Professional Documents
Culture Documents
Corruption and Foreign Direct Investment
Corruption and Foreign Direct Investment
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve and extend access
to Journal of International Business Studies
Mohsin Habib*
UNIVERSITY OF MASSACHUSETTS, BOSTON
Leon Zurawicki**
UNIVERSITY OF MASSACHUSETTS, BOSTON
This study examines the impact of ined. The analysis provides support
corruption on foreign direct invest- for the negative impacts of both.
ment (FDI). First, the level of cor- The results suggest that foreign
ruption in the host country is ana- investors generally avoid corrup-
lyzed. Second, the absolute differ- tion because it is considered wrong
ence in the corruption level between and it can create operational in-
the host and home country is exam- efficiencies.
291
JOURNAL OF INTERNATIONAL BUSINESS STUDIES, 33, 2 (SECOND QUARTER 2002): 291-307
FDI, Belgium, which is similarly rated bilities, and the reduction in transaction
on corruption (by Transparency Interna- costs. The third element, location (L) ad-
tional), attracts substantial FDI. Address- vantages, has attracted renewed atten-
ing such a paradox requires careful anal- tion due to (1) changes in the extent,
ysis. So far, scholars in economics and character and geography of MNE activity
public policy have dealt with corruption in the last two decades, and (2) the link-
(Bardhan, 1997). Analogous empirical ages of the spatial aspects of the MNE
research from the international business
value-added activities to firm competi-
perspective is scanty. tiveness (Dunning, 1998). The location
Understanding the pernicious role ofof FDI is driven by the search for (1)
corruption in FDI is important since itmarkets, (2) resources, (3) efficiency, and
produces bottlenecks, heightens uncer-(4) strategic assets (Dunning, 1998). This
tainty, and raises costs. Further, corrup-implies that different location selection
tion creates distortions by providing criteria apply to projects of different mo-
some companies preferential access totivation. Thus, identifying the variables
profitable markets. The difference in theexerting the strongest impact on particu-
exposure to corruption between the host lar types of investment is important.1
and home countries can also be a con- The majority of previous studies have
relied on total FDI inflow data, obscuring
cern for investors. The greater such dif-
ference between the two countries, the the links between specific location char-
lower the likelihood that they know how acteristics and the type of motivation.
to deal mutually. Given a choice be- This study also uses aggregate FDI data,
tween a familiar and a less familiar en-
however, when appropriate, the location
variables relevant to specific FDI orien-
vironment, firms will prefer the former
(Davidson, 1980). tation are highlighted.
This paper adopts a dual approach in Identifying additional variables can
assessing the impact of corruption on expand our knowledge of the (L) advan-
FDI. First, the effect of the level of cor- tages and their influence on FDI. To this
ruption in the host country is investi- end, corruption, a factor that has re-
gated. Second, the effect of the difference ceived attention lately, is included
in the corruption levels between the among the descriptors of the attractive-
home and host countries is examined. ness of a location. Corruption is defined
The paper is organized as follows. Ain various ways. International organiza-
re-
view of corruption and its impact on tions
FDI (UN) and Western governments
equate corruption with "improbity",
are presented first. Based on this review,
hypotheses are developed. Analysis which
and encompasses not only what is il-
results are then reported. Finally, legal,
the but also improper (Malta Confer-
end section discusses the findings ence,
and 1994). The World Bank empha-
their implications. sizes the abuse of public power for pri-
vate benefit (Tanzi, 1998). Whereas
THEORY
corruption of public employees is a dom-
inant theme, Coase (1979) also argues
The OLI paradigm (Dunning, 1988)
that corruption exists between private
provides guidance for FDI activities. The
ownership (0) and internalization (I) parties.
ad- Although corruption can be op-
vantages are derived from the exploita-erationalized as an all-inclusive variable,
comprising bribes, bureaucratic ineffi-
tion of firm-specific resources and capa-
measure including corruption. However, under control. Also, the host country's
this result could be affected by the mea- willingness to accept the presence of TI
sure itself since, besides corruption, it chapter sends a positive signal. If, how-
combined twelve other indicators, some ever, corruption is deeply ingrained,
of them marginally important for FDI bringing in TI can help improve the cur-
(Wei, 2000). Drabek and Payne (1999) rent FDI situation but may not do much
tested how FDI was affected by non- to actually combat corruption. Empiri-
transparency, a composite of corruption, cally, the impact of corruption on FDI
unstable economic policies, weak and when a corporate watchdog is present is
poor property rights protection, and poor yet to be determined. Hence, the follow-
governance. Results indicated the nega- ing hypothesis is proposed.
tive impact of high levels of non-trans-
Hypothesis lb: Corruption will con-
parency on FDI.
tinue to have a negative effect on FDI
Other studies identified a negative re-
when TI in the host country is in-
lationship between corruption and FDI.
cluded as an independent variable in
Wei (2000) analyzed data on FDI in the
the analysis.
early 1990s from 12 source countries to
45 host countries. Corruption revealed a It is believed that political stability
significant and negative effect on FDI. positively affects FDI (Kobrin, 1976) and
Busse et al. (1996) looked at the relation- that political stability and corruption are
ship between FDI and the level of cor- negatively correlated (Wei, 2000). Uncer-
ruption exposed by the local media. tain political situations make investors
Reviewing the inconsistent results per- and public officials short-term oriented
taining to different countries, they hy- and pursuing personal gains while sacri-
pothesized that FDI increased when in- ficing the legality. Alternatively, a stable
vestors believed that the government political environment encourages a long-
would curb corruption. Alternatively, term orientation and reduces incentives
FDI would decrease with media expo- for quick illegal returns. In the end,
sure should investors assume that the
whether the dual effects of political sta-
government was unwilling to improve
bility will actually change the FDI-cor-
the environment. This suggestion, how- ruption relationship is an open question.
ever, was not rigorously validated. Given
The following hypothesis is suggested.
the state of the literature, the following
basic hypothesis is proposed first. Hypothesis lc: Corruption will con-
tinue to have a negative effect on FDI
Hypothesis la: Corruption will havewhen
a political stability is included as
negative effect on FDI. an independent variable in the analy-
Various institutions monitor and ex- sis.
dogs like Transparency International (TI) For a long time, researchers looked at
can serve two purposes: as a deterrent to the order of selection of foreign markets
corrupt officials and as a reassurance to Since Cyert and March (1963), the be-
havioral model attempted to identify th
the investors that the situation is getting
markets firms enter first and determine and risky commitment among the inter-
how they organize at each stage of devel-national business activities (Johanson
opment within the new market. A deci- and Wiedersheim-Paul, 1975). Two in-
sion to enter a foreign market is a func- fluences of opposing directions can oc-
tion of knowledge and experience, andcur. First, psychic distance can prove a
the selection of a similar market reduces
greater impediment to FDI in a particular
uncertainty (Aharoni, 1966). In the 70's,
country than to other modes of interna-
the Scandinavian authors categorized tional business. Second, applying previ-
the "psychic distance," to assert that ous lessons from exporting and licens-
companies enter markets perceived to being, the international investors may
psychologically closer before consider-bridge their knowledge and experience
ing the remote ones (Johanson and Wied- gap.
ersheim-Paul, 1975; Johanson and In this study, the notion of similarity
Vahlne, 1977, 1990). Two importantthat as- closely resembles the psychic dis-
pects are emphasized: (1) lower leveltanceof idea is adopted. The similarity ap-
uncertainty in psychically close coun- proach is simpler and more clear-cut as
tries (Johanson and Vahlne, 1990), and it does not presuppose managers' level of
(2) psychic closeness facilitating learn-
knowledge about the host country. Cor-
ing about the target countries (Kogutruptand practices represent a component of
Singh, 1988). local business and administrative cus-
Over the years, more studies were un- toms. Inability to handle corruption
dertaken. Some (Davidson, 1980; Nord- makes FDI challenging for companies
strom and Vahlne, 1992) corroborated from less corrupt countries and can re-
the hypothesis that firms expand first to sult in a negative FDI decision. Alterna-
the psychically close and then to the tively, exposure to corruption at home
more psychically remote countries. Yet, provides a learning experience preparing
Benito and Gripsrud (1992) and Engwall the individual companies to handle
and Walenstal (1988) found no support them abroad. Hence, acquiring skills in
for the hypothesis that the first FDIs are managing corruption helps develop a
made in the countries that are culturally certain competitive advantage. This ad-
closer to the home country. Recently, vantage is lost or turned into disadvan-
Ghemawat (2001) suggested that four di- tage when expertise in corruption be-
mensions of distance namely cultural, comes redundant in "clean" markets.
administrative, geographic and eco- While it is easy to comprehend that the
nomic, influence companies considering difference in corruption level between
global expansion. Within this taxonomy, two countries would be more problem-
difference in corruption levels between atic for investors coming from a less cor-
the host and home countries is part of rupt environment than those from a
the administrative distance that creates
more corrupt one, it still constitutes a
significant barriers for foreign investors.
"distance" that both types of investors
To complicate matters further, the psy-
have to overcome. Applying the similar-
chic distance can exert different effects
ity approach to corruption, the following
depending on the mode of firms' in-
hypothesis is proposed.
volvement in international business. To
date, only a handful of psychic distance Hypothesis 2: The greater the absolute
studies addressed FDI, the most serious difference in the corruption levels be-
tween the home and the host coun- ables in the study. In this model, the
tries, the lesser the FDI inflow for dependent
the variable is the log of FDI. FDI
host country. is the annual inflow between the home
and the host country. Log of FDI is used
METHODOLOGY
in the analysis to render the distribu-
tions
The study includes a large number of nearly normal and the error term
host countries offering a broad perspec- homoscedastic. Data on FDI is collected
tive on FDI and corruption. The statistics from the IMF (2000).
on aggregate bilateral FDI flows are ana- The key independent variables are cor-
lyzed in conjunction with key observa- ruption (measured by CPI) and the abso-
tional variables. Most of the data for the lute difference in corruption (Abs. Diff.
sample is derived from the Internationalin CPI) between the host and the home
Monetary Fund (IMF, 2000). The latestcountry. Measuring corruption is prob-
three years with significant country datalematic. There is no consensus among
(1996-1998) are chosen. Altogether, 89researchers regarding what should be
countries are included. These countries measured (Lambsdorff, 1999). Objective
represent the whole spectrum, compris- measures are hardly available because of
ing developed, developing, and the tran- the secrecy of corrupt dealings. Subjec-
sition economies. The selection of the tive measures relying on questionnaire-
home countries: Germany, Italy, Japan, based surveys represent an acceptable al-
Korea, Spain, UK and the USA aims to
ternative for this problem. They measure
provide a fair representation of the
theperception of corruption rather than
Triad, while incorporating countries
corruption per se. Such surveys are com-
piled by organizations like Transparency
with markedly varying levels of Corrup-
International (TI), Political Risk Ser-
tion Perception Index (CPI), a measure
produced by Transparency International and World Economic Forum. In-
vices,
(TI). This is very important, as based terestingly,
on the respective indices are
the methodology applied by TI relatively highly correlated (Tanzi, 1998). In this
small differences in CPI for individual study, the corruption measure collected
by TI (1999) is used. TI, a non-govern-
countries are not statistically significant.
The three-year bilateral FDI data wasmental organization, publishes the Cor-
ruption Perception Index (CPI), which is
pooled together for statistical analysis.
an average of multiple surveys of coun-
Accordingly, a maximum of (7 x 89 x 3)
1869 observations for the variables wastry and business experts. The index
expected. However, a large number of ranks up to 100 countries from the most
missing values related to a country ortoa the least corrupt on a 0 to 10 scale.
year produced fewer total observations.Since its inception in 1995, academics
Two different models are used in this and companies have used the CPI exten-
sively.
study to analyze the effects of corruption
on FDI. The first is an OLS regressionBesides the key independent vari-
model. The second is a PROBIT model, ables, the regression model includes
devised to assess the impact on FDI of"control" variables reflecting the deter-
the absolute differences in corruptionminants of FDI suggested in the litera-
ture. First, variables that affect all FDI
levels between the home and the host
endeavors are included. Second, vari-
countries. The OLS approach is appro-
ables that influence the location of spe-
priate given the use of continuous vari-
cific types of FDI are highlighted. A brief (Billington, 1999). Under high unem-
discussion of the variables follows. ployment, workers value their current
A classic reason for FDI is the search job higher and are willing to accept
lower wages to keep the jobs. Thus, a
for new markets. FDI is positively influ-
enced by the size of the host economy high unemployment would positively af-
measured by its GDP or population (Ko-fect the resource-seeking FDI. The unem-
brin, 1976). Large markets provide a rea-
ployment data was gathered from the In-
sonable scope for investment and hence ternational Labor Organization (ILO,
influence market-seeking FDI. Log of 2001).
population is our measure of countryStrategic linkage theory (Nohria and
size. Garcia-Pont, 1991) and the network ap-
Host countries' growth rates of GDP proach (Johanson and Mattson, 1988)
seem to be positively related to FDI (Ko- have been presented as other explana-
brin, 1976; Nigh, 1986). High growth en-tions for FDI. Accordingly, FDI serves to
sures demand for the output of the localacquire strategic resources that the firms
market-oriented FDI. GDP growth is in- are lacking. Dunning (1998) and Doz,
cluded in this study. Santos & Williamson (2001) emphasized
GDP/capita is a significant explanatorythe need for most foreign firms to find
variable for FDI (Wells and Wint, 2000; locations supporting innovative activi-
Grosse and Trevino, 1996). High GDP/ ties. Host countries with higher scores
capita reflects high consumption poten- on science and technology will be pre-
tial in the host country. Log of GDP per ferred for asset-seeking FDI. Country rat-
capita is used in this study. ings for science and technology are
Export orientation of the host country adopted for this study from the World
can stimulate FDI (Jun and Singh, 1996). Competitiveness Yearbook (1999).
Countries open to international trade It has been shown that the duration of
provide a good platform for global busi- diplomatic and economic ties between
ness operations. Also, a country's inter- the home and host country increases the
national orientation reflects its competi- likelihood of FDI (Tse, Pan and Au,
tiveness. International orientation of a 1997). Greater business interactions
country is measured as the trade/GDP should promote understanding between
ratio from the IMF (2000) data. the home and host country that is con-
ducive to FDI. Economic ties reflecting
Political stability strongly affects FDI
(Kobrin, 1976). It is considered an imper-
the home and the host country participa-
ative for planning, profitability, andtion in the same common market areas
(EU, NAFTA, ASEAN) or preferential
long-run success. Political Risk Services,
Inc. (2000) publishes Political Risk In-
trade agreements (Lome Convention, Ca-
dex, which is on a scale of 0 to 100, with
ribbean Initiative) are included as a
100 being the most politically stable. The
(dummy) variable in this study. Data was
index is used in this study. collected from the official membership
Labor is another factor importantliststo of the regional blocs.
foreign investors. The more abundant As mentioned earlier, cultural and
(with lower costs) labor is, the more at-
geographic distance affects FDI. Cultural
tractive the location becomes. Country-proximity between the two countries
level unemployment figures have been will facilitate FDI operations. It was cal-
culated applying the Kogut and Singh
considered a proxy for labor availability
(1988) formula to Hofstede's four dimen- Then a categorical variable was devel-
sions of culture. Geographic proximity oped. 0 is assigned if the host country's
exerts a multifaceted impact on FDI, fa- share in global FDI exceeds its share in
cilitating contact with host country and the home country's total outgoing FDI
reducing the transportation costs. Also, (under-investment). 1 applies to cases
in case of large geographic distance, the when the host country's share in global
transportation costs encourage substitu- FDI is equal to or lower than its share in
tion of exports by market-oriented FDI the home country's total FDI (over-in-
(Brainard, 1997). Log of geographic dis- vestment). This categorical variable
tance between the home and the host which focused on relative "under-invest-
countries was calculated from Hen- ment," as opposed to relative "over-in-
geveld (1996). vestment," is subsequently used as a de-
pendent variable in the PROBIT analy-
Finally, the presence of the corporate
watchdogs in the host countrysis, is to
in-estimate the probability that the
cluded as a variable affecting FDI.bilateral
It is FDI flows would fit one or an-
other category as a function of a set of
expected that the presence of organiza-
independent variables. Data on a coun-
tions such as TI will improve the invest-
ment climate. Representation of TI try's
in theshare in global FDI is calculated
from the IMF (2000). Data on the coun-
host country was included as a dummy
variable in the study (TI, 1999). try's share in the total FDI of the home
country
In order to avoid a problem known as is calculated from the OECD
(2000) statistics.
the "omitted variable bias," the regres-
Most of the independent variables
sion model is developed to be as inclu-
sive as possible. All the independent included referred to the same factors as
variables described earlier are included in the OLS regression model.2 How-
in the same model. A test of multicol- ever, since the PROBIT model focuses
linearity among independent variables on the impact of similarities, except for
the geographic distance and economic
using the variance-inflation factor (VIF)
ties, all other independent variables
did not suggest any serious problem.
None of the VIF values exceeded 3. VIF are considered in terms of absolute dif-
values of 5.3 (Hair, Anderson, Tatham ferences between the home and the
and Black, 1992) and 10 (Studenmund, host country levels. It is expected tha
1992) have been suggested as cutoffs for these differences would lead to greate
multicollinearity. Table 1 reports the de- difficulties in operations of foreign
scriptive statistics. firms and, hence, deter FDI. The only
The second (PROBIT) model attempts new variable added is the log of the
to explain the differences between the absolute difference in Gross Capital
host country's share in the global (total) Formation/capita. Like most other vari-
incoming FDI and the same country's ables, it is included to capture as much
share in a specific home country's FDI. as possible the macro environmental
First, the raw differences were calcu- differences between the host and home
lated: countries. GCF data is calculated from
the IMF (2000). Table 2 presents the
Difference in share of FDI = Global
descriptive
Share - Countryi Share, where i denote a statistics for the PROBIT
particular home country of outgoingmodel.
FDI.
t TABLE 1
MEANS, ST
Stability 72.41 11.71 1582 -.26**" -.07** .78"** -.19**" .31*** .41*** -.05
9. Log Distance 8.11 0.89 1848 .09*** .06* -.25*** -.13*** -.09*** -.14*** .14*** -.27
10. Economic Ties 0.20 0.40 1830 .07** -.05 .31** .11 .07** .12"'* -.17*** .2
11. Corruption
Perceptio
Index (CPI) 4.96 2.45 1491 -.30*** -.03 .868' -.07 .20** .55**
12. Abs. Diff. In
CPI 2.79 1.72 1465 .12**' .01 -.31*** -.00 -.14'** -.17*** .42*** -.
13. LogFDI 4.34 2.45 785 .20*** .02 .23"* -.00 -.06 .22'' -.14'* .1
N~~c~ ~ ~ ~ ~ ~~D~~IN
CD
CO
*p <.05, **pK<.01.
TABLE 3
REGRESSION RESULTS FOR THE FULL SAMPLE
(DEPENDENT VARIABLE = LOG FDI)
Variable Model 1 Model 2 Model 3 Model 4 Model 5
Log Population 0.33*** (0.14) 0.43*** (0.16) 0.40*** (0.17) 0.39*** (0.16) 0.44*** (0.16)
GDP Growth 0.01 (0.06) -0.03 (0.06) -0.02 (0.06) -0.04 (0.06) -0.04 (0.06)
Log GDP/Capita 0.22** (0.25) 0.05 (0.31) -0.08 (0.35) 0.14 (0.32) 0.05 (0.31)
Unemployment 0.04 (0.03) 0.07 (0.03) 0.11t (0.03) 0.07 (0.03) 0.06 (0.03)
Trade/GDP 0.11t (0.22) 0.15** (0.22) 0.13* (0.22) 0.11t (0.23) 0.12* (0.22)
Science &
Technology 0.02 (0.01) -0.02 (0.01) 0.04 (0.01) -0.03 (0.01) -0.05 (0.01)
Cultural Distance -0.05 (0.01) -0.06 (0.01) -0.07 (0.01) -0.04 (0.01) 0.02 (0.01)
Log Distance -0.01 (0.15) -0.05 (0.15) -0.07 (0.15) -0.03 (0.16) -0.05 (0.15)
Economic Ties 0.19** (0.31) 0.15* (0.32) 0.13* (0.32) 0.15* (0.32) 0.16** (0.32)
CPI 0.25* (0.11) 0.18t (0.11) 0.18t (0.11) 0.26** (0.11)
Political Stability 0.18* (0.02)
TI Chapters 0.14** (0.30)
Abs. Diff. In CPI -0.18*** (0.09)
Adjusted R2 0.151 0.165 0.173 0.179 0.203
Change in Adj. R2 0.014 0.008 0.006
F 9.01*** 8.97*** 8.68*** 9.00*** 9.48***
N 405 405 405 405 403
Standardized regressio
tp < 0.10, *p < 0.05,
0.203). Model 5 supports both Hypothe- was done because it has been suggested
ses la and 2.3 that GDP/capita and political stability
Hypothesis 2 is also tested using the not only affect FDI, but also corruption
PROBIT model. The results are shown in (Wei, 2000; Husted, 1999). In both mod-
Table 4. Model 1 includes all the control els, difference in CPI remains significant
variables, several of which show signifi-and negatively affects FDI. In fact, the
cant results (discussed later). In model 2,significance level increases to p < 0.05
the difference in the CPI variable is in- level when the two variables are ex-
TABLE 4
PROBIT ESTIMATON RESULTS FOR FDI
(0 = LESS SHARE OF FDI, 1 = MORE SHARE OF FDI)
Variable Model 1 Model 2 Model 3 Model 4
Apart from the difference in CPI, tries. This further suggests that foreign
geographic distance and economic ties firms are unwilling to deal with the plan-
show expected significant results in the ning and operational pitfalls related to
PROBIT analysis (negative and positive an environment with a different corrup-
relationships to FDI, respectively). How- tion level. This important fine point is
ever, some unexpected results emerge. worth emphasizing, as it has not yet been
Log of absolute difference in GCF/capita, empirically tested in the literature.
absolute difference in internationaliza- Including corruption in the FDI model
tion, and absolute difference in scienceshould help the companies realize the
and technology all showed positive sig- individual importance of that factor in
nificant relationships to the share of FDI the site selection.4 More importantly, in
(some only marginally significant at the a dynamic business environment, con-
p < 0.10 level). It seems the greater the sidering corruption relative to other di-
difference the higher the FDI flows into mensions will help the managers imple-
the host country. These results require ment complex analyses and refine coun-
further inspection for possible explana- try evaluation procedures. Based on this
tions. study, even small changes in corruption
matter. Thus, once companies determine
DISCUSSION AND CONCLUSION
the importance of corruption for FDI,
This study examined the relationship their response to the expected deteriora-
between corruption and FDI based on tion/improvement in the host country
the recent three-year data. The findings corruption level would have to be pro-
are consistent with the arguments pre- grammed.
sented in the literature and suggest that Foreign investors should take an ag-
corruption is a serious obstacle for in- gressive stance and combat corruption
vestment. The data for this study are ob- for their own long-term interest. For ex-
tained from international statistics onample, if a competitor obtains a license
FDI, aggregated by countries of origin to operate or secures a contract through
and destination. As such, it generalizes
corrupt dealings it is up to the disad-
the individual experiences of thousandsvantaged companies left behind to blow
of investment projects and adds to the ourwhistle. In many national legal sys-
understanding of the pattern of inves- tems, one can render null and void the
tors' reactions to corruption. agreements obtained through corrupt
activities provided convincing proof is
The theoretical arguments against cor-
presented (Malta Conference, 1994). Ex-
ruption derive from both ethics and eco-
nomics. Foreign investors may shun cor-change of information within the busi-
ruption because they believe it is mor-ness community can foster the process.
ally wrong. They may also try to avoid Examples can be found where compa-
corruption because it can be difficult nies
to have made a difference in the way
manage, risky, and costly. The negativethey managed FDI in a relatively "cor-
rupt" country. McDonalds opened its
effect of corruption on FDI found in this
study suggests that firms, as a whole,successful
do outlet in Moscow instilling its
not support corruption. However, in ad-
international standards and making no
dition, the study also found a negative
compromise for the local "corrupt" envi-
ronment. The name McDonald, obvi-
effect due to the difference in corruption
levels between the home and host coun- ously helped maintain a strong stand
against corruption. Another example in- 2. The only omission is the cultural
volving Russia is the 3M Inc., which cap- distance variable. The regression analy-
italized on both the strengths and weak- sis did not find it significant (Table 3).
nesses of the Russian culture positioning Given its correlation with several control
itself as a socially and morally responsi- variables (Table 1), we decided to ex-
ble company. It raised its involvement clude it from further analysis.
with the locals by creating successful 3. As an additional test of the hypoth-
networks but at the same time actively eses, regressions for the FDI samples of
promoted ethical behavior, training its individual home countries are run.
salespeople to avoid illegal acts and per- Apart from CPI and difference in C
sonal harm (Gratchev, 2001). Notwith- variables, fewer control variables are in-
standing such encouraging but infre- cluded so as to accommodate the samp
quently reported initiatives, a continu- size. The majority of the results are co
ing challenge for the companies is how sistent with hypotheses la and 2.
to achieve a unified stance when some 4. Highlighting, for example, that
companies can obtain an unfair advan- change in the level of corruption m
tage through corruption and, in that re- have the same impact on FDI as a chang
spect, outperform their competitors. in the tax rate on earned income.
This study has its limitations. It relied
on the perception-based measure of cor- REFERENCES
ruption. The measure is broad and fails
Aharoni, Yair. 1966. The Foreign Invest-
to capture the different forms of corrup-
ment Decision Process. Division of Re-
tion, which can exert varying impacts on
search, Harvard Business School, Bos-
FDI. Including specific indices of corrup-
ton.
tion in future analyses can help uncover
Alam, M. 1995. A Theory of Limits on
important nuances regarding the overall
Corruption and Some Applications.
burden for the investors (e.g. between a
Kyklos, 48: 419-435.
petty and grand corruption). Another
Bardhan, Pranab. 1997. Corruption and
limitation of this study is that it could
Development: A Review of Issues.
not probe into the specific impact of cor-
Journal of Economic Literature, XXXV
ruption on FDI driven by contrasting mo-
(September): 1320-46.
tives (e.g. market vs. asset-seeking FDI).
Beck, Paul & Michael Maker. 1986. A
FDI data classified by projects (reflecting
Comparison of Bribery and Bidding in
different motives) will be required to ad-
Thin Markets. Economic Letters, 20:
dress this issue. Finally, it is worthwhile
1-5.
to examine if the impact of corruption on
Benito, Gabriel & Geir Gripsrud. 1992.
FDI varies with the size of the company,
The Expansion of Foreign Direct
the project, or the nature of the industry
Investments: Discrete Rational Loca-
in question. Addressing those issues is
tion Choices or a Cultural Learning
an important and timely task.
Process? Journal of International Busi-
ness Studies, 3: 461-476.
NOTES
Billington, Nicholas. 1999. The Location
1. Various motives can be present si- of Foreign Direct Investment: An Em-
multaneously (Chudnovsky, Lopez and pirical Analysis. Applied Economics,
Porta, 1995) and blur the distinctions. Jan, 31: 65-80.
Boatright, John. 2000. Ethics and the Economy. Boston: Harvard Business
Conduct of Business. Third Edition. School Publishing.
New Jersey: Prentice Hall. Drabek, Zdenek & Warren Payne. 1999.
Braguinsky, Serguey. 1996. Corruption The Impact of Transparency on For-
and Schumpeterian Growth in Differ- eign Direct Investment. Staff Working
ent Economic Environments. Contem- Paper ERAD-99-02, Geneva, World
porary Economic Policy, XIV, July: 14- Trade Organization.
25. Dunning, John. 1988. The Eclectic Para-
Brainard, Lael. 1997. An Empirical As- digm of International Production: A
sessment of the Proximity-Concentra- Restatement and Some Possible Exten-
tion Tradeoff Between Multinational sions. Journal of International Busi-
Sales and Trade. American Economic ness Studies, 19 (Spring): 1-31.
Review, 87: 520-544. . 1998. Location and the Multina-
Busse, Laurence, Noboru Ishikawa, Mor- tional Enterprise: A Neglected Factor?
gan Mitra, David Primmer, Kenneth Journal of International Business
Surjadinata & Tolga Yaveroglu. 1996. Studies, 29 (1): 45-66.
The Perception of Corruption: A Mar- Engwall, Lars & M. Wallenstal. 1988. Tit
ket Discipline Approach. Working Pa- for Tat in Small Steps: The Interna-
per, Emory University, Atlanta, GA. tionalization of Swedish Banks. Scan-
Caves, Richard. 1996. Multinational En- dinavian Journal of Management, 4 (3/
terprise and Economic Analysis. Sec- 4): 147-55.
ond Edition. Cambridge: Cambridge Ensign, Prescott. 1996. An Examination
University Press. of Foreign Direct Investment Theories
Chudnovsky, D., Andre Lopez, & Fer- and the Multinational Firm: A Busi-
nando Porta. 1995. New Foreign Direct ness/Economics Perspective. In M.
Investment in Argentina: Privatiza- Green & R. McNaughton, editors, The
tion, the Domestic Market, and Re- Location of Foreign Direct Investment,
gional Integration. In Agosin, Manuel, Avebury: Aldershot.
editor, Foreign Direct Investment in Ghemawat, Pankaj. 2001. Distance Still
Latin America. Washington, D.C.: In- Matters: The Hard Reality of Global
ter-American Development Bank and Expansion. Harvard Business Review,
the Johns Hopkins University Press. September: 3-11.
Coase, Ronald. 1979. Payola in Radio Gratchev, Mikhail. 2001. Forethought:
and Television Broadcasting. Journal Making the Most of Cultural Differ-
of Law and Economics, 22: 269-328. ences. Harvard Business Review,
Cyert, Richard & James March. 1963. A October: 2-3.
Behavioral Theory of the Firm. Engle- Grosse, Robert & Len Trevino. 1996. For-
wood Cliffs, NJ: Prentice Hall. eign Direct Investments in the United
Davidson, William. 1980. The Location States: An Analysis by Country of Or-
of Foreign Direct Investment Activity: igin. Journal of International Business
Country Characteristics and Experi- Studies, 27(1): 139-55.
ence Effects. Journal of InternationalGupta, Sanjeev, Hamid Davoodi & Rosa
Business Studies, 11(2): 9-22. Alonso-Terme. 1998. Does Corruption
Doz, Yves, J. Santos, & P. Williamson. Affect Income Inequality and Poverty?
2001. From Global to Metanational: IMF Working Paper, International
How Companies Win in the Knowledge
Monetary Fund, Washington D.C.
Hair, Joseph, Rolph Anderson, Ronald Jun, K. & H. Singh. 1996. The Determi-
Tatham & William Black. 1992. Multi- nants of Foreign Direct Investment in
variate Data Analysis. New York, NY:Developing Countries. Transnational
Macmillan. Corporation, 5(2): 67-105.
Hengeveld, W.A.B. 1996. World Dis- Kaufmann, Daniel & Shang-Jin Wei.
tance Tables, 1948-1974 [Computer 1999. Does "Grease Money" Speed Up
File]. Ann Arbor, MI: ICPSR. the Wheels of Commerce? Policy Re-
Hines, James. 1995. Forbidden Pay- search Working Paper 2254, The
ments: Foreign Bribery and American World Bank.
Business After 1977. Working Paper Kobrin, Steven. 1976. The Environmen-
5266, National Bureau of Economic tal Determinants of Foreign Direct
Research, Cambridge. Manufacturing Investment: An Ex-
Husted, Bryan. 1994. Honor Among post Empirical Analysis. Journal of In-
Thieves: A Transaction-Cost Interpre- ternational Business Studies, 7(1): 29-
tation of Corruption in the Third 42.
World. Business Ethics Quarterly, 4(1): Kogut, Bruce & Harbir Singh. 1988. The
17-27. Effect of National Culture on the
.1999. Wealth, Culture, and Cor- Choice of Entry Mode. Journal of Int
ruption. Journal of International Busi- national Business Studies, 19(3): 411-
ness Studies, 30(2): 339-360. 432
ILO. 2001. Yearbook of Labor Statistics. Lambsdorff, Johann. 1999. The Transpar-
International Labor Organization Pub- ency International Corruption Percep-
lications. tions Index 1999-Framework Docu-
IMF. 2000. International Financial Sta- ment, www.transparency.org.
tistics. (CD-ROM). Washington, LaPalombara,
D.C.: Joseph. 1994. Structur
International Monetary Fund. and Institutional Aspects of Corrup
Johanson, Jan & Jan-Erik Vahlne. 1977. tion. Social Research, 61(2): 325-350.
The Internationalization Process of the Leff, Nathaniel. 1964. Economic Devel-
Firm-A Model of Knowledge Devel- opment Through Bureaucratic Corrup-
opment and Increasing Foreign Market tion. American Behavioral Scientist,
Commitment. Journal of International 8-14.
etta 14-15th of June, 1994. Strasbourg: Shleifer, Andrei & RobertVishny. 1993.
Council of Europe Publishing. Corruption. Quarterly Journal of Eco-
Mauro, Paolo. 1995. Corruption and nomics. 108(3): 599-617.
Growth. The Quarterly Journal of Eco- Studenmund, A. H. 1992. Using Eco-
nomics, August: 686-706. nometrics: A Practical Guide. New
Nigh, Douglas. 1986. Political Events York, NY: Harper Collins.
and the Foreign Direct Investment Tanzi, Vito. 1998. Corruption Around
Decision: An Empirical Examination. the World: Causes, Consequences,
Managerial and Decision Economics, Scope, and Cures. IMF Working Pape
7: 99-106.
WP/98/63, International Monetar
Nohria, Nitin & Carlos Garcia-Pont.
Fund, Washington, D.C.
1991. Global Strategic Linkages and
TI. 1999. www.transparency.org/
Industry Structure. Strategic Manage- documents/cpi.
ment Journal, 12 (Special Issue): 105-
Tse, David, Yigang Pan & Kevin Au.
24.
1997. How MNCs Choose Entry Modes
Nordstrom, Kjell & Jan-Erik Vahlne.
and Form Alliances: The China Expe-
1992. Is the Globe Shrinking? Psychic
Distance and the Establishment of
rience. Journal of International Busi-
ness Studies, 28(4): 779-805.
Swedish Sales Subsidiaries During the
Wei, Shang-Jin. 2000. How Taxing is
Last 100 years. International Trade
and Finance Association's Annual Corruption on International Investors?
The Review of Economics and Statis-
Conference, April 22-25, Laredo,
Texas.
tics, 82(4): 1-12.
OECD. 2000. International Direct Invest-
Wells, Louis & Alvin Wint. 2000. Mar-