Professional Documents
Culture Documents
Harris 3
Harris 3
net/publication/286061234
CITATIONS READS
12 212
5 authors, including:
Akosua Darkwah
University of Ghana
21 PUBLICATIONS 129 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Leila M. Harris on 07 December 2015.
TABLE OF CONTENTS
Introduction ...................................................................................................................... 1
Understandings of gender and water from a feminist political ecological perspective . 2
Context: Details on methods and research sites in Accra and Cape Town ..................... 5
Contextual differences: Water access in Cape Town and Accra .................................. 5
Gender differences with respect to water access, uses, knowledge, governance, and
experiences ...................................................................................................................... 8
Access to and uses of water ....................................................................................... 10
Knowledges ................................................................................................................ 13
Participation and governance ..................................................................................... 15
Everyday lived dimensions and emotional lives related to water ............................... 16
Discussion: Relating back and conclusions ................................................................... 20
References ..................................................................................................................... 24
Appendix .......................................................................... Error! Bookmark not defined.
List of Tables
Table 1. Overview of questions of interest for the analysis.
Table 2. Variables used in intersectional analysis
Table 3. Women and men's participation in community governance
Table A-1. Country and within country community test results
Table A-2. Cross tab results for question F.10. I know about water related meetings in
my community.
Table A-3. Cross tab results for question D.18. I worry about lack of water.
Table A-4. Cross tab results for question E.1. I am satisfied with the water services in
my community.
List of Figures
Figure 1. Primary sources of household water in Ghana and South Africa
Figure 2. The gender of primary water fetchers reported by female and male
respondents
Figure 3. Time spent fetching water
ii
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
iii
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
INTRODUCTION
In underserved settlements of Accra, Ghana and Cape Town, South Africa, men and women
both work to negotiate access to water on a daily basis. In a community such as Teshie in
Accra, residents might travel several minutes carrying heavy water buckets from a nearby
vendor, perhaps making several trips in a day to meet household needs. In South Africa, in a
community such as Khayletisha, some residents have in-home access through a tap in newly
built Reconstruction and Development Program (RDP) homes1, while others might walk up to
50 meters to a communal standpipe to fill buckets as needed. The particular experiences of
women are frequently highlighted as central for questions of water access and conditions,
with the linked suggestion that women are likely to be among the most vulnerable to access
challenges or fluctuations in water quality or quantity. The focus on women’s specific
experiences has also been highlighted in policy discourses, for instance, with the third Dublin
principle stating that ‘women play a central role in the provision, management, and
safeguarding of water’ (Dublin Principles 1992). As such, work of the past several decades
has highlighted the need to explicitly interrogate the gendered aspects of water access, uses,
and conditions. Feminist political ecologists, in particular, have contributed to our
understanding of gender-water linkages in diverse contexts, with particular focus on gender-
differentiated:
1) Access and uses of water
2) Knowledges of water and water related institutions
3) Participation in governance
4) Lived experiences and emotional dimensions of water use, access, and governance
Here we report on an analysis of a 487 household survey conducted in 2012 in four
underserved and relatively impoverished settlements, two each in Cape Town, South Africa
and Accra, Ghana. We speak to these four focal themes, while using this analysis as a
starting point to consider broader gender and water, and linked epistemological and
methodological debates. While our emphasis here is on the quantitative analysis of everyday
water access and experience in relation to the four focal themes, the work is part of a
broader research project that also includes qualitative analysis of water access, governance,
and narratives in Accra and Cape Town.2
Directing our analysis of the survey data to the four focal themes noted above, this article has
several linked goals. First, we statistically analyze the differences in water related access and
experiences as reported by female and male survey respondents to consider the evidence
for, and characterization of, these differences across our study sites . Second, doing so
allows us to contribute to the broader gender and water literature, and debates in feminist
political ecology, through a quantitative multi-sited analysis—adding to discussions that have,
to date, drawn heavily from single sited, in-depth, qualitative case-study research (see
Agarwal, 2010, as a key exception, among others). Among other insights, our results
1
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
foreground key differences across the study sites, underscoring the importance of context
specificity for analysis of complex gender related negotiations and realities. Third, and finally,
we aim to contribute to broader discussions about method, epistemology and ways of
knowing by drawing attention to tensions between empirical evidence and theory, as well as
tensions that arise when quantitative analyses do– or do not – validate insights derived from
qualitative studies. To this last point, we find that even as many of the expected gender
differences are not evidenced as statistically significant, we do not consider that this negates
the broader themes highlighted in the literature (as discussed below). We take these points
to argue for triangulation, not necessarily to validate particular ‘truths’, but to more carefully
examine the tensions between different methodologies and to highlight, and explicitly
theorize, the types of information that are made visible (or remain hidden) through different
approaches (cf. Nightingale 2009, 2003). Before presenting our results, we first provide more
background from the literature on gender and water in general, and feminist political ecology
in particular, to better understand the claims common to this body of work, and to
substantiate the key questions that guided the analysis. This is followed by a characterization
of the water situation in the study sites, to contextualize results and provide necessary
background for the analysis that follows.
Understandings of gender and water from a feminist political ecological
perspective
Our rationale for interrogating gender differences with respect to water access, uses,
knowledges, governance, and experiences stems from a large and growing literature on
gender and water, much of it informed by feminist political ecology (e.g. Bennett, Davila-
Poblete & Rico 2005, special issue of Gender, Place, and Culture [2009] or recent special
issue in the Feminist Review [2013], Buechler & Hanson 2015). While the case studies and
theoretical approaches that comprise this literature are diverse, several frequently cited
insights include: (1) That women are often primarily responsible for water provision,
especially for domestic needs (e.g., cleaning, cooking), and that this results in differentiated
uses of water, as well as gender differentiated effects when water quality or access is
compromised; (2) Linked to gender differentiated uses, and labour practices, women and
men often have differentiated knowledges of water and other environmental resources (e.g.
men may have certain knowledges in terms of water for irrigation or other productive uses,
while women may be more focused on quality and quantity considerations for domestic uses,
although this should not be assumed from the outset (Barnes 2013). (3) Men and women
participate differently, and often unequally, in water governance institutions (e.g. women
often participate less than men in water-related institutions, and even when women do
participate, the character of the participation is often qualitatively different), with potential
implications for all of the above (e.g. uses, access, and/or knowledges). Finally, (4), the
experiences and emotional lives of water are themes of growing interest, with several recent
studies showing that women may face particular stress related to compromised water quality
or changing environmental conditions (e.g. Wutich & Ragsdale 2008). We provide further
detail on these four focal themes as they are commonly discussed in the literature,
2
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
particularly as we use these insights to guide our analysis of the survey data, as reported
below.
On the issue of water fetching and gender differentiated labours and responsibilities, cross-
national analyses and multiple case studies have argued that women are most often
responsible for water procurement for domestic uses (bathing, cooking, cleaning, drinking),
and that women in particular often spend significant time on this task (Singh, Jacks,
Bhattacharya, & Gustafsson 2006, Galvin 2011). Cross-national analyses have also
suggested that these patterns are more pronounced when there is less access to improved
water sources (Sorenson, Morssink, & Campus 2011). This recognition has led water policy
communities to focus on women as primary purveyors of water (Dublin Principles1992), and
to recognize that adverse water conditions, such as water scarcity or drought, may
particularly affect women and girls (in terms of work burdens or responsibilities in caring for
sick family members who may be suffering from waterborne illness (e.g. Alston 2006, Arku &
Arku 2010). An important cross-national study of 160 countries also demonstrates that there
are gender differences in terms of morbidity and mortality associated with natural disasters
(including floods and droughts). In brief, women and girls are more likely to die following
disaster events—a pattern that was more pronounced when disaster events are larger, and
in contexts where gender inequalities are greater (Neumayer & Plumper 2007).
With respect to knowledges, a number of studies have considered the ways that water
related knowledge and expertise often becomes codified as masculine, or is connected to
particular practices. For instance, in a recent study from Nepal, Zwarteveen and Liebrand
(2015) highlight the embodied gendered performances that situate scientifically trained male
(and often foreign) engineers as ‘experts’ and local people, especially women, as
unknowledgeable about agricultural and hydrologic realities. In other studies, women
demonstrated more interest in, and knowledge of, water related budgeting and
economization compared with male counterparts ([Wutich, 2009], based on work in Bolivia).
Concern with knowledges is also intrinsically linked with participation. Work by Harris (2005)
in southeastern Turkey shows that women are often marginalized from water user groups in
newly irrigated areas, a situation that leaves women without access to the training,
resources, and networking possibilities available through those institutions. As such, any
differences in terms of expertise and knowledges may be further entrenched through
differential participation in water management institutions. It is worth noting that other studies
have also documented enhanced participation of women in civic life linked to their initial
engagement in water related associations, boosting their confidence, and enhancing
community support for their participation (e.g. Aguilar [2005] for an example from Costa Rica,
and Ennis-McMillan [2005] for an example from Mexico).
Highlighting lived experiences of water uses and governance, recent work suggests that
stress, worry and other emotions can be strongly linked to differentiated water access and
uses. Sultana (2011), for instance, examines the arsenic crisis in Bangladesh, considering
the physical and emotional toll the crisis takes on women’s lives, particularly when they have
to rely on power-laden and tenuous social networks to secure access to safe water. Wutich
3
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
and Ragsdale (2008), provide a detailed examination of water insecurity in the urban context
of Cochabamba, Bolivia—finding that being a woman is a primary predictor of water-related
stress. In this context they interpret the stress as linked to inequities in water access and
distribution, rather than water scarcity per se. Wutich (2009) also highlights self-esteem and
acceptance issues when water is less available, and women are unable to conform to social
expectations in terms of hygiene and beauty, while arguments put forward by Goldin (2013,
2010) suggest in parallel that shame and self-esteem are elements important for tracking
processes of inclusion and exclusion from decision making processes. As such, water (and
sanitation) access and conditions are understood as linked to worry, embarrassment, or a
range of other affective responses that characterize differential lived and emotional
experiences of water use and access, notably for men and women, but also likely tracking
along class, caste, or other intersectional differences (Hawkins & Ojeda 2011).
The examples outlined above, as well as the broader literatures on feminist political ecology
and gender and water (Resurreccion & Elmhirst 2008, Hawkins & Ojeda 2011, special issue
of Geoforum [2011]) highlight that gendered relationships to water are not essential (linked to
male and female attributes or bodies), but rather are mediated by work responsibilities, socio-
cultural expectations, and a range of contextual factors. The themes specific to water map
against key insights and themes from work in feminist political ecology more broadly,
including work that has focused on forestry, land, or other key resources (Rocheleau,
Thomas-Slayter, and Wangari 1996, Nightingale 2009, Agarwal 2010). Although the various
studies cited draw on different contexts, they nonetheless provide a compelling rationale
related to the need to interrogate gender as a key dimension of difference that is likely to
matter for understanding water uses, knowledges, participation, and emotional lives. As
such, our expectation in undertaking the statistical analysis was that we would observe a
number of key differences between male and female respondents related to a number of the
questions from our survey (see full list of survey questions evaluated in Table 1). A key goal
was to determine if any of these relationships could be validated statistically, and whether
any patterns were consistent across our diverse study sites.
4
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
Congress) government took office and the new Constitution was adopted, apartheid era
planning and policy continues to shape access to water and sanitation – still markedly
differentiated along racial and class lines. As an attempt to redress inequalities resulting from
the past, South Africa boasts a Constitutional right to water and sanitation, even as there are
ongoing implementation challenges (Mehta 2006, Rodina 2013). Specific efforts seek to
make water more affordable and available include the Free Basic Water (FBW) policy that (in
theory) guarantees a minimal amount of water will be provided to all households regardless
of their ability to pay (currently set at 6 kiloliters per month, but with variable implementation
across the country, and without specific consideration of household needs or size). Over the
past several decades, there has been some progress in extending services to informal
settlements and former townships—including in Khayletisha and Philippi where our survey
was carried out. Despite these efforts, concern about unequal access to water, sanitation and
other basic services is frequently in the news, with a spate of ‘service delivery protests’, and
so-called ‘poo’ wars during which protestors dumped human waste in the urban core to
protest the appalling status of sanitation in township settlements. These ongoing protests are
part of the public response to continuing unevenness of services, as well as tensions that
exist between consideration of water as a public ‘right’ versus as an ‘economic good’ as part
of demand management policies—for instance, contestation related to the increasing use of
meters in Cape Town and elsewhere (Environmental Monitoring Group [EMG], 2014).
Although not detailed here, the situation with service access is changing with the ongoing
housing formalization process. In short, as residents are increasingly moved into formal RDP
homes, effectively this means moving from shacks with communal water and sanitation
access—to structures with an in home tap and toilet (Rodina 2013)6.
6
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
0%#
Other#
3%# Ghana#(n=228)#
Primary'source'of'household'water'
South#Africa#(n=208)#
14%#
Communal#tap#or#well*#
27%#
15%#
Communal#water#tank#
12%#
'
5%#
Private#water#tank#
0%#
19%#
In#house#or#yard#connec>on#
59%#
47%#
Purchase#water#from#a#vendor#
0%#
*”Tap” includes standpipes and “well” includes boreholes. The Private water tank, or in-house or in-yard
connection, generally refers to private/individual access (although it is possible that some might have given this
response for shared communal access, particularly in Cape Town). Note the heavy reliance on vendors in
Ghana, while no respondents in South Africa provided this response.
Relative to the country on the whole, the Western Cape region (where Cape Town is the
major metropolitan area) is among the better-served provinces when it comes to water
access. Currently reports state that 88% of the population has access to piped water inside
their house (our survey—focusing on underserved communities of Khayletisha and Philippi
places that number at 55% for in-yard connections, and 3% for in-house connections). In
Cape Town, 83% of our survey respondents reported having ‘easy access to water’ and only
11% said that they do not. This is in stark contrast to the low percentages in the data from
Accra—29% and 64% respectively. Also a noteworthy difference between sites –– none of
the respondents surveyed in Cape Town rely on private vendors.
While the case study sites are certainly distinct, the specific research sites in both urban
areas were purposefully selected to highlight impoverished and underserved settlements.
Given that gender differentiated patterns are often exaggerated under conditions of stress,
we might also expect that observable gender differences might be more pronounced in our
selected sites as relatively impoverished, relatively water-stressed, locales7.
7
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
Question Answer
C.3 Who primarily fetches the water for this household? Categorical
C.7.1 For which of the following activities do you use most water? Categorical
C.8.1 (GHANA) Do you know of Aqua Vitens Rand Limited (AVRL)? Binary
C.8.1 (SOUTH AFRICA) Do you know of Free Basic Water Policy? Binary
F.10.1. I know that meetings are held about water-related issues in my Binaryb
community.a
G.8.1 If you had a problem related to water supply, who would you go to? Categorical
8
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
Question Answer
F.11 I believe that meetings about things that concern our community are Likert
open and anyone can attend/participate.
G.1 I feel safe when I fetch water from a community water sources (stand Likert
pipe, or a tanker).
9
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
10
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
11
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
C.3'Who'primarily'fetches'the'water'for'this'
household?''
2%$
GHANA$
7%$
Female$
Male$(n$=$94)$ 43%$ 50%$
Male$
SOUTH$AFRICA$
We ran analyses for how water is used and whether or not respondents consider that there is
enough water. The full results are provided in the appendix. In brief, there were no significant
differences in the ways that men and women answered questions about uses of water overall
(however there were ‘nearly’ significant gender differences in South Africa). For having
‘enough’ water, again there were no significant gender differences, but it of interest to note
that in Ghana, women more often stated that they had ‘enough’ water for all four categories
(drinking, cooking, bathing, ), and there were significant gender differences for one of the
Ghanaian sites-Ashaiman, where women more frequently agreed that they had enough water
for drinking (X2 = 6.65, df = NA, p-value = 0.036), cooking (X2 = 8.51, df = NA, p-value =
0.018), and bathing (X2 = 6.58, df = 2, p-value = 0.037), when compared with male
counterparts. It is worth noting that what is ‘enough’ is a relative category, and people in
underserved areas might have lower expectations in this regard.
Time spent fetching water
More Ghanaian respondents agreed that they spend a substantial amount of time fetching
water as compared with South African respondents (Figure 3) although a considerable
proportion of respondents in South Africa also suggested significant time investments for
water fetching. Interestingly, even as we have the important gender skew in terms of who
fetches water, we do not see any significant differences in responses of women and men
related to time spent fetching water (ibid). This could be a function of expectations (women
may expect to spend more time on this task), as well as other responsibilities that may also
condition expectations (e.g. employment, particularly given that men are more often
employed). To this point, we observe from the intersectional analyses that female
respondents who were home owners or formally employed most often report that they spend
a significant amount of time fetching water, and male respondents who were unemployed
and did not own a home had the lowest frequency of reporting that they spend a significant
12
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
amount of time fetching water. These results reinforce the idea of relative notions of time,
where those who are employed might put more weight on the time spent for domestic chores
as they might consider their time to be at a premium.
Female n=107
Ghana
χ2 = 2.14
p−value = 0.34
Male n=103
Female n=80
South Africa
χ2 = 1.37
p−value = 0.50
Male n=68
100 50 0 50 100
Percentage
2
The X analysis was done on three categories of “Agree”, “Neutral”, and “Disagree”, with the “Strongly Agree”
13
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
and “Strongly Disagree” being grouped in their coresponding categories. The tests are between women and
men within the same country. All tests had a degree of freedom = 2.
Knowledges
Overall knowledge about water governance structures and policies was fairly low in both
Ghana and South Africa—corresponding to the fact that participation in water governance is
also generally low. In Ghana there were some gender differences in knowledge: men were
more than twice as likely to report knowledge of the AVRL consortium that had operated for
five years in urban Accra, 26% of men compared to 10% of women, a significant difference
(Χ2 = 9.09, df = 1, p-value = 0.003). In the study communities of Ghana, more men (14%)
than women (6%) knew of meetings regarding water-related issues that were held in their
community, with this difference being just shy of statistical significance (Χ2 = 3.33, df = 1, p-
value = 0.07), however, when tested in individual communities there was a significant
difference between women and men from Teshie: 9% of women and 27% of men reported
knowing about meetings (Χ2 = 4.12, df = 1, p-value = 0.042). In Ashaiman, 3% of women
and 7% of men report knowing of meetings (Χ2 = 0.31, df = 1, p-value = 0.58). The fact that
this difference would show up in Teshie but not for Ashaiman makes sense in that there are
existing water related institutions (local water boards) in Teshie, but none in Ashaiman at
present (Peloso 2013, Morinville and Harris 2014). In terms of the direction of this result,
again it is consistent with what one might expect from the broader literature, given the
expectation of higher participation among men in water and community governance.
In South Africa, one measure of water related knowledge is familiarity with the Free Basic
Water Policy (FBW): surprisingly, only 7% of women and 10% of men surveyed were aware
of the FBW, revealing no statistical difference (Χ2 = 0.29, df = 1, p-value = 0.59). A greater
number of respondents were aware of water-related meetings occurring in their communities
(25% of men, 19% of women), again with no statistical gender difference (Χ2 = 0.77, df = 1,
p-value = 0.38). It is also clear that in South Africa, there is less of a gender divide in terms of
civic participation generally, with women often taking on roles in a range of community
activities (broadly defined). From our survey, South African women participated in community
governance in the highest numbers (52%), while Ghanaian women were the least likely to
participate (12%). Numbers for male respondents were 42% for South Africa and 30% for
Ghana.
When we compare the proportion of women and men within each category of employment
and home ownership, we find knowledge of water-related meetings to be highest among
South African homeowners (both women and men), and lowest among women in Ghana,
particularly women who were home owners (Table A-2). The only significant difference found
was that in South Africa, home owning women were more likely than non-home owner
women to know about meetings (Table A-2)9. Other differences were consistent with the
direction of the hypotheses, but were not significant (i.e. more unemployed men had
knowledge of meetings than unemployed women). In other cases, the results were the
14
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
opposite of our hypothesis, but again, these were not significant. For example it is of interest
to note that knowledge of meetings among unemployed men and women was slightly higher
than their employed counterparts in both countries—an interesting result for tracking senses
of ‘enfranchisement’ among different segments of the population. As well, it is noteworthy
that homeownership correlated with knowledge of meetings in South Africa, but this did not
carry over to Ghana, where home-owning women were the subgroup least likely to know
about meetings. It may be the case, in this example, the gendered patterns of
disenfranchisement are more difficult to overcome.
Participation and governance
Participation in water governance was generally very low in all of our study sites and there
were no significant differences between women and men. Broadening the question to other
forms of civic engagement (not limited to water governance, but including community
associations and government consultation meetings), there were much greater overall levels
of participation in South Africa, where women also appear to be participating more than men.
In Ghana, a statistically significant difference was observed, but in the opposite sense—more
men participated in community governance than women (more consistent with the general
trends common in the literature, Table 3).
Table 3. Women and men's participation in community governance
Gender Gender
(GHANA) (SOUTH AFRICA)
Question
p- p-
Female Male X2 n Female Male X2 n
value value
F.1 I participate
in civic groups 10.2
12% 30% 0.001 250 52% 43% 1.63 0.20 237
0
F.1.4 I participate
in water
committees 2% 4% 0.19 0.66 204 10% 6% 0.65 0.42 240
In both Ghana and South Africa there was a great desire to participate in community
meetings, as compared with actual participation. While in both countries women were less
likely than men to report that they desired to participate more, the differences were not
15
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
significant (Figure 4). In relation to the literature, this a somewhat counterintuitive result as
we would have expected that women may ‘wish’ to participate in higher proportions, given
informal and formal barriers that curtail their actual participation. However, the lower
responses among female respondents might precisely reflect their understanding of those
barriers, and thus the senses of disengagement reflected in lower numbers of women who
‘wish’ to participate in addition to the lower rates of ‘actual’ participation. This highlights the
possibility again that gender works to condition variable expectations related to participation
and as well as for our ‘desire to participate’ proxy.
Figure 4. Wish to participate
Female n=121
Ghana
χ2 = 4.66
p−value = 0.10
Male n=115
Female n=150
South Africa
χ2 = 1.19
p−value = 0.55
Male n=98
100 50 0 50 100
Percentage
2
The X analysis was done on three categories of “Agree”, “Neutral”, and “Disagree”, with the “Strongly Agree”
and “Strongly Disagree” being grouped in their coresponding categories. The tests are between women and
men within the same country. All tests had a degree of freedom = 2.
Consistent with our expectations, men in South Africa were significantly more likely than
women to believe that community meetings were open to all (Figure 5), yet, there was no
statistically significant difference between women and men in Ghana.
Figure 5. Open community meetings
16
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
F11 I believe that meetings are open and anyone can participate
Female n=121
Ghana
χ2 = 1.85
p−value = 0.40
Male n=116
Female n=78
South Africa
χ2 = 6.02
p−value < 0.05
Male n=53
100 50 0 50 100
Percentage
2
The X analysis was done on three categories of “Agree”, “Neutral”, and “Disagree”, with the “Strongly Agree”
and “Strongly Disagree” being grouped in their coresponding categories. The tests are between women and
men within the same country. All tests had a degree of freedom = 2.
17
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
2
The X analysis was done on three categories of “Agree”, “Neutral”, and “Disagree”, with the “Strongly Agree”
and “Strongly Disagree” being grouped in their coresponding categories. The tests are between women and
2
men within the same country. All tests had a degree of freedom = 2. * Indicates X with a Monte Carlo simulation
(2000 interations) because of low frequencies (df = NA)
Worry and conflict over water
Overall Ghanaians were more likely to report that they worried about water, but there were
no significant differences between women and men (overall, nor within each country context).
Similarly, Ghanaians had disagreements with their family and community members over
water more frequently than South Africans, but again, women and men in Ghana and South
Africa gave similar answers (See Appendix for full data).
Intersectional analyses: Worry and satisfaction
In conducting intersectional analyses we found some significant differences between different
groups of men and women, even if not between men and women. In South Africa,
unemployed men were significantly more worried about water than employed men, and
women who owned homes were more worried about water than women that did not own
homes. The difference in worry by home ownership in South Africa is consistent with our
understanding of the context in that currently home owners in our study contexts are worried
about metering, and eventual requirements for payments, and linked concerned related to
indebtedness (see Rodina 2013, EMG, 2014, even if they are not always paying for water at
present). So, in this sense, somewhat counter-intuitively, home ownership in our study sites
of South Africa may elevate worry related to water, rather than the inverse. For satisfaction,
even as homeowners are more likely to enjoy in-home taps, there was more pronounced lack
of satisfaction with water services among homeowners—women with homes stated
significantly more dissatisfaction with water services compared with the women who did not
own homes. Again, this likely tracks against variable expectations, rather than some absolute
18
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
condition of those services (or may be read through the worry over payment and metering, as
noted above).
19
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
that gender differences do not exist, but rather, we must consider the ways that gender works
to condition different expectations and relative notions of what is to be expected among men
and women, or among different men and women (according to income, employment, or other
category). Where relative notions might be different, a key follow-up research task would be
to characterize and assess those different relative expectations—whether qualitatively or
quantitatively.
Second, several factors linked with the context specificities of our study sites are worth
highlighting, primary among them, the fact that they are located in urban contexts. This
context is crucial as much of the literature on gender and water focuses on rural contexts
(notable examples that do focus on urban contexts include the work of Wutich and Ragsdale
[2008] and Wutich [2009]). As such, urban sites may reveal specific experiences of water that
likely differ from other trends in the literature. For instance, in all of our sites, although people
spend significant time fetching water, access points are generally quite close to the residents’
homes (although people do at times queue to get water and may have to collect water
several times a day on a bucket-by-bucket basis). As such, it is possible that the gender
differences may not be as pronounced in these contexts as they may be elsewhere, i.e. in
locales where women have to travel long distances to water sources, or if they may be
accessing water from a natural stream or other ‘unimproved source’.10 Our observations in
these sites also confirm that both women and men are involved in fetching water for home
use, even if the bulk of this labor still falls on women. For Ghana in particular, the economic
benefits derived from water vending may also result in men being more closely linked with
water access and provision than they may be in other locales—again highlighting context
specificities that help to situate, and complicate, our results.
Third, our country specific analyses reveal interesting differences across our study sites of
Ghana and South Africa. Although these differences were anticipated (given the key
differences related to water access in the sites, as detailed in the introduction), we find that
there are again lessons for gender and water discussions. While few would argue that
gender-differentiated patterns are the same everywhere, our results nonetheless reinforce
the importance of context to understand gender dimensions of water access and experience.
In some instances, patterns observable in our Ghanaian sites were not observable for South
Africa. At times as well, there are counter-trends in terms of what is observable from site to
site11. To recall two specific examples, overall governance participation is much lower in
Ghana (21%) than in South Africa (48%). Looking at gender, there is a gender skew, for
instance, with 30% of men in Ghana reporting that they participate in community
organizations or activities in contrast with 12% of women who report the same. In South
Africa, the opposite appears to be true, with 52% of women reporting participation in
community governance compared with 42% of men. These differences suggest the
difficulties, yet also benefits, of cross-national comparative studies and of multi-sited
research design. Given the observed differences, the survey results can serve as starting
points to further open up, and unpack, these processes more fully, particularly through in-
depth qualitative examination. As such, observed quantitative differences can be understood
21
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
as invitations to work towards nuance and clarity with respect to how and why gender related
processes might work differently in various times and places. The context specificities in our
results also underscore the need for caution in terms of abstracting generalizable insights
with regards to gender-differentiated patterns. This is particularly important given that some
themes from the gender and water literature are often taken as ‘truisms’, e.g. that women
more often fetch water, or that women and men have different experiences of, and
vulnerabilities to, water given labour requirements and household uses. Instead of assuming
these patterns exist, it is important to work to clarify the specific pathways and contexts in
terms of how gender or other key differences might work to condition water-related conditions
and realities.
Fourth, our results also give us reason to pause and consider the limitations of different
research methods more generally. For instance, it is possible that some gender-differentiated
experiences are subtle, and perhaps experienced by subsets of the population, even if not by
‘most’ women in a way that would allow these patterns to be easily decipherable through
quantitative analysis of survey data. As such, the idea of statistical significance may need to
be challenged when we give weight to the lived experience and ways that water access
matters for people’s daily lives. As one key example, we know from our qualitative work that
women, particularly in South Africa, are very worried about water fetching for safety reasons
(e.g., fearing sexual assault, particularly at night). The fact that this does not show up in our
quantitative analysis does not negate the reality of these experiences for many women. We
are left with the strong sense that while there is no reason that quantitative analysis cannot
be part of a feminist and critical toolkit (Sheppard 2001, Schwanen & Kwan 2009), some
gender related complexities are likely to be subtle and nuanced enough that they may
require in-depth qualitative and ethnographic work. While we find value in testing trends, or in
finding more general insights through statistical analysis, we are also cognizant of the
reductive and glossing-over tendencies inherent in summing up complex processes and
negotiations into a ‘yes’ or ‘no’ answer—as well as linked concerns related to the
‘authoritative’ veneer such work often enjoys (ibid, Hanson 1993). While we are encouraging
critical engagement with quantitative methods (and results), we do not consider these sorts
of limitations to only be valid for quantitative studies. With qualitative work, it is possible that if
one goes looking for gender differences, selection bias or interpretive lenses may enable us
to find those very differences we seek. Theorists have cautioned that one’s theoretical
approach, or vision of the world, or what one expects to find, is important to structure the
‘truths’ revealed by the research (cf. Botkin 1990, Scott 1991). With this observation, we are
not suggesting that qualitative work is not valid or robust—rather we are suggesting that each
set of approaches likely has its associated limits, as well as benefits. We have shown that
quantitative analysis can be revealing in terms of opening up key puzzles, or targeting
specific relationships—inviting further work to better understand patterns that might be
observable in some sites, but not others. As such, we are left with the strong suggestion that
whenever possible, it is useful to pursue qualitative and quantitative work in tandem—
allowing the quantitative work to reveal patterns that can then be explained and understood
with more in-depth work. For instance, our work has revealed important questions such as
22
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
aiming to understanding why homeowners might express less satisfaction, and significant
worry, related to water access—questions that can then be pursued qualitatively in follow-up
work. Doing the qualitative work as well, and emphasizing the lived realities that come to the
fore through these methods can also allow us to give needed weight to the important
experiences in men’s and women’s daily lives—rather than only taking those ideas and
‘differences’ seriously if they can be validated statistically.
Finally, taken together, above all we find that these divergences and seeming contradictions
offer forceful suggestion to pursue diverse methods not only in tandem, but also in
conversation—taking the tensions and divergences between approaches seriously to
interrogate and query these approaches, and what they might offer to challenge and unsettle
our understanding. As Nightingale (2003, 2009) convincingly argues, ‘triangulation’ between
different data sources is not necessarily to ‘validate’ particular knowledges, but rather to be
able to query, and consider why certain truths appear in certain modes of inquiry, while other
realities may be visible through other methods. The point, she argues, is thus not to find the
truth, but to be better able to speak to the silences, tensions, and convergences and
divergences between diverse realities revealed by different approaches (see also Hesse-
Biber 2010, Harris 2011). In this way, we are not necessarily working for complete knowledge,
but rather, seeking to learn from the necessarily partiality and situatedness of diverse
knowledges (Haraway 1988). Our work may not offer decisive evidence of the ways that
gender matters for water in all the ways that we had expected. It does, however, validate the
urgent and ongoing need to better speak to divergences and convergences in our
knowledges and research approaches, particularly as we engage with diverse realities and
contexts. It is only in this way that conceptual engagements with ideas of gender can be
interrogated, and that understandings of gender in general, and gender and water (access,
use, knowledges and experiences) in particular, can be better understood—with attention to
the nuance and complexity required for the complex and ever-changing terrain of water and
society.
23
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
REFERENCES
Adank, M., Darteh, B., Moriarty, P., Osei-Tutu, H., Assan, D., & van Rooijen, D. (2011).
Towards integrated urban water management in the Greater Accra Metropolitan Area. Accra,
Ghana: SWITCH.
Agarwal, B. (2010). Does women’s proportional strength affect their participation? Governing
local forests in South Asia. World Development, 38(1), 98-112.
Aguilar, L. (2005). Water as a source of equity and empowerment in Costa Rica. In V.
Bennett, S. Davila-Poblete, & M. N. Rico (Eds.), Opposing currents: The politics of water and
gender in Latin America (pp. 123–134). Pittsburgh, PA: University of Pittsburgh Press.
Ainuson, K. G. (2010). Ensuring adequate water supply to disadvantaged urban communities
in Ghana. African Studies Quarterly, 11(4), 59-82.
Alston, M. (2006). The gendered impact of drought. In B. Bock, and S. Shortall (Eds.), Rural
gender relations: Issues and case studies. Cambridge, MA: CABI.
Arku, F., & Arku, C. (2010). I cannot drink water on an empty stomach: A gender perspective
on living with drought. Gender and Development, 18(1), 115-124.
Barnes, J. (2013). Who is a water user? The politics of gender in Egypt’s water user
associations. In L. Harris, J. Goldin, & C. Sneddon (Eds.), Contemporary water governance
in the Global South (pp. 185-198). London: Routledge.
Bennett, V., Davila-Poblete, S., & Rico, N. (2005). Opposing currents: The politics of water
and gender in Latin America. Pittsburgh, PA: University of Pittsburgh Press.
Botkin, D. (1990). Discordant harmonies: A new ecology for the twenty-first century. New
York: Oxford.
Buechler, S., & Hanson, A. S. (2015). A political ecology of women, water and global
environmental change. London: Routledge.
Butler, J. (1990). Gender Trouble: Feminism and the Subversion of Identity. London and New
York, Routledge.
CCT. 2013a. City of Cape Town — 2011 Census — Ward 018 (January 2013).
(https://www.capetown.gov.za/en/stats/Pages/wards_2011census.aspx) (accessed 24 March
2013)
CCT. 2013b. City of Cape Town – 2011 Census – Ward 087 (January 2013)
(https://www.capetown.gov.za/en/stats/Pages/wards_2011census.aspx) (accessed 24 March
2013)
24
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
Dapaah, E. (2014). Water access and governance among indigenous and migrant low
income communities in the Greater Accra Metropolitan Area (GAMA), Ghana. (Unpublished
master’s thesis). University of British Columbia.
Dublin Principles. (1992). Dublin Statement on Water and Sustainable Development.
Retrieved May 6, 2015, from http://www.un-documents.net/h2o-dub.htm
Ennis-Mcmillan, M. (2005). La vida del pueblo: Women, equity, and household water
management in the valley of Mexico. In V. Bennett, S. Davila-Poblete, & M. N. Rico (Eds.),
Opposing currents: The politics of water and gender in Latin America (pp. 137-153).
Pittsburgh, PA: Pittsburgh University Press.
Environmental Monitoring Group. (2014). Retrieved May 4, 2015, from http://www.emg.org.za.
Galvin, M. (2011). Participating in urban myths about women’s rural water struggles. Agenda:
Empowering Women for Gender Equity, 25(2), 87-100.
Goldin, J. (2010). Water policy in South Africa: Trust and knowledge as obstacles to reform.
Review of Radical Political Economics Sage Publications, 42(2), 195–212.
Goldin, J. (2013). Proposition for a sector toolkit: Addressing the gap between policy and
practice through the use of a multi-dimensional poverty framework. Water Policy, 15, 309-
324.
Hanson, S. (1993). Never question the assumptions and other scenes from the revolution.
Urban Geography, 14(6), 552-556.
Haraway, D. (1988). Situated knowledges: The science question in feminism and the
privilege of partial perspective. Feminist Studies, 14(3), 575 – 599.
Harris, L. (2009). Contested sustainabilities: assessing narratives of environmental change in
southeastern Turkey. Local Environment 14(8), 699 - 720.
Harris, L. (2005). Negotiating inequalities: Democracy, gender and politics of difference in
water user groups in Southeastern Turkey. In F. Adaman & M. Arsel (Eds.),
Environmentalism in Turkey: Between democracy and development? (pp. 185-200).
Aldershort, UK: Ashgate.
Harris, L. (2006). Irrigation, gender, and social geographies of waterscape evolution in
Southeastern Turkey. Environment and Planning D: Society and Space, 24(2), 187-213.
Hawkins, R., & Ojeda, D. (2011). Gender and environment: Critical tradition and new
challenges. Environment and Planning D: Society and Space, 29(2), 237-253.
Hesse-Biber, S. (2012). Feminist approaches to triangulation: Uncovering subjugated
knowledge and fostering social change in mixed methods research. Journal of Mixed
Methods Research, 6(2), 137-146.
25
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
Mehta, L. (2006). Do human rights make a difference to poor and vulnerable people?
Accountability and the right to water in South Africa. In J. Wheeler & P. Newell (Eds.), Rights,
resources and accountability (pp. 63-78). London and New York: Zed Books.
Morinville, C. and Harris, L. (2014) Participation, politics and panaceas: exploring the
possibilities and limits of participatory urban governance in Accra, Ghana Ecology and
Society 19 (3), 36.
Morinville, C. (2012). Beyond the pipe: Participation and alternative water provision in
underserved areas of Accra, Ghana. (Unpublished master’s thesis). University of British
Columbia. Available at www.edges.ubc.ca.
Nightingale, A. (2003). A feminist in the forest: Situated knowledges and mixing methods in
natural resource management. Acme: An International E-Journal for Critical Geographies
2(1), 77-90.
Nightingale, A. (2009). Triangulation. In N. Thrift & R. Kitchin (Eds.), International
encyclopedia of human geography. Oxford: Pergamon.
Neumayer, E., & Plumper, T. (2007). The gendered nature of natural disasters: The impact of
catastrophic events on the gender gap in life expectancy, 1981-2002. Annals of Association
of American Geographers, 97(3), 551-566.
Peloso, M. and Morinville, C. (2014) Chasing for water: Everyday practices of water access
in peri-urban Ashaiman, Ghana. Water Alternatives 7 (1), 140-159.
Peloso, M. (2014). Navigating water access and governance in peri-urban Ashaiman, Ghana:
a case study. (Unpublished master’s thesis) University of British Columbia. Available at
www.edges.ubc.ca.
Resurreccion, B., & Elmhirst, R. (Eds.). (2008). Gender and natural resource management:
Livelihoods, mobility, and interventions. London and Sterling, VA: Earthscan.
Rocheleau, D., Thomas-Slayter, B., & Wangari, E. (1996). Feminist political ecology: Global
issues and local experiences. New York: Routledge.
Rodina, L. (2013). Lived notions of citizenship and the human right to water in Site C,
Khayelitsha, Cape Town, South Africa. (Unpublished master’s thesis). University of British
Columbia. Available at www.edges.ubc.ca.
Schwanen, T., & Kwan, M. (2009). Doing critical geographies with numbers. The
Professional Geographer, 61(4), 459-464.
Scott, J. W. (1991). The evidence of experience. Critical Inquiry, 17, 773-797.
Sheppard, E. (2001). Quantitative geography: Representations, practices, and possibilities.
Environment and Planning D: Society and Space, 19, 535-555.
26
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
Singh, N., Jacks, G., Bhattacharya, P., & Gustafsson, J. E. (2006). Gender and water
management: Some policy reflections. Water Policy, 8(2), 183-200.
Sorenson, S. B., Morssink, C., & Campos, P. A. (2011). Safe access to safe water in low
income countries: Water fetching in current times. Social Science and Medicine, 72, 1522-
1526.
Special issue on feminist political ecologies of water. (2009). Gender Place and Culture,
16(4).
Special issue on rethinking feminist political ecologies. (2011). Geoforum, 42(2).
Special issue on feminist approaches to water. (2013). Feminist Review, 103.
Sultana, F. (2011). Suffering for water, suffering from water: Emotional geographies of
resources access, control and conflict. Geoforum, 42(2), 163-172.
WHO/UNICEF Joint Monitoring Programme. (2010). Progress on Sanitation and Drinking
Water. Retrieved May 6, 2015, from
http://www.who.int/water_sanitation_health/publications/9789241563956/en/
Wutich, A., & Ragsdale, K. (2008). Water insecurity and emotional distress: Coping with
supply, access, and seasonal variability of water in a Bolivian squatter settlement. Social
Science & Medicine, 67, 2116-2125.
Wutich, A. 2009. Intrahousehold disparities in women and men's experiences of water
insecurity and emotional distress in urban Bolivia. Medical Anthropology Quarterly, 23.
Zwarteveen, M., & Liebrand, J. (2015). Performing modernity: The scalar politics of irrigation
development in Nepal. In E. Norman, C. Cook, & A. Cohen (Eds.), Negotiating water
governance: Why the politics of scale matters. Ashgate.
27
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
NOTES
1
RDP refers to the ‘Reconstruction and Development Program’, an effort of democratically elected ANC
(African National Congress) that includes focus on providing housing and other basic services to impoverished
populations, particularly as part of redressing apartheid era inequalities. At present in Site C, Khayletisha, the
predominantly black township that is the focus of our study, as many as 1/3 of residents might be living in newly
build RDP homes, while the majority live in corrugated metal or other ‘shack’ dwellings (CCT 2013a, 2013b).
With ongoing formalization, the expectation is that all residents will eventually be living in formal RDP homes, or
will be resettled elsewhere.
2
More details on the broader research effort, and linked publications, is available on our website,
www.edges.ubc.ca.
3
Several surveys were taken out for the analysis due to lack of completeness.
4
The survey was funded by the Centre for International Governance Innovation (CIGI) and was implemented
through a partnership between the EDGES Research Collaborative (Environment and Development: Gender,
Equity and Sustainability Perspectives) at The University of British Columbia (UBC), the Anthropology of Water
Research Group at the University of the Western Cape, and Professor Akosua Darkwah of the University of
Ghana-Legon. Follow-up work was funded by the Social Sciences and Humanities Research Council of Canada
(SSHRC), and has included qualitative fieldwork conducted in communities of Accra (2011-present) and in
Cape Town (2012-present). More details on the survey instrument, implementation, and related research
publications are available at www.edges.ubc.ca.
5
The WHO/UNICEF Joint Monitoring Program reports that 90% of Accra’s population has access to improved
drinking water, although the report also specifies that only 33% do so through their own direct connection to the
GWCL (WHO/UNICEF Joint Monitoring Programme, 2010). Other accounts report that between 51% (Adank,
Darteh, Moriarty, Osei-Tutu, Assan, & van Rooijen, 2011) and 59% of Accra’s population is served by the
GWCL (Ghana News Agency, cited in [Ainuson, 2010]).
6
In either case, residents would be considered ‘homeowners’ for purposes of our survey, even as the home
might be a corrugated metal shack or backyard dwelling.
7
Recall several key citations highlighted above that highlight gender and water stress, disasters, or
floods/drought, including Wutich and Ragsdale (2008); Neumayer and Plumper (2007); Alston (2006).
8
Respondents with an in-house water source were removed from this analysis, which accounts for the low
sample sizes, particularly in South Africa where in home/yard water sources were more common.
9
It is likely, but not certain, that homeownership indicates time living in the community, as more recent arrivals
to the community might be less likely to own homes, in addition to being a rough proxy for wealth and
enfranchisement (though again, few –to none—of the respondents would be considered ‘wealthy’ as all study
sites are relatively impoverished settlements.
10
Recall that cross-national data on water fetching cited in the introduction confirms that women are more likely
to fetch water when improved sources are unavailable.
28
INSTITUTE FOR RESOURCES, ENVIRONMENT AND SUSTAINABILITY
UNIVERSITY OF BRITISH COLUMBIA
11
This was very common in another analysis we conducted in the impact of material water conditions on
governance participation, see Author et al in process). In several of the analyses run, gender was significant,
but only at the country level, and at times the trends were in the opposite direction, so cancelled each-other out
in the overall dataset.
29