Crime, Law & Social Change 41: 179–194, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands. 179
Democracy and political corruption: A cross-national comparison
The National Center on Addiction and Substance Abuse, Columbia University, New York,
NY 10017, USA (e-mail: firstname.lastname@example.org)
Abstract. Past research on democracy and political corruption produced mixed results be-
cause of differences in sampling and analytical methods. Moreover, an important shortcoming
has been researchers’ focus on detecting linear effects alone. In the current study, I statistically
controlled for potentially confounding economic factors and used hierarchical polynomial re-
gression to evaluate the form of the democracy-corruption relationship. Results showed that a
cubic function best ﬁtted the data. Despite eruptions of corruption among intermediate demo-
cracies, the consolidation of advanced democratic institutions eventually reduced corruption.
Ultimately, the initial political conditions and the ﬁnal democratic achievements determined
the magnitude of political corruption in a country.
That democratization inﬂuences political corruption in a profound way is an
indisputable truism, yet the directions of the impact of democratic reforms
on incidence of corruption remain hotly contested. Research ﬁndings on the
relationship between democracy and corruption have even been described as
simply “contradictory” (Harris-White & White, 1996, p.3). The main reason
for the disagreement among scholars resides in the multidimensionality of
the concept of “democracy” or “democratization.” Whereas certain aspects
of the democratic process – such as party-based competition and free elec-
tions – motivate unscrupulous politicians to prevail over their rivals through
vote buying or illegal party ﬁnancing (Little, 1996; Johnston, 1997; della
Porta and Vannucci, 1999), the protection of freedom of speech nurtures a
professional investigative journalism that exposes and deters corrupt public
dealings (Giglioli, 1996; da Silva, 2000). By the same token, the defense of
civil liberties and the materialization of an independent judiciary – key ele-
ments that deﬁne a “liberal” democracy – can restrain corruptive inﬂuences,
and maximize the efﬁcacy of anti-corruption campaigns, respectively (Rose-
Ackerman, 1999; Schwartz, 1999; Jamieson, 2000; Moran, 2001). Such is
the confused state of affairs, that a recent review of the issue concluded that
“democratization is in practice a complex phenomenon with unpredictable
effects” (Moran, 2001: p. 390).
When insightful qualitative studies found democracy to be a multifaceted
process affecting corruption in numerous and sometimes conﬂicting manners,
180 H.-E. SUNG
statistical analyses have mostly detected a linear and negative democracy-
corruption association (Goldsmith, 1999; Sandholtz and Koetzle, 2000; see
Treisman’s work (2000) for exception). This ﬁnding is both assuring and
disturbing. On the one hand, the positive statistical link between ﬁrmer demo-
cratic institutions and lower levels of corruption seems to corroborate the
widely accepted belief about the ability of democratic reforms to bring about
transparent and accountable governance. On the other hand, it ﬂies in the face
of numerous observations of renewed corrupt practices induced by political
liberalization in Southeast Asia, Latin America, and former Soviet republics
(Cohen, 1995; Harris-White and White, 1996). This conceptual quagmire
shows that corruption theory and research are still in their infancy.
What proves clearer is the anti-corruption efﬁcacy of improved economic
performance in many countries that have implemented broad economic or
market reforms along with democratization. Independent of labor productiv-
ity and economic output, greater democratic achievements lead to higher
wages (Goldsmith, 1995; Rodrik, 1999), which in turn reduce incentives and
opportunities for corruption among elected and appointed ofﬁcials (Sand-
holtz and Koetzle, 2000; Van Rijckeghem and Weder, 2001). Also, direct
democracy impels monetary, ﬁscal, and social policy-making toward fuller
employment, greater welfare spending, and a bigger public sector (Barro,
1997; Frey and Stutzer, 2000; Boix, 2001). As citizens’ participation in polit-
ical processes brings micro- and macroeconomic outcomes closer to voters’
preferences, life satisfaction increases and corruption decreases. This indir-
ect check of democracy on corruption via material wellbeing can easily be
confounded with the process democratization itself in empirical evaluations,
if no adequate controls are in place.
In the study presented in this article, I examined the relationship between
democracy and corruption in various cross-sections of countries. The goal
was to empirically isolate the democratization process from its economic
correlates, which have been known to inﬂuence corruption opportunities, and
to determine the direction and magnitude of its impact on political corruption.
Results from this analysis provided the ﬁeld of corruption research with some
of the empirical discipline needed to make the debate policy-relevant.
Exploring the democracy-corruption link
If very few relationships in political science are exactly linear in shape, demo-
cracy and democratization are nonlinear phenomena par excellence. Mobil-
izing public preferences to determine who gets what, when, and how, and
building functioning institutions to process them, are such a complex en-
deavor that studies on democracy and economic growth, income distribution,
DEMOCRACY AND CORRUPTION 181
and war-making all reported curvilinear functions as better solutions (Barro,
1997; Burkhart, 1997; Hegre, Ellingsen and Gates, 2001). The degree of
non-linearity in these relationships was so evident that linear equations were
considered inadequate even for approximation.
Unfortunately, the few statistical analyses of the democracy-corruption
connection only assumed, and thus found almost exclusively, a linear corres-
pondence between the two variables (Goldsmith, 1999; Sandholtz and Koet-
zle, 2000). The study that did not ﬁnd signiﬁcant relationship between demo-
cracy and corruption, only tested the linear hypothesis (Treisman, 2000).
Although this simplistic linear approach was able to uncover statistically
signiﬁcant results, it remains unable to accommodate a larger amount of qual-
itative data that contradicts the democracy-reduces-corruption axiom. Ethno-
graphic or country-speciﬁc studies completed in the 1990s emphasized how
democratization may actually increment the opportunities and magnitude of
corruption without strengthening countervailing institutional capacities. The
radical embrace of democratic rules cast discredit on previous authoritarian,
albeit sometimes effective, controls over corruption without providing an al-
ternative set. Incentives for corrupt behavior in emerging democracies stem
from the enormous costs of mounting electoral campaigns, the luring attract-
iveness of public assets and inﬂuence, and the openness of the state apparatus
to the ambitions of elite groups.
Far from reducing corruption in the short term, political liberalization
has made matters worse in most of those countries that embarked on the
democratic transition in the 1980s and 1990s. Once-celebrated reborn demo-
cracies like Argentina, Philippines, and Russia have become prototypes of
poor governance. However, social scientists refuse to give up on democracy.
They argue that corruption may well be a transitional phenomenon common
in ﬂedgling democracies where procedural practices have not been under-
pinned by a ﬁrm liberal culture and effective institutions (Harris-White and
White 1996; Rose-Ackerman, 1999). Democratization opens up a long-term
prospect of institutional remedies for corruption that require sustained ef-
forts to make them work successfully. This view has recently been supported
by Gabriella R. Montinola and Robert W. Jackman’s cross-national studies
(2002), which hinted toward a nonlinear relationship between political demo-
cracy and corruption. Their analysis of corruption data from the 1980s found
that corruption was typically lower in dictatorships than in partial democra-
cies, but once attained a threshold, democratic practices suppress corruption.
Nevertheless, the small size of the sample (n = 66), the unstable statistical
signiﬁcance of the key ﬁndings, and the aged data used in the analysis render
Montinola and Jackman’s conclusions very informative but of limited gen-
182 H.-E. SUNG
eralizability. Most importantly, only one type of nonlinear hypothesis – the
quadratic form – was considered in that study.
Following the same thread of exploration, the study presented in this art-
icle tested three statistical equations (linear, quadratic, and cubic) using recent
data covering a larger number of countries to assess the link between demo-
cratization and corruption. These simple equations represent major forms
of statistical relationships in social sciences. The linear model predicts a
straightforward positive or negative correspondence between democracy and
corruption and can be expressed as corruption = f (democracy). The quadratic
equation, ascribed by Montinola and Jackman, hypothesizes that as demo-
cratization progresses, political corruption ﬁrst increases and then decreases,
or ﬁrst decreases and then increases. It can be represented as corruption =
f (democracy, democracy2). And lastly, the cubic model anticipates that as
democratic institutions consolidate, levels of corruption can either increase,
then decrease, then increase, or decrease, then increase, then decrease. This
polynomial function is expressed as follows: corruption = f (democracy,
Instead of focusing on individual components of democratic polity, I ad-
opted Moran’s comprehensive deﬁnition that democratization is a process
typiﬁed by “the vote as a basis for constitutional mechanisms for the transfer
of power; political competition through parties; guaranteed individual liber-
ties; freedoms to form public organizations and private organizations” (2001:
pp. 379–380). Despite the broadness of this approach, Moran’s study concep-
tually and empirically distinguished the liberalization of the political system
itself from its economic outcomes. Finally, by comparing a large collection
of countries, my analysis generated ﬁndings of very high generalizability that
transcend the unique features and histories of single societies.
Data, variables and methods
The study sample comprised 103 countries that have complete information
on both the Political Rights Index compiled by the Freedom House and the
1995–2000 annual Corruption Perceptions Index (CPI) compiled by Trans-
parency International, an international non-governmental organization de-
voted to measuring and combating corruption with 80 international chapters
around the world (see Appendix). The Freedom House indicator of demo-
cracy measures citizens’ rights to vote, efﬁcacy of elected representatives,
multi-party system, meaningful opposition, and exercise of sovereignty over
national territories (Karatnycky, 2001), thus it operationally deﬁnes a valid
and holistic construct of democracy. The CPI is a secondary rating system
compiled from a number of surveys of business executives’ perceptions of
DEMOCRACY AND CORRUPTION 183
Table 1. Description of variables (N = 416)
Name Description and data source Minimum Maximum Mean Std.
Inverted Corruption Perceptions
Index, –10 = extremely
non-corrupt; 0 = extremely
–10.00 –.69 –5.01 2.49
Inverted Political Rights Index,
–7 = strongly undemocratic;
–1 = strongly democratic
–7.00 –1.00 –2.59 1.90
Purchasing power parity
(International Monetary Fund)
.01 21.73 1.33 3.03
% work force unemployed from
(The Economist Intelligence
.40 27.50 8.64 5.14
Inﬂation rate Inﬂation rate (International
–8.50 325.00 11.86 28.28
corruption among politicians and public ofﬁcials conducted by different or-
ganizations (Transparency International, 2002). Both indicators were inverted
(by multiplying by –1) in this study, for readability. Measures of purchas-
ing power, unemployment, and inﬂation were incorporated in multivariate
analyses to control for the potential disturbances from economic hardships.
These indicators were obtained from the International Monetary Fund (2002)
and The Economic Intelligence Unit (2002). The resulting time-series cross-
section data contained 520 country-year units.
Since bivariate measures such as the Pearson correlation do not capture
curvilinear relationships, only multivariate tests were conducted. As the
pooled Durbin-Watson dstatistics for the linear model of 1.83 indicated the
absence of a serious problem of serially correlated errors, there was no need
to implement additional procedures such as adding a lagged dependent vari-
able in the regression equations or transforming the error terms to correct
for this common problem in pooled time series data (Beck and Katz, 1995).
Only year-speciﬁc dummy variables were included to correct for unique dis-
turbance effects in time (Sayrs, 1989). Polynomial regression results were
analyzed to judge the form and magnitude of the relationship. Because the
coefﬁcient of multiple determination (R2) increases with each addition of
184 H.-E. SUNG
Figure 1. Scatterplot showing relationship between democracy and political corruption.
predictors, adjusted R2were presented, although discussion of ﬁndings was
solely based on unadjusted coefﬁcients.
All three models yielded statistically and substantively signiﬁcant goodness
of ﬁt, although the coefﬁcients of multiple determination notably differed in
size. In other words, while all models were useful in interpreting changes
in the levels of corruption, some achieved considerably larger reductions in
prediction errors and thus were more powerful in interpreting the vicissitudes
of democracy-corruption connections.
The linear model accounted for 45% of the variation in political corruption
and detected a statistically powerful negative effect of democracy on cor-
ruption. Controlling for disturbances from the economic factors, a one-point
increase in democracy index was associated with a .737-point decrease in the
scale of political corruption. This association is stout by conventional social
science standards. Economic factors also made an inﬂuential contribution to
the explanatory power of the model: unemployment and inﬂation positively
and strongly correlated with the outcome variable, conﬁrming the thesis that
economic hardships foster dishonest practices among government ofﬁcials.
The quadratic model improved the explanatory power of the equation by
5 percentage points, raising the R2to .516. In other words, the inclusion
of the second-order polynomial term enhanced the model’s goodness of ﬁt
by 14%. The two democracy variables retained the same statistical signiﬁc-
ance level of .001. An unmistakable answer existed regarding the form of
this curvilinear relationship. The negative sign attached to the second-degree
DEMOCRACY AND CORRUPTION 185
Table 2. Results from hierarchical polynomial regression (N = 416)
Linear model Quadratic model Cubic model
Purchasing power parity –.016 .020 .005
(.035) (/033) (.030)
Unemployment rate .122*** .114*** .121***
(.020) (.019) (.017)
Inﬂation rate .001* .001* .008*
(.004) (.004) (.003)
Democracy index –.737*** –2.210*** –7.881***
(.056) (.231) (.672)
Democracy index: quadratic term — –.207*** –2.038***
Democracy index: cubic term — — –.163***
Year 1995 –.640 –.610 –.685
(.391) (.368) (.331)
Year 1996 –.654 –.241 –.326
(.357) (.336) (.302)
Year 1997 –.153 .021 –.128
(.364) (.344) (.309)
Year 1998 –.200 –.156 –.214
(.310) (.291) (.262)
Year 2000 –.117 –.075 .025
(.305) (.287) (.208)
R2.452*** .516*** .611***
(Adjusted R2) (.437***) (.501***) (.598***)
Incremental R2.450*** .064*** .095***
∗p<.05; ∗∗ p<.01; ∗∗∗ p<.001.
NOTE: Regression coefﬁcients and standard errors (between parentheses) are reported.
186 H.-E. SUNG
polynomial democracy indicator revealed that a concave function better ﬁt-
ted the data than the simple linear function. This mound-shaped relationship
changed its direction at the point X=–βdemocracy/2βdemocracy squared (Agresti
and Finlay, 1997), which was the point –5.34 on the democracy scale for
this sample. Undemocratic countries with extremely low scores on the demo-
cratic index experienced an increase in political corruption in the early stage
of democratization until they reached the –5.34 point, at which the average
score of corruption reached its maximum (see Figure 1). Beyond this critical
point, countries began to reap very important anti-corruption beneﬁts from
strengthening their democratic institutions, as the steady downward slope
indicated. The slope also implied that transitional countries with an initial
democracy index score above the critical point were likely to see immediate
anti-corruption effects as soon as political reforms took place. The overall
long-term trend of the entire process might resemble the downward slope
portrayed by the simple linear function, but the quadratic function was able
to discriminate experiences of undemocratic countries from the performances
of highly democratic countries or democratizing countries with better initial
Best regression results were observed for the cubic model, whose R2of
.611 increased the same coefﬁcients for the linear and quadratic equations
by 35% and 18% respectively. The regression coefﬁcients showed a negative
linear term, negative squared term, and negative cubed term, which meant
that, beginning at the origin (which is negative due to the inversion of the
democracy and corruption index scores), the function ﬁrst tended downward,
then upward, then downward again (see Figure 1). The association began
with a convex shape whose curvature changed at the point of inﬂection (Y=
–βdemocracysquared/3βdemocr acy cubed = –4.17), to be continued by a concave
shape. The average corruption index in the convex range reached its min-
imum at the democracy index score –5.50 (Zmin =Y+(sqrt(–3βdemocracy cubed
βdemocracy squared)/3d emocracy cubed )), whereas the average corruption score in
the following concave range was at its maximum at the democracy score
–2.74 (Zmax =Y-(sqrt(–3βd emocracy cubed βd emocracy squared)/3d emocracy cubed )).
Therefore, in general terms, three moments distinguished this cubic relation-
ship: an initial negative democracy-corruption connection among undemo-
cratic countries with a democracy index score lower than –5.50, followed
by a positive association among partly democratic countries with democracy
scores between –5.50 and –2.74, and ﬁnally a steep downward slope among
advanced democracies beyond the last critical point.
The most undemocratic countries (rated –7 on the democracy scale) were
likely to experience a slight improvement in the transparency and accountab-
ility of governance during the very beginning stage of their democratization.
DEMOCRACY AND CORRUPTION 187
But as they entered a fairly lengthy intermediate phase of the process, they
tended to suffer worsening corruption as new values and institutions took root
amidst continuous power realignments and resource reallocation. For coun-
tries that began their democratization journey at an intermediate level of polit-
ical liberalization (rated between –6 and –3), a growing problem of corruption
was a common occurrence. The persistence and advancement of undemo-
cratic countries and intermediate democracies in their progression toward
a ﬁrmer and more mature democratic polity yielded substantial reductions
in corruption once their democracy scoring reached –2. The achievement of
honest and efﬁcient government was even more substantial among developed
democracies at the highest rank of –1. This curvilinear relationship clearly
implies that the manner in which democracy affected corruption depended
on the initial democratic conditions as well as on the eventual democratic
attainments of each country. Greatest rewards (in the form of a clean and
transparent state) were granted to countries that were able not only to realize
but also to maintain the strongest and healthiest democratic institutions.
Just as the different pictures painted by qualitative studies and statistical re-
search on democracy and corruption complement each other, ﬁndings from
the three models tested in this study shed more light on the problem if in-
terpreted side by side. Democratization generally, and eventually, decreases
corruption (as evidenced by the valid linear model). But it should be re-
cognized that temporary upsurges in government corruption are to be ex-
pected during the early stages of the process of political liberalization (as
demonstrated by the stronger quadratic model). Yet most importantly, it is
the initial conditions and the ﬁnal achievements of each society, rather than
the democratization process itself, that determine the shape and magnitude of
the impact of democratic reforms on political institutions.
Although democracy and market reform are partly meant to clean up pub-
lic life in developing countries, corruption proves resistant to intervention
efforts. For countries that depart from a long history of authoritarianism or
totalitarianism, the enthusiasm and optimism aroused by the early expansion
of political rights are often cooled down by an alarming growth in corruption
and “money politics” (Moran, 2001). The massive scale of political bickering
and state restructuring (including the privatization of state-owned enterprises)
make democratization open to suspicion. The ability to check on the power
of the elites and to garner enough support for deepening institutional reforms
(i.e. erecting an independent judiciary, and protecting freedom of speech,
which fosters an investigative journalism) then become the most critical tasks
188 H.-E. SUNG
the government, the business community, and the civil society must together
undertake. Escalation of corruption in democratizing countries is a product
of a changing environment as well as the result of illiberal practices (Rose-
Ackerman, 1999; Bunce, 2000). When efforts at consolidating checks and
balances of state powers and establishing a ﬁrmer rule of law are aborted,
more often than not by corrupt elite groups (Mbaku, 1995; Weyland, 1998;
Fairbanks, 2001), the country risks drifting back into authoritarianism or
away into a kleptocracy.
Findings from this study make a relevant contribution to the stagnant
debate about whether international donors and lenders should promote the lib-
eral democratic form of political organization when attaching anti-corruption
requirements to their assistance packages (Marquette, 2001). The stalemate
was mainly caused by the lack of consensus on the link between democracy,
development, and corruption. As evidence points to a descending curvilinear
democracy-corruption relationship, international agencies should recognize
the role of the political system in determining the efﬁcient utilization of for-
eign assistance; and they should encourage aid recipients to broaden citizens’
participation in political processes, to create an environment friendly to the
successful implementation of anti-corruption policies.
DEMOCRACY AND CORRUPTION 189
Year-country units with complete democracy and corruption data included in
Country 1995 1996 1997 1998 1999 2000
Armenia X X
Azerbaijan X X
Belarus X X X
Botswana X X X
Bulgaria X X X
Cameroon X X X X
190 H.-E. SUNG
Appendix – continuation
Country 1995 1996 1997 1998 1999 2000
Cote d’Ivoire X X X
Croatia X X
Czech Republic X XXXXX
Ecuador X X X X
Egypt X X X X
El Salvador X X X
Estonia X X X
Ghana X X X
Greece X XXXXX
Guatemala X X
Honduras X X
Hungary X XXXXX
Iceland X X X
Jamaica X X
Jordan X X X X
Kazakhstan X X
Kenya X X X X
Kyrgyz Republic X
DEMOCRACY AND CORRUPTION 191
Appendix – continuation
Country 1995 1996 1997 1998 1999 2000
Latvia X X X
Lithuania X X
Luxembourg X X X X X
Malawi X X X
Mauritius X X X
Moldova X X
Morocco X X X
Mozambique X X
Namibia X X X
Nicaragua X X
Paraguay X X
Peru X X X
Senegal X X X
Slovakia X X X
Slovenia X X
192 H.-E. SUNG
Appendix – continuation
Country 1995 1996 1997 1998 1999 2000
Tanzania X X X
Tunisia X X X
Uganda X X X X
Ukraine X X X
United Kingdom X XXXXX
Uruguay X X X
Uzbekistan X X
Vietn am X X X
Yugoslavia X X X
Zambia X X X
Zimbabwe X X X
Agresti, A. and B. Finlay, Statistical Methods for the Social Sciences, 3rd edn. (Upper Saddle
River, NJ: Prentice Hall, 1997).
Barro, R.J., Determinants of Economic Growth: A Cross-Country Empirical Study (Cam-
bridge, MA: The MIT Press, 1997).
Beck, N. and J. Katz, “What to Do (and not to Do) with Times-Series Cross-Section Data,”
American Political Science Review 1995 (89), 634–647.
Boix, C., “Democracy, Development, and the Public Sector,” American Journal of Political
Science 2001 (45), 1–17.
Bunce, V., “Comparative Democratization: Big and Bounded Generalizations,” Comparative
Political Studies 2000 (33), 703–734.
Burkhart, R.E., “Comparative Democracy and Income Distribution: Shape and Direction of
the Causal Arrow,” Journal of Politics 1997 (59), 148–164.
Cohen, A., Crime and Corruption in Eurasia: A Threat to Democracy and International
Security (Washington, DC: The Heritage Foundation, 1995).
da Silva, C.E.L., “Journalism and Corruption in Brazil,” in J.S. Tulchin and R.H. Espach (eds.),
Combating Corruption in Latin America (Washington, DC: Woodrow Wilson Center
Press, 2000) pp. 173–192.
DEMOCRACY AND CORRUPTION 193
della Porta, D. and A. Vannucci, Corrupt Exchanges: Actors, Resources, and Mechanisms of
Political Corruption (New York: Aldine de Gruyter, 1999).
Fairbanks, C.H., Jr., “Disillusionment in the Caucasus and Central Asia,” Journal of Demo-
cracy 2001 (12), 49–56.
Frey, B.S. and A. Stutzer, “Happiness, Economy and Institutions,” The Economic Journal
2001 (110), 918–938.
Giglioli, P.P., “Political Corruption and the Media: The Tangentopoli Affair,” International
Social Science Journal 1996 (48), 381–394.
Goldsmith, A.A., “Democracy, Property Rights, and Economic Growth,” Journal of Develop-
ment Studies 1995 (32), 157–174.
Goldsmith, A.A., “Slapping the Grasping Hand: Correlates of Political Corruption in Emer-
ging Markets,” American Journal of Economics and Sociology 1999 (58), 865–883.
Harris-White, B. and G. White (eds.), Liberalization and New Forms of Corruption (Brighton:
Institute of Development Studies, 1996).
Hegre, H., T. Ellingsen and S. Gates, “Toward a Democratic Civil Peace? Democracy, Political
Change, and Civil War, 1816–1992,” American Political Science Review 2001 (95), 33–48.
International Monetary Fund, “The World Economic Outlook Database April 2002,” [On-line
database]. 2002, Available
Jamieson, A., The Antimaﬁa: Italy’s Fight against Organized Crime (New York: Palgrave
Johnston, M., “Public Ofﬁcials, Private Interests, and Sustainable Democracy: When Polit-
ics and Corruption Meet,” in K.A. Elliot (ed.), Corruption and the Global Economy
(Washington, DC: Institute for International Economics, 1997) pp. 61–82.
Karatnycky, A. (ed.), Freedom in the World: The Annual Survey of Political Rights and Civil
Liberties, 2000–2001 (Somerset, NJ: Transaction, 2001).
Little, W., “Corruption and Democracy in Latin America,” IDS Bulletin 1996 (27), 64–70.
Marquette, H., “Corruption, Democracy and the World Bank,” Crime, Law and Social Change
2001 (36), 395–407.
Mbaku, J.M., “Post-Independence Opportunism and Democratization in Africa,” Journal of
Social, Political and Economic Studies 1995 (20), 405–422.
Montinola, G.R. and R.W. Jackman, “Sources of Corruption: A Cross-National Study,” British
Journal of Political Science 2002 (32), 147–170.
Moran, J., “Democratic Transitions and Forms of Corruption,” Crime, Law and Social Change
2001 (36), 379–393.
Rodrik, D., “Democracies Pay Higher Wages,” Quarterly Journal of Economics 1999 (114),
Rose-Ackerman, S., Corruption and Government: Causes, Consequences, and Reform (New
York: Cambridge University Press, 1999).
Sandholtz, W. and W. Koetzle, “Accounting for Corruption: Economic Structure, Democracy,
and Trade,” International Studies Quarterly 2000 (44), 31–50.
Sayrs, L.W., Pooled Time Series Analysis, Quantitative Applications in the Social Sciences No
70 (Thousand Oaks, CA: Sage, 1989).
Schwartz, H., “Surprising Success: The New Eastern European Constitutional Courts,” in A.
Scheller, L. Diamond and M.F. Plattner (eds.), The Self-Restraining State: Power and
Accountability in New Democracies (Boulder, CO: Lynne Rienner Publishers, 1999) pp.
194 H.-E. SUNG
The Economist Intelligence Unit, EIU Country Data [On-line database]. 2001, Available
Transparency International, Corruption Perceptions Index [On-line database]. 2002, Available
Treisman, D., “The Causes of Corruption: A Cross-National Study,” Journal of Public
Economics 2000 (76), 399–457.
Van Rijckeghem, C. and B. Weder, “Bureaucratic Corruption and the Rate of Temptation: Do
Wages in the Civil Service affect Corruption, and by how Much?” Journal of Development
Economics 2001 (65), 307–331.
Weyland, K., “The Politics of Corruption in Latin America,” Journal of Democracy 1998 (9),