cluster Total 134 150 30
Madiun 11 114 146
Magetan 9 21 89
Ponorogo 4 69 7
Ponorogo Cluster Total 24 204 242
TOTAL 158 354 272
Table A16 : Conflict Impacts by District, NTT (2001-2003)

District # deaths # injuries # properties damaged
Manggarai 52 78 86
Flores Timur 22 101 27
Ende 15 49 24
Sikka 21 40 24
Ngada 7 45 21
TOTAL 117 313 182
207
3. Impacts of Conflict in Qualitative Research Areas
Tables A17 and A18 report the prevalence and impacts of conflict in the research sub-districts and
villages, respectively.
Table A17: Levels of Impacts of Conflict in Research Sub-Districts (2001-2003)
# Cases with # #
# Violent # Cases # Cases # People
No Districts Sub-district Type # Cases Damaged People Properties
Cases with Death with Injury Killed
Property Injured Damaged
1 Pamekasan Pademawu Treatment 13 1 1 1
2 Pamekasan Palengaan Control 4 2 2 1 2 1
3 Pamekasan Pasean Treatment 3
4 Pamekasan Proppo Treatment 3 2 2 2
5 Ponorogo Badegan Treatment 1
6 Ponorogo Jenangan Treatment 1
7 Ponorogo Sampung Control 7 3 3 1 6 1
8 Ponorogo Slahung Treatment 4 1 1 2
Total East Java 36 9 6 4 1 7 7 1
1 Manggarai Kota Komba Control 11 5 2 4 3 5

2 Manggarai Lambaleda Treatment 3 2 2 2
3 Manggarai Ruteng Treatment 23 10 2 7 3 3 9 4
4 Manggarai Cibal Treatment 3
5 Sikka Maumere Control 13 6 2 2 2 4 2 3
6 Sikka Nita Treatment 7 3 1 3 1 3
7 Sikka Paga Treatment 3 2 2 1 1 2 2 3
8 Sikka Talibura Treatment 10 5 3 3 3 4
Total NTT 73 33 12 22 6 16 27 10
TOTAL 109 42 18 26 7 23 34 11
Table A18: Levels and Impacts of Conflict in Research Villages (2001-2003)
# Cases # Cases # Cases with # #
District Sub-district Village Type # Violent # Properties
No # Cases with with Damaged People People
Cases Damaged
Death Injury Property Killed Injured
1 Pamekasan Pademawu Sumedangan Treatment
2 Pamekasan Pademawu Padelegan Treatment 2 1 1 1
3 Pamekasan Palengaan Banyupele Control 2 1 1 1 1 1
4 Pamekasan Palengaan Palengaan Laok Control 1
5 Pamekasan Palengaan Palengaan Daja Control
6 Pamekasan Pasean Tlontoh Raja Treatment
7 Pamekasan Pasean Sanadaya Treatment
8 Pamekasan Proppo Proppo Treatment
9 Pamekasan Proppo Panangguan Treatment
10 Pamekasan Proppo Tattanggoh Treatment
11 Ponorogo Jenangan Kemiri Treatment
12 Ponorogo Jenangan Panjeng Treatment
13 Ponorogo Slahung Wates Treatment
14 Ponorogo Slahung Slahung Treatment
15 Ponorogo Badegan Badegan Treatment
16 Ponorogo Badegan Dayakan Treatment
17 Ponorogo Badegan Biting Treatment
18 Ponorogo Sampung Pagerukir Control
19 Ponorogo Sampung Gelangkulon Control
20 Ponorogo Sampung Sampung Control
East Java Total 5 2 2 1 0 2 1 0
1 Manggarai Ruteng Wae Belang Treatment

2 Manggarai Ruteng Cumbi Treatment
3 Manggarai Lambaleda Ten gkuleda Treatment
4 Manggarai Lambaleda Satar Punda Treatment
5 Manggarai Lambaleda Golo Manggung Treatment
6 Manggarai Cibal Pagal Treatment 1
7 Manggarai Cibal Lando Treatment
8 Manggarai Kota Komba Golo Meni Control 1
9 Manggarai Kota Komba Tanah Rata Control 2 1 1 1
10 Manggarai Kota Komba Watu Nggene Control 2
11 Sikka Talibura Werang Treatment
12 Sikka Talibura Nebe Treatment
13 Sikka Maumere Watugong Control
14 Sikka Maumere Koting A Control
15 Sikka Maumere Nele Wutung Control 8 4 2 1 1 4 1 1
16 Sikka Nita Nita Treatment 2 1 1 1
17 Sikka Nita Bloro Treatment
18 Sikka Nita Magepanda Treatment
19 Sikka Paga Paga Treatment 1 1 1 1 1 1 2 3
20 Sikka Paga Loke Treatment
NTT Total 17 7 3 4 2 5 5 4
TOTAL 22 9 5 5 2 7 6 4
208
References
Abidin, Zainal (2005). Penghakiman Massa: Studi Psikologi Sosial Tentang Kekerasan Kolektif
Terhadap Orang-Orang Yang Dipersepsi Sebagai Pelaku Kejahatan. Jakarta: Gramedia.
Addison, Tony (1998). Rebuilding Post-Conflict Africa: Reconstruction and Reform. Helsinki:
WIDER/United Nations University.
Alatas, Vivi (2005). “Economic Analysis of KDP Infrastructure.” Mimeo. World Bank Office Jakarta.
Alsop, Ruth and Nina Heinsohn (2005). “Measuring Empowerment in Practice: Structuring Analysis and
Framing Indicators.” Policy Research Working Paper No. 3510. Washington, DC: World Bank.
Anderson, Benedict (1991). Imagined Communities: Reflections on the Origins and Spread of
Nationalism (revised edition). London: Verso.
Anderson, Mary B. (1999). Do No Harm: How Aid Can Support Peace – Or War. Boulder, CO:
Lynne Rienner.
Anggraini, Novia Cici (2004). “The Village Representative Board (BPD) in Kecamatan Badegan
and Sampung, Kabupaten Ponorogo.” Mimeo. World Bank Office Jakarta.
Anggraini, Novia Cici, and Imron Rasyid (2004). “KDP Implementation and Dreams of Social
Change.” Mimeo. World Bank Office Jakarta.
Antlov, Hans (2001). “Village Governance and Local Politics in Indonesia.” Paper presented at session
on Decentralization and Democratization in Southeast Asia at SOAS, London, 6-9 September.
Antlov, Hans (2003). “Village Government and Rural Development in Indonesia: The New
Democratic Framework.” Bulletin of Indonesian Economic Studies 39(2): 193-214.
Appadurai, Arjun (2004). “The capacity to aspire: culture and the terms of recognition?”, in
Vijayendra Rao and Michael Walton (eds.) Culture and Public Action, Stanford, CA: Stanford
University Press, pp. 59-84.
Ashari, Luthfi (2003). “Power Sharing Between the Kyai, Bajingan and the Village Head.” Mimeo.
World Bank Office Jakarta.
Ashari, Luthfi (2005). “The Dynamics of District Governance: Forums, Budgetary Processes and
Transparency: Case Studies of Bangkalan and Poso Districts.” Indonesian Social
Development Paper No. 1. Jakarta: World Bank.
Ashari, Luthfi and Alexa Hergesell (2006). “Youth and Reconciliation: A Study of Youth Initiatives
for Peace and Reconciliation in Poso, Central Sulawesi.” Indonesian Social Development
Paper No. 9. Jakarta: World Bank.
209
Aspinall, Edward and Greg Fealy (eds.) (2003). Local Power and Politics in Indonesia:
Decentralisation and Democratisation. Singapore: Institute of Southeast Asian Studies.
Baare, Anton (2004). “Policing and Local Level Conflict Management in Resource Constrained
Environments: Case Studies in Kabupaten Sikka (NTT) and Kabupaten Ponorogo (East
Java), Indonesia.” Jakarta: NCG/World Bank. Mimeo. Available on-line at:
www.conflictanddevelopment.org
Badan Pusat Statistik (2002a). Jawa Timur Dalam Anka. Jawa Timur: BPS.
Badan Pusat Statistik (2002b). Nusa Tenggara Timur Dalam Anka. Nusa Tenggara Timur: BPS.
Badan Pusat Statistik (2002c). Kabupaten Ponorogo Dalam Anka 2001. Ponorogo: BPS.
Baker, Judy (2000). Evaluating the Impact of Development Projects on Poverty: A Handbook for
Practitioners. Washington, DC: World Bank.
Barnawi, Saifullah (2003). “KDP under the Hegemony of the Klebun: A General Picture of the
Implementation of KDP in Kec. Proppo, Pamekasan.” Mimeo. World Bank Office Jakarta.
Barron, Patrick, Samuel Clark, and Muslahuddin Daud (2005). Conflict and Recovery in Aceh: An
Assessment of Conflict Dynamics and Options for Support to the Peace Process. Jakarta:
World Bank.
Barron, Patrick, Samuel Clark, and Ambar Mawardi (2004). “The Links Between KDP and Local
Conflict: Results from a Key Informant Survey in East Java and NTT.” Mimeo. Revised
version. World Bank Office Jakarta.
Barron, Patrick, and Rachael Diprose (forthcoming). “Nine Cases of Development and Conflict in
Indonesia.” Jakarta: World Bank.
Barron, Patrick, Rachael Diprose, David Madden, Claire Q. Smith, and Michael Woolcock. (2004).
“Do Participatory Development Projects Help Villagers Manage Local Level Conflicts? A
Mixed Methods Approach to Assessing the Kecamatan Development Project, Indonesia.”
Conflict Prevention and Reconstruction Working Paper No. 9 (revised edition). Washington
DC: World Bank, Social Development Department.
Barron, Patrick, Melina Nathan, and Bridget Welsh (2005). “Consolidating Indonesia’s Democracy:
Conflict, Institutions and the ‘Local’ in the 2004 Legislative Elections.” Conflict Prevention
and Reconstruction Working Paper No. 31. Washington DC: World Bank, Social
Development Department.
Barron, Patrick, Rachael Diprose, and Claire Q. Smith (2004). Field Research Guides: KDP &
Community Conflict Negotiation Study. World Bank: Jakarta. Available on-line at:
www.conflictanddevelopment.org
210
References
Barron, Patrick, Rachael Diprose, Claire Q. Smith, Katherine Whiteside, and Michael Woolcock
(2004). “Applying Mixed Methods Research to Community Driven Development Projects
and Local Conflict Mediation: A Case Study from Indonesia.” Mimeo. World Bank: Jakarta.
Available on-line at: www.conflictanddevelopment.org
Barron, Patrick, Kai Kaiser, and Menno Pradhan (2004). “Local Conflict in Indonesia: Measuring
Incidence and Identifying Patterns.” Policy Research Working Paper No. 3384. Washington,
DC: World Bank.
Barron, Patrick, and David Madden (2004). “Violence and Conflict Resolution in Non-Conflict
Regions: The Case of Lampung, Indonesia.” Indonesian Social Development Paper No. 2.
Jakarta: World Bank.
Barron, Patrick, and Joanne Sharpe (2005). “Counting Conflicts: Using Newspaper Reports to
Understand Violence in Indonesia.” Conflict Prevention and Reconstruction Paper No. 25.
Washington, DC: World Bank, Social Development Department.
Barron, Patrick, Claire Q. Smith, and Michael Woolcock (2004). “Understanding Local Level Conflict
in Developing Countries: Theory, Evidence and Implications from Indonesia.” Conflict
Prevention and Reconstruction Paper No. 19. Washington DC: World Bank, Social
Development Department.
Bates, Robert H. (2000). Prosperity and Violence: The Political Economy of Development. New
York: W.W. Norton & Co.
Bertrand, Jacques (2004). Nationalism and Ethnic Violence in Indonesia. Cambridge: Cambridge
University Press.
Bjork, Christopher (2003). “Local Responses to Decentralization Policy in Indonesia.” Comparative

Education Review 47(2) (May): 184-216.
Bowen, John R. (2003). Islam, Law and Equality in Indonesia: An Anthropology of Public Reasoning.
New York: Cambridge University Press.
Brass, Paul (1997). Theft of an Idol: Text and Context in the Representation of Collective Violence.
Princeton: Princeton University Press.
Burton, John W. (1990). Conflict: Resolution and Prevention. New York: St Martin’s Press.
Chopra, Jarat and Tanya Hohe (2004). “Participatory Intervention.” Global Governance 10: 289-305.
Clark, Samuel (ed.) (2004). “More than Just Ownership: Ten Land and Natural Resource Conflict
Case Studies from East Java and NTT.” Indonesian Social Development Paper No. 4. Jakarta:
World Bank.
211
Cliffe, Sarah, Scott Guggenheim, and Markus Kostner (2003). “Community-Driven Reconstruction
as an Instrument in War-to-Peace Transitions.” Conflict Prevention and Reconstruction
Working Paper No. 7. Washington, DC: World Bank, Social Development Department.
Colombijn, Freek (2002). “Maling, Maling! The Lynching of Petty Criminals,” in Freek Colombijn,
Freek and J. Thomas Lindblad (eds.), Roots of Violence in Indonesia: Contemporary Violence
in Historical Perspective. Leiden: KITLV Press.
Colombijn, Freek (2005). “A Cultural Practice of Violence in Indonesia: Lessons from History,” in
Dewi Fortuna Anwar, Helene Bouvier, Glenn Smith, and Roger Tol (eds.), Violent Internal
Conflicts in Asia Pacific: Histories, Political Economies and Policies. Jakarta: KITLV/
Buku Obor.
Cribb, Robert (2005). “Legal Pluralism, Decentralization and the Roots of Violence in Indonesia,”
in Dewi Fortuna Anwar, Helene Bouvier, Glenn Smith, and Roger Tol Helene Bouvier,
Glenn Smith, and Roger Tol (eds.), Violent Internal Conflicts in Asia Pacific: Histories,
Political Economies and Policies. Jakarta: KITLV/Buku Obor.
Cutura, Josephina (2003). “Women in Village Governance and Problem-Solving: the Case of Sikka,
Flores.” Mimeo. World Bank Office Jakarta.
Das, Jishnu and Jeffrey Hammer (2005). “Which Doctor? Combining Vignettes and Item Response
to Measure Clinical Competence.” Journal of Development Economics 78(2): 348-383.
De Jonge, H. (1998). Madura dalam Empat Zaman: Pedagang, Perkembangan Ekonomi dan Islam.
Suatu Studi Antropologi Ekonomi. Jakarta: Gramedia.
Didakus, Stanis (2004). “Civic Interaction and Identity Groups in Sikka.” Mimeo. World Bank
Office Jakarta.
Diprose, Rachael (forthcoming). “Conflict Pathways in Indonesia: Conflict, Violence and

Development in East Java.” Mimeo. World Bank Office Jakarta.
Duflo, Esther and Michael Kremer (2005). “Use of Randomization in the Evaluation of Development
Effectiveness,” in George Pitman, Osvaldo Feinstein and Gregory Ingram (eds.), Evaluating
Development Effectiveness. New Brunwick, NJ: Transaction Publishers, pp. 205-31.
Eade, Deborah (1997). Capacity-Building: An Approach to People-Centred Development. Oxford:

Oxfam.
Easterly, William (2001). “Can Institutions Resolve Ethnic Conflict?” Economic Development and
Cultural Change 49(4): 687-706.
Ellickson, Robert C. (1991). Order Without Law: How Neighbors Settle Disputes. Cambridge,
MA: Harvard University Press.
212
References
Erb, Maribeth (1999). The Manggaraians: A Guide to Traditional Lifestyles. Singapore: Times.
Evers, Pieter (2000). “Resourceful Villagers, Powerless Communities: Rural Village Governance
in Indonesia.” Mimeo. World Bank Office Jakarta.
Fauzanafi, Muhammad Zamzam (2002). Reog: Sebuah Ritus Pemranataan (Konstruksi Tradisi
dalam Pertunjukan Reog Ponorogo). Skripsi S-1. Yogyakarta: Universitas Gadjah Mada.
Fukuyama, Francis (2004). State Building: Governance and World Order in the Twenty-First Century.
London: Profile Books.
Fung, Archon and Erik Wright (2003). Deepening Democracy: Institutional Innovations in
Empowered Participatory Governance. London: Verso.
Galtung, Johan (1969). “Violence, Peace, and Peace Research.” Journal of Peace Research 6(3):
167-191.
Geertz, Clifford (1976). The Religion of Java. Chicago and London: University of Chicago Press.
Geertz, Clifford (1987). Abangan, Santri, Priyayi. Jakarta: Pustaka Jaya.
Geertz, Clifford (1992). “Ritus dan Perubahan Sosial: Pernakaman di Jawa,” Kebudayaan dan
Agama. Yogyakarta: Kanisius.
George, Alexander and Andrew Bennett (2005). Case Studies and Theory Development in the Social
Sciences. Cambridge, MA: MIT Press.
Ghewa, Yohanes (2003). “The Role of KDP in Kabupaten Manggarai.” Mimeo. World Bank Office
Jakarta.
Gibson, Christopher and Michael Woolcock (2005). “Empowerment and Local Level Conflict
Mediation in the Kecamatan Development Project in Indonesia: Concepts, Measures, and
Project Efficacy.” Policy Research Working Paper No. 3713. Washington, DC: World Bank.
Government of Indonesia (2002). “Kecamatan Development Program: KDP in Conflict Areas.”

Jakarta: Ministry of Home Affairs, Community Development Agency (PMD) and National
Management Consultants (NMC).
Government of Indonesia (2004). “Kecamatan Development Program Phase II Results.” Jakarta:

Ministry of Home Affairs, Directorate General of Rural and Community Empowerment,
KDP National Secretariat and National Management Consultants (NMC).
Grimes, Charles E., Tom Therik, Barbara Dix Grimes, and Max Jacob (1997). A Guide to the
Peoples and Languages of Nusa Tenggara. Kupang: Artha Wacana Press/Alfa Omega
Foundation.
213
Guggenheim, Scott E. (2006). “The Kecamatan Development Program, Indonesia.” in Anthony

Bebbington, Scott E. Guggenheim, Elisabeth Olson, and Michael Woolcock (eds.) The Search
for Empowerment: Social Capital as Idea and Practice at the World Bank. Bloomfield, CT:
Kumarian Press.
Gurr, Ted (2000). People versus States: Minorities at Risk in the New Century. Washington DC:
United States Institute of Peace Press.
Gurr, Ted (2001). Peace and Conflict 2001: A Global Survey of Armed Conflicts, Self-Determination
Movements, and Democracy. Maryland: Center for International Development and Conflict
Management, University of Maryland.
Haggard, Stephan, and Robert R. Kaufman (1995). The Political Economy of Democratic Transitions.
Princeton, NJ: Princeton University Press.
Hefner, Robert W. (2000). Civil Islam: Muslims and Democratization in Indonesia. Princeton, NJ:
Princeton University Press.
Hegre, Havard, Tanja Elingsen, Scott Gates, and Nils Petter Gleditsch (2001). “Towards a Democratic
Civil Peace? Democracy, Political Change, and Civil War, 1816-1992.” American Political
Science Review 95(1): 33-41.
Heifetz, Ronald (1994). Leadership Without Easy Answers. Cambridge, MA: Harvard University Press.
Horowitz, Donald (2000). Ethnic Groups in Conflict (2nd edition). Berkeley, Los Angeles, London:
University of California Press.
International Crisis Group (2003). “Indonesia: Managing Decentralization and Conflict in Indonesia.”
44. Jakarta/Brussels: International Crisis Group.
International Crisis Group (2005). “Decentralisation and Conflict in Indonesia: The Mamasa Case.”
Asia Briefing 37. Singapore/Brussels: International Crisis Group.
KDP National Secretariat and National Management Consultants (2003). Indonesia: Kecamatan
Development Program. December 2003.
King, Charles (2004). “The Micropolitics of Social Violence.” World Politics 56(3): 431-455.
King Gary, Christopher J.L. Murray, Joshua A. Salomon, and Ajay Tandon (2004). “Enhancing the
Validity and Cross-cultural Comparability of Survey Research.” American Political Science
Review 98(1): 191-207.
King, Gary and Jonathan Wand (2004). “Comparing Incomparable Survey Responses: New Tools
for Anchoring Vignettes.” Mimeo. http://gking.harvard.edu/files/c.pdf
214
References
Kuntowijoyo (2002). Perubahan Sosial dalam Masyarakat Agraris: Madura 1850 – 1940.
Yogyakarta: Mata Bangsa.
Lawang, M.Z.R. (1996). Konflik Tanah di Manggarai, Flores Barat – Pendekatan Sosiologik. Jakarta:
Penerbit Universitas Indonesia.
Mahur, Agus (2003). “Land Conflicts and the Mechanisms Used to Resolve Them in Kabupaten
Manggarai.” World Bank Office Jakarta.
Mallaby, Sebastian (2004). The World’s Banker: A Story of Failed States, Financial Crises, and the
Wealth and Poverty of Nations. Washington, DC : Council on Foreign Relations/Penguin Press.
Mansuri, Ghazala, and Vijayendra Rao (2004). “Community Based (and Driven) Development: A
Critical Review.” World Bank Research Observer 19(1): 1-39.
McAdam, Doug, Sidney Tarrow, and Charles Tilly (2001). Dynamics of Contention. New York:
Cambridge University Press.
McCarthy, John (2004). “Changing to Gray: Decentralization and the Emergence of Volatile Socio-
Legal Configurations in Central Kalimantan, Indonesia.” World Development 32(7): 1199-
1223.
McNeil, Mary and Michael Woolcock (2004). “Capacity Enhancement and Social Development:
Building on Local Context and Process.” World Bank Institute Working Paper No. 37245.
Washington, DC: World Bank.
Meliala, Adrianus (ed.) (2005). “Evaluation of the Police Role in the 2004 General Elections”.
Jakarta: Partnership for Governance Reform in Indonesia.
Mills, C. Wright (1959). The Sociological Imagination. New York: Oxford University Press.
Moore, Barrington (1967). Social Origins of Dictatorship and Democracy: Lord and Peasant in
the Making of the Modern World. Boston: Beacon Press.
Morishita, Akiko (2005). “Today’s Local Politics in Indonesia 1998-2004: Who Controls the Political
and Economic Interests in Kalimantan?” Paper presented at the University of Indonesia 4th
International Symposium of Journal Anthropologi Indonesia, Depok, 12th-15th July.
Mosse, David (2005). Cultivating Development: An Ethnography of Aid Policy and Practice. London:
Pluto Press.
Mosse, David, Alan Rew and John Farrington (1998). Development as Process: Concepts and
Methods for Working with Complexity. London: Routledge.
215
North, Douglass (1990). Institutions, Institutional Change and Economic Performance. New York:
Cambridge University Press.
Olken, Ben (2005). “Monitoring Corruption: Evidence from a Field Experiment in Indonesia.”
NBER Working Paper No. 11753. Cambridge, MA: NBER
Olken, Ben (2006). “Corruption and the Cost of Redistribution: Micro Evidence from Indonesia”
Journal of Public Economics 90 (4-5) : 853-870.
Parry, Richard Lloyd (2005). In the Time of Madness. London: Jonathan Cape.
Polanyi, Karl (1944). The Great Transformation. Boston: Beacon Press.
Prior, John (2003). “The Church and Land Disputes: Sobering Thoughts From Flores.” Maumere:
Chandraditya Research Center for the Study of Religion and Culture. Mimeo.
Prior, John (2004). “Land, Church and State: Forestry Business Packaged in Ecological Concerns
in Flores.” Inside Indonesia 78 (April-June).
Pritchett, Lant (2002). “It Pays to be Ignorant: A Simple Political Economy of Program Evaluation.”
Journal of Policy Reform 5(4): 251-269.
Pritchett, Lant, and Michael Woolcock (2004). “Solutions when the Solution is the Problem: Arraying
the Disarray in Development.” World Development 32(2): 191-212.
Probo, Endro (2003a). “On KDP in Pamekasan.” World Bank Office Jakarta.
Probo, Endro (2003b). “The Phenomenon of the Martial Arts Schools in Kabupaten Ponorogo.”
World Bank Office Jakarta.
Rao, Vijayendra (2005). “Symbolic Public Goods and the Coordination of Collective Action: A
Comparison of Local Development in India and Indonesia.” Policy Research Working Paper
No. 3685. Washington, DC: World Bank.
Rasyid, Imron (2004). “KDP Implementation and its Spillover Effects on the Lives of the Villager
in Ponorogo.” World Bank Office Jakarta.
Rieffel, Lex (2004). “Indonesia’s Quiet Revolution.” Foreign Affairs 83(5) (July): 98-110.
Riefqi Muna (2004). “Security Reform: Reform of the Security Sector is Fundamental for Achieving
a Democratic Indonesia,” Inside Indonesia (Jan-March).
Rosenbaum, Paul R., and Donald B. Rubin (1983). “The Central Role of the Propensity Score in
Observational Studies for Causal Effects,” Biometrika (70): 41-55.
216
References
Rosenbaum, Paul R., and Donald B. Rubin (1985). “Constructing a Control Group Using
Multivariate Matched Sampling Methods that Incorporate the Propensity,” American
Statistician (39): 33-38.
Ross, Marc (1993). The Culture of Conflict: Interpretations and Interests in Comparative Perspective.
New Haven and London: Yale University Press.
Rothchild, Donald (1997). Managing Ethnic Conflict in Africa: Pressures and Incentives for
Cooperation. Washington, DC: Brookings Institution Press.
Sage, Caroline, and Michael Woolcock (2005). “Breaking Legal Inequality Traps: New Approaches
to Local Level Judicial Reform in Low Income Countries.” Background paper for the World
Development Report 2006, on Equity and Development. Washington, DC: The World Bank.
Satu, Adam, and Patrick Barron (forthcoming). “Understanding Local Conflict Pathways in Flores.”
Mimeo. World Bank Office Jakarta.
Sayogyo (1994). “Pembangunan Daerah dan Masyarakat di NTT Dalam Sekilat Sorotan Kritis” in
Sayogyo (ed.), Kemiskinan dan Pembangunan di Propinsi Nusa Tenggara Timur. Jakarta:
Yayasan Obor Indonesia.
Scott, James C. (1985). Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven:
Yale University Press.
Scott, James C. (1998). Seeing Like a State: How Certain Schemes to Improve the Human Condition
Have Failed. New Haven: Yale University Press.
Simatupang, G. R. Lono Lastoro (2002). Play and Display: An Ethnographic Study of Reyog
Ponorogo in East Java, Indonesia. Doctoral thesis, Department of Anthropology, University
of Sydney, Sydney.
SMERU (2004). Peta Kemiskinan Indonesia 2000 (Poverty Map of Indonesia 2000). CD-Rom.
www.smeru.or.id
Smith, Claire Q. (2005). “The Roots of Violence and Prospects for Reconciliation: a Case Study of
Ethnic Conflict in Central Kalimantan, Indonesia.” Conflict Prevention and Reconstruction
Paper No. 23. Washington, DC: World Bank.
Stephens, Matthew (2003). “Local-level Dispute Resolution in Post-Reformasi Indonesia: Lessons

from the Philippines.” The Australian Journal of Asian Law 5(3): 213-259.
Stewart, Frances (2005). “Policies towards Horizontal Inequalities in Post-Conflict Reconstruction”,

CRISE Working Paper No.7. Oxford: CRISE.
217
Suryadinata, Leo, Evi Nurvidya Arifin, and Ariss Ananta (2003). Indonesia’s Population: Ethnicity
and Religion in a Changing Political Landscape. Singapore: Institute of Southeast Asian
Studies.
Tadjoeddin, Mohammad Zulfan (2002). “Anatomy of Social Violence in the Context of Transition:
The Case of Indonesia’. UNSFIR Working Paper 02/01. Jakarta: UNSFIR.
Tadjoeddin, Mohammad Zulfan, Widjajanti I. Suharyo, and Satish Mishra (2001). “Regional
Disparity and Vertical Conflict in Indonesia.” Journal of the Asia Pacific Economy 6(3):
283-304.
Tajima, Yuhki (2004). “Mobilizing for Violence: the Escalation and Limitation of Identity Conflicts,
The Case of Lampung, Indonesia.” Indonesian Social Development Paper No. 3. Jakarta:
World Bank.
Tendler, Judith (2000). “Why are Social Funds so Popular?” in Shahid Yusuf, Weiping Wu, and
Simon Evenett (eds.) Local Dynamics in an Era of Globalization. New York: Oxford
University Press, pp. 114-29.
Toda, Dami (1999). Manggarai Mencari Pencerahan Historiografi. Ende: Nusa Indah.
UNDP (2002). United Nations Human Development Report 2002: Deepening Democracy in a
Fragmented World. Oxford and New York, NY: UNDP/Oxford University Press.
UNDP (2004). “Indonesia: Decentralized Conflict Sensitive Planning and Development: an

Assessment and Project Formulation Report.” Mimeo. Jakarta: UNDP.
Uvin, Peter (1998). Aiding Violence: The Development Enterprise in Rwanda. West Hartford, CT:
Kumarian Press.
Uvin, Peter (2002). “The Development/Peacebuilding Nexus: A Typology and History of Changing
Paradigms” Journal of Peacebuilding and Development 1(1): 5-24.
Varshney, Ashutosh (2002). Ethnic Conflict and Civic Life: Hindus and Muslims in India. New
Haven: Yale University Press.
Varshney, Ashutosh, Rizal Panggabean, and Mohammad Zulfan Tadjoeddin (2004). “Patterns of
Collective Violence in Indonesia (1990-2003)”. UNSFIR Working Paper 04/03. Jakarta:
UNSFIR.
Verheijen, J.A.J. (1991). Manggarai dan Wujud Tertinggi. Jakarta: LIPI.
Welsh, Bridget (2003). “Street Justice: Vigilantism in West Java.” Paper prepared for the Workshop
on Contentious Politics, Columbia University, February.
218
References
Whiteside, Katherine, Michael Woolcock, and Xavier Briggs (2005). “Assessing the Impact of
Social Development Projects: Integrating the Science of Evaluation and the Art of Practice.”
Mimeo. Washington, DC: World Bank.
Wiyata, A. Latief (2002). Carok: Konflik Kekerasan dan Harga Diri Orang Madura. Yogyakarta:
LKiS.
Wolf, Eric (1964). Anthropology. Englewood Cliffs, NJ: Prentice Hall.
Wolf, Eric (1969). Peasant Wars of the Twentieth Century. New York: Harper & Row.
Wolf, Eric (1999). Envisioning Power: Ideologies of Dominance and Crisis. Berkeley and Los
Angeles: University of California Press.
Wong, Susan (2003). “Indonesia’s Kecamatan Development Program: Building a Monitoring and
Evaluation System for a Large-Scale Community-Driven Development Program.” Discussion
Paper. Environment and Social Development Unit. East Asia and Pacific Region. Washington,
DC: World Bank.
Woodhouse, Andrea (2005). “Village Corruption in Indonesia: Fighting Corruption in the World
Bank’s Kecamatan Development Program.” Indonesian Social Development Paper No. 6.
Jakarta: World Bank.
Woolcock, Michael (1998). “Social Capital and Economic Development: Toward a Theoretical
Synthesis and Policy Framework.” Theory and Society 27(2): 151-208.
World Bank (2002). “Back to Office Report: KDP Supervision Mission and Assessment of Social
Service Delivery.” Mimeo. World Bank Office Jakarta.
World Bank (2003). World Development Report 2004: Making Services Work for Poor People New
York: Oxford University Press.
World Bank (2004). Village Justice in Indonesia: Case Studies on Access to Justice, Village
Democracy and Governance. Jakarta: World Bank.
219
Indonesian Social Development Papers
No. Title Language Author(s) Date
1 The Dynamics of District Governance: Forums, English Luthfi Ashari May 2004
Budget Processes and Transparency
Dynamika Pemerintahan Kabupaten: Forum, Indonesian
Perencanaan Anggaran dan Transparenci
2 Violence and Conflict Resolution in Non-Conflict English Patrick Barron Aug 2004
Regions: The Case of Lampung, Indonesia David Madden
3 Mobilizing for Violence: The Escalation and English Yuhki Tajima Aug 2004
Limitation of Identity Conflicts
4 More Than Just Ownership: Ten Land and English Samuel Clark (ed.) Dec 2004
Natural Resource Conflict Case Studies
from East Java and Flores
Bukan Sekedar Persoalan Kepemilikan: Indonesian
Sepuluh Studi Kasus Konflik Tanah dan Sunber
Daya Alam dari Jawa Timur dan Flores
5 Crisis, Social Ties, and Household Welfare: English Anna Wetterberg Apr 2005
Testing Social Capital Theory with Evidence
From Indonesia
6 Village Corruption in Indonesia: Fighting English Andrea Woodhouse Apr 2005

Corruption in Indonesia’s Kecamatan
Development Program
7 Counting Conflicts: Using Newspaper Reports English Patrick Barron May 2005
to Understand Violence in Indonesia Joanne Sharpe
8 Aceh: Reconstruction in a Conflict Environment English Adam Burke Oct 2005

Afnan
9 Youth and Reconciliation: A Study of Youth English Luthfi Ashari July 2006
Initiatives for Peace and Reconciliation in Poso, Alexandra Hergesell
Central Sulawesi
10 Local Conflict and Community Development in English Patrick Barron July 2006
Indonesia: Assessing the Impacts of the Rachael Diprose
Development Program Michael Woolcock
Papers can be accessed on-line at: www.conflictanddevelopment.org

Buku Local Conflict and Community Development in Indonesia

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Buku Local Conflict and Community Development in Indonesia

Uploaded by

Copyright:

Available Formats

Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized

Luthfi Ashari and Patrick Barron (series editors)

Assessing the Impact of the Kecamatan

Indonesian Social Development Paper No. 10

Email addresses for correspondence:

Copies of this paper are available from:

Decentralization Support Facility

Photographs by Poriaman Sittangang

In East Java: In NTT:

All of them demonstrated patience and perseverance beyond the call

James Scott, Weapons of the Weak 1

1. Introduction and Context

There are two primary questions we seek to answer in this paper:

These questions give rise to a related set of secondary questions:

using propensity scoring to match on ‘observable’ variables and qualitative verification to

4. The Impact of KDP on Local Conflict and Conflict Management

Development projects and conflict inevitably go hand in hand, because development is an

KDP and Other Development Projects as Sources of Conflict

Direct Impacts of KDP on Local Conflict and Conflict Management

Indirect Impacts of KDP on Local Conflict and Conflict Management

Second, KDP also appears to contribute positively to the reconfiguration of citizen-state

Pre-existing Local Capacity to Manage Conflict, Program Functionality and KDP

The Development-Conflict Nexus

(a) Conceptualizing conflict and development: the political dimensions of development

(b) Understanding local context: prioritizing and investing in local knowledge

Ensuring Projects ‘Do No Harm’

(a) Involve communities at every stage

(b) Ensure programs work as intended: socialization, monitoring and follow-up

(c) Enhance internal complaints mechanisms and avenues of redress

seeks to establish more inclusive forms of decision-making, it should be expected (and it

(a) Design complementary programs/strategies to constructively address the drivers of conflict

(b) Improving inter-group relations: the impact of KDP on social structure

(c) Normative and behavioral changes

Evaluation of Community Development Projects

EXECUTIVE SUMMARY VII

TABLE OF CONTENTS XXI

CHAPTER 1. INTRODUCTION: UNDERSTANDING THE CONFLICT-

1. Development Projects and Conflict: Part of the Problem, or Part of a Solution? 1

2. Research Questions and Paper Structure 3

4. Making the Case 5

CHAPTER 2. ASSESSING KDP AND LOCAL CONFLICT: ANALYTICAL AND

1. The Kecamatan Development Program (KDP) and Local Conflict 7

2. Analytical Framework: How Might KDP Influence Local Conflict? 9

3. Methodology and Research Locations 13

CHAPTER 3. DEVELOPMENT PROJECTS AS ‘PROBLEM’: KDP AND OTHER

1. The Prevalence and Impact of KDP-related Conflicts 21

2. Forms of KDP-related Conflict 23

3. Why Do KDP Conflicts Not Escalate? 46

4. Comparison of KDP and Other Development-Related Conflicts 56

CHAPTER 4. DEVELOPMENT PROJECTS AS PART OF ‘THE SOLUTION’? THE IM-

1. Forms of Conflict in East Java and NTT 61

2. Conflict in KDP and non-KDP Areas 64

CHAPTER 5. THE DIRECT IMPACTS OF KDP ON CONFLICT MANAGEMENT: THE

1. The Use of KDP Forums for Managing Non-Project Conflict 71

2. The Role of KDP Facilitators in Conflict Management 76

2.2 Willingness to Act to Seek a Solution 79

CHAPTER 6. INDIRECT IMPACTS OF KDP ON CONFLICT MANAGEMENT:

3. Changes in Norms and Perceptions 106

4. The Impact on Conflict Outcomes 111

CHAPTER 7. EXPLAINING VARIATION: THE INFLUENCE OF CONTEXT ON

1. Explaining Variation 121

2. KDP-Triggered Conflict and Local Context 123

2.1 The Effect of Local Capacity on KDP-Triggered Conflict 123

4. Indirect Impacts and Local Context 137