Does Foreign Aid Promote Democracy?
Introduction
Stephen Knack of The World Bank examines the impact of foreign aid on democratization in recipient nations during 1975-2000.
The study uses multivariate analysis and various democracy indexes and aid intensity measures.
The conclusion reached is that there is no evidence that foreign aid promotes democracy.
The analysis considers alternative model specifications, estimation techniques, and the post-Cold War period (1990-2000).
Foreign aid can potentially contribute to democratization through:
Technical assistance for electoral processes.
Strengthening legislatures and judiciaries.
Promoting civil society organizations.
Conditionality.
Improving education and per capita incomes.
However, aid could also undermine accountability, encourage conflict, and promote coup attempts.
The study analyzes democracy-building efforts of bilateral donors and reviews literature on the negative effects of aid on democratization and governance.
The data used in the analysis are described, and the implications of the findings are discussed.
Background
Many foreign aid donors prioritize the promotion of democratic government.
USAID spends over $700 million annually on democracy-related programs, including:
Elections support.
Parliament and judiciary strengthening.
Political party development.
Fostering civil society organizations.
Technical assistance for democracy promotion is a small fraction of total aid.
Donors often condition general budget support on civil liberties, elections, and rule of law.
The Foreign Assistance Act of 1975 linked U.S. aid to respect for human rights and civil liberties.
Aid can indirectly promote democracy by modernizing societies through increased literacy and incomes.
Empirical studies link education and per capita income to democratization.
The World Bank commits resources to "good governance" programs that strengthen accountability and the rule of law.
Effective technical assistance and conditionality should associate higher aid levels with increased democracy.
USAID claims credit for the spread of democracy, noting an increase from 58 democratic nations in 1980 to 115 in 1995.
However, correlation does not equal causation.
Conditioning aid on reform is generally ineffective.
The U.S. provided 11.4 billion in Official Development Assistance (ODA) in 2001.
Total ODA by all donors in 2001 was 52.3 billion, down from 58.9 billion in 1995.
Arguments Against Foreign Aid
Scholars and practitioners argue that foreign aid can undermine democratic government.
Friedman (1958) argued that aid strengthens the government sector relative to the private sector, which is detrimental to civil liberties and democracy.
Aid may weaken governmental accountability by retarding the development of civil society.
The evolution of democracy in the West was related to monarchs' need for tax revenues and elites demanding accountability.
Foreign aid may short-circuit these processes by reducing government dependence on citizens for tax revenues.
Recipient governments become accountable to foreign donors rather than taxpayers.
High aid levels may reinforce executive dominance by funding off-budget projects, precluding parliamentary review.
The International Financial Institution Advisory Commission (2000) noted that IMF conditionality undermines sovereignty and democratic processes.
Aid may encourage coup attempts and political instability by making control of government and aid receipts a valuable prize.
Violent competition for food aid contributed to government breakdown in Somalia.
Comprehensive analyses are needed to determine aid's impact on democracy.
This study examines the issue empirically using cross-country data from 1975–2000.
There is a growing body of literature on aid's effectiveness in promoting economic development.
Svensson (1999) found that aid's impact on income growth is higher in more democratic nations.
Isham, Kaufmann, and Pritchett (1997) found higher economic rates of return on World Bank projects in countries with higher civil liberties scores.
Few cross-country studies have addressed aid's impact on democracy and governance.
Knack (2001) found that higher aid levels worsen corruption and bureaucratic quality.
Svensson (2000) found aid is associated with greater corruption in ethnically heterogeneous countries.
Crawford (1997) analyzed 29 cases of politically motivated aid sanctions and found limited success.
This study finds no evidence that aid has promoted or undermined democracy in recipient nations.
Studies on human rights' protection and aid allocations have mixed findings and do not correct for reverse causation.
The empirical work below corrects for the potential endogeneity of aid to democratization.
Data and Hypotheses
Freedom House provides annual ratings on political freedoms and civil liberties since the early 1970s.
These ratings are used as measures of democracy.
Countries are rated on a 1–7 scale, with higher values indicating less freedom.
The scales are reversed so that a higher value indicates a country is more democratic.
The political freedoms index better approximates standard definitions of democracy than the civil liberties index, but the latter better captures civil society dimensions.
Civil liberties are necessary for free and fair elections and are an integral part of democracy.
A combined index ranging from 2 to 14 in value will be used in analyses.
From 1972 until 1989, Raymond Gastil produced all of the Freedom House ratings.
The judgments and methods of those replacing him may have differed, creating a break in the data.
However, examination of year-to-year correlations in the indexes before, during, and after the transition provide no indication of a break in the data.
Results do not differ when only the post-1989 Freedom House data are used.
All results reported using the Freedom House index are also replicated using an alternative democracy index based on indicators provided in the Polity IV data set (Marshall and Jaggers, 2002).
The Polity IV data set provides a 10-point scale of democracy based on evaluations of how executives are selected, and of whether or not there are any effective institutional checks on their power.
Also included is a 10-point autocracy scale, with higher values representing some combination of less competitive processes for selecting chief executives, and fewer constraints on their authority.
The democracy index used in the analysis is constructed by taking the difference in the Polity IV democracy and autocracy indexes.
The resulting index can vary from a high of +10 to a low of -10. While the Freedom House democracy indexes attempt to reflect rights afforded a country’s population, the Polity index is intended to reflect institutional characteristics of a country’s government, particularly how open and competitive are the processes for selecting chief executives.
The conceptual links between the Freedom House indexes and donors’ democracy-promoting programs are obvious; as noted, donors even use those indexes as indicators of progress.
Aid in the form of support for elections and electoral processes could equally affect the Polity index, by enhancing the openness and competitiveness of elections.
The Freedom House and Polity democracy indicators are in close (but not perfect) agreement, not only on levels but also on changes in democracy over the 1975–2000 period.
For 1975, the Spearman rank correlation is 0.73, and for 1996 levels is 0.89, for the 96 countries included in the aid-democracy tests reported in section 3 below. Changes over the 1975–2000 period are correlated at 0.76 (r = 0.0001).
Two alternative measures of aid intensity or dependence are used in the analysis: "official development assistance" (ODA) as a percentage of GNP, and as a percentage of government expenditures.
Averages over the 1975–1999 period were constructed for aid/GNP and aid as a share of government spending.
Aid as a share of GNP and aid as a share of government expenditures are correlated at 0.86 (r = 0.0001).
Annual data on aid are available from the World Bank’s World Development Indicators, based on data collected by the OECD’s Development Assistance Committee, dating from 1975.
Aid includes grants, and concessionary loans with a grant element of more than 25 percent.
Aid is measured net rather than gross, that is, current loan repayments are subtracted. Military assistance is excluded.
-Table 1. Summary Statistics
-Variable N Median Mean Std. dev. Minimum Maximum
-Freedom House index change 105 +2 +1.9 3.7 –8 +10
-Initial Freedom House 105 6 6.7 3.3 2 14
-Polity index change 97 +5 +5.6 7.5 −14 +17
-Initial Polity index 97 −7 −2.4 7.0 −10 +10
-Annual income growth (%) 105 0.09 0.38 2.95 −11.7 9.3
-Initial log GDP per capita 105 7.86 7.84 0.83 6.34 9.47
-Illiteracy change, 1975–99 (%) 105 −14.6 −16.1 10.2 0.0 −38.2
-Initial illiteracy (%) 105 37.9 40.8 28.6 0.2 93.3
-Infant mortality/1000 births, 1975 105 90.2 88.6 48.8 16.3 90.2
-Log population, 1975 105 15.4 15.5 1.7 11.8 20.6
-Aid/GNP, mean 1975–95 105 4.3 6.7 7.8 0.1 47.3
-Aid/govt. exp., mean 1975–95 96 11.2 23.7 26.9 0.1 105.1
-If aid promotes democracy, then countries with higher aid levels should exhibit improving ratings on the democracy indexes over time, other things equal.
-Accordingly, the dependent variable analyzed here is the end of period (2000) minus the initial period (1975) ratings on the democracy indexes.
-Control variables include the initial democracy index value, initial (log of) per capita income, growth in per capita income from 1975 to 1999, initial levels of illiteracy, the change in illiteracy from 1975 to 1999, and a set of regional dummy variables.
-The initial democracy index captures regression-to-the-mean effects, and controls for the limited opportunity of highly democratic countries to increase their ratings (the Freedom House and Polity indexes respectively have upper-bound values of 14 and 10).
-Socioeconomic development has long been believed to be conducive to the emergence or survival of democracy, so changes in democracy may be related to changes in per capita income.
-Growth of per capita income is measured as the average annual change from 1975 to 1999.
-Literacy is an alternative measure of socioeconomic development emphasized by "modernization" theories of democratization. The measure used here is the change in illiteracy from 1975 to 1999, with larger numbers indicating a larger decline (i.e., improvement) over time.
-There is some evidence that democratization occurs in waves. Regardless of whether or not a nation is "modernizing," its likelihood of democratizing is partly dependent on political changes in its neighbors. The "third wave" of democratization (Huntington, 1991) occurred during the period analyzed here, and involved primarily nations in Sub-Saharan Africa, Latin America and the Caribbean, and in Eastern Europe. Dummy variables for each of these country groups are therefore included as regressors.
Results
Figure 2 displays the low (r = 0.12) and insignificant correlation between the 1975– 2000 change in the Freedom House index and aid as a share of GNP, averaged over 1975–1999.
Multivariate tests confirm that aid on balance has been ineffectual in promoting democracy.
-Table 2 presents these results.
-Equation 1, the dependent variable is the change in the Freedom House democracy index from 1975 to 2000, and the aid variable is aid/GNP averaged over 1975–1999. In equation 2, aid as a share of government spending is substituted for aid/GNP as the aid measure. Equation 3 returns to aid/GNP as the aid measure, but substitutes the Polity democracy index for the Freedom House index. In these regressions, the aid coefficients are small, of varying sign, and never come close to attaining statistical significance.
-Among the control variables, the initial value of the democracy index is the most powerful predictor of changes in the democracy indexes; its coefficient is negative as expected, and highly significant in every regression. Initial income and illiteracy levels are never significant.
-Larger declines in illiteracy, surprisingly, are associated with significant declines in democracy over the period: each 10-point rise in illiteracy is associated with a 1-point rise on the 14-point Freedom House scale (equations 1 and 2). Changes in illiteracy are not significantly related to the Polity index, however (equation 3).
-Economic growth is positively related to democratization as expected, but more often than not this relationship is not significant at the .05 level in Table 2.
-The "third wave" regional dummies are positive and significant in the majority of cases. Of these, the coefficient for the Latin America dummy is the largest and most significant.
-Equation 4 of Table 2 replicates equation 1, but using ordered logit instead of OLS. Ordered logit is an estimation method designed for dependent variables that are ordinal but not interval level. The Freedom House and Polity indexes are typically treated as interval-level measures (e.g., Barro, 1997; Burkhart and Lewis- Beck, 1994), but are technically only ordinal-level scales, despite having up to 25 different possible values (ranging from +12 to -12) in the case of Freedom House, and 41 possible values (+20 to -20) for the Polity index.
-In equation 4, ordered logit estimates confirm the absence of a link between aid and democratization. Equation 5 also replicates equation 1, but using median regression, which reduces the sensitivity of results to outlying values, instead of OLS. Aid still shows no impact on democratization in this test. Equation 6 analyzes changes in the Freedom House index from 1990 to 2000, adjusting aid/GNP and other independent variables to conform to this 10-year period.
-Aid was conceivably more effective in this post–Cold War period, in which donors likely gained leverage relative to recipient nations, which could no longer play the superpowers off against one another. In particular, aid was less likely to be used for propping up pro-U.S. dictators, which may have forestalled democratization during the Cold War. Equation 6 shows, however, that aid is unrelated to democratization in the post–Cold War period. Equation 7 returns to the specification in equation 1, but is estimated using two- stage least squares (2SLS) to correct for possible reverse causation. Endogeneity of aid is potentially a serious problem in studying the impact of aid on democratic transition.
-For example, USAID currently has an explicit policy of directing more aid toward countries that appear to be making greater progress on democratiza- tion. Targeting aid toward countries that appear to be making progress in democratizing would bias aid coefficients in a positive direction in OLS regressions. Even with this potential upward bias, equations 1–6 fail to link aid with democratization.
-However, it is possible that other donors may allocate aid differently in ways that produce a bias in the opposite direction. Accordingly, equation 7 corrects for potential endogeneity by estimating the relationship between democratization and the exogenous component of aid. Exogenous instruments for aid in the first-stage regression include measures of recipient need and of donor interest.
-Infant mortality (in 1975) is a measure of recipient need. Country size, measured as the log of population of the recipient nation in 1975, is a measure of donor interest that strongly predicts aid levels. Smaller nations receive much more aid on average than larger nations, presumably because of donors’ desire to "show the flag" widely. Other measures of donor interest included as instruments are a set of colonial heritage dummies, indicating whether recipient nations are ex-colonies of any one of eight donor nations (Britain, France, Spain, Portugal, Belgium, Italy, the Netherlands, and the U.S.). Collectively these instruments are effective in explaining aid: the first-stage regression of aid/GNP on these instruments and on the other regressors in the democracy equation in Table 2 produces an R2 of .69, compared to only .49 for a regression of aid/GNP on only the other regressors in Table 2. Aid as instrumented by these variables remains an insignificant predictor of changes in democracy in equation 7. This result indicates that the insignificant OLS coefficients on aid are not biased downward from any propensity of donors to target aid toward countries experiencing setbacks in democratization.
-Table 2. Aid Dependence and the Freedom House Index
-Equation 1 2 3 4 5 6 7 Aid/govt. expenditure Polity index Ordered logit Median regression Post- 1990 2SLS
-Intercept 1.506 2.702 2.949 F 3.115 4.910 −0.793
-(4.432) (4.457) (9.313) (4.916) (5.023) (4.860)
-Aid −0.023 0.016 0.011 −0.007 0.023 0.028 0.060
-(0.044) (0.019) (0.091) (0.032) (0.043) (0.032) (0.080)
-Initial democracy index value −0.833nn −0.909nn −0.806nn −0.544nn −0.875nn −0.318n −0.845nn
-(0.110) (0.105) (0.095) (0.096) (0.114) (0.126) (0.116)
-Illiteracy change/ initial illiteracy 0.093n 0.102n 0.073 0.070n 0.149nn 0.155 0.077
-(0.041) (0.042) (0.080) (0.028) (0.047) (0.097) (0.045)
-Initial illiteracy 0.006 −0.009 −0.010 0.001 −0.009 −0.004 −0.002
-(0.017) (0.020) (0.037) (0.013) (0.021) (0.017) (0.020)
-Annual growth of per capita income 0.185 0.282n 0.118 0.134 0.270n 0.099 0.202
-(0.119) (0.137) (0.188) (0.101) (0.110) (0.091) (0.127)
-Initial (log) per capita income 0.755 0.733 0.102 0.468 0.695 −0.219 1.023
-(0.536) (0.529) (1.103) (0.376) (0.601) (0.606) (0.590)
-Latin America & Caribbean 2.688nn 2.489nn 4.901nn 1.880nn 2.661nn 1.039 2.524nn
-(0.716) (0.751) (1.328) (0.565) (0.837) (0.603) (0.726)
-Africa 1.664n 1.303 0.013 1.239n 1.903n 1.451n 1.129
-(0.808) (0.822) (1.742) (0.625) (0.846) (0.650) (0.930)
-Eastern Europe 2.924nn 2.406n 3.741 1.904nn 2.496n 2.773nn 2.822nn
-(0.946) (1.049) (2.276) (0.731) (1.205) (0.733) (0.978)
-N 105 96 97 105 105 105 105
-R2/pseudo R2 .50 .57 .59 .12 .29 .29 .49
-Dependent variable is the change in the democracy index (Freedom House, 1975–2000 except 1990–2000 in equation 6, and Polity, 1975–2000 in equation 3). Standard errors are in parentheses. A n (nn) indicates significance at .05 (.01) level for 2-tailed tests.
-An alternative, informal approach to exploring the direction of causality is to compare the strength of democratization’s relationships with aid early in the period and late in the period. If aid averaged over 1975–1989, for example, turned out to be strongly related to democratization (measured over the full period 1975–2000), that would suggest aid encourages democratization. On the other hand, if aid disbursed later in the period, for example 1990–1999, were more strongly related to democratization (again, measured over the full period), that would suggest aid is allocated as a reward for prior or ongoing democratization.
-In fact, aid over each of these sub-periods has equally small and insignificant relationships with democratization, when they are substituted for the full-period aid variable in regressions otherwise identical to those in Table 2.
-Table 3 reports OLS regressions for specifications that differ somewhat from those reported in Table 2. If aid promotes democracy indirectly through increasing per capita incomes or through improving access to education, those variables could capture part of any positive impact aid might have on democracy. Therefore, equation 1 of Table 2 omits per capita income growth and changes in illiteracy from the regressions.
-Results in equation 2, however, show no evidence that aid promotes democracy through increasing incomes or improving education, as aid remains insignificant. Equation 2 shows that aid remains insignificant in a specification that includes a more exhaustive set of regressors. This test adds an alternative measure of modernization: urbanization (the 1975 level, and the change from 1975 to 1999).
Religious composition has also been linked to democratization, so percent Muslim, percent Protestant, and percent Catholic were added. Adding all of these regressors leaves the aid coefficient insignificant. Of the added regressors, only percent Muslim is significant (at the .06 level). An increase of nearly 40 percentage points in percent Muslim is associated with a decline of about 1 point on the Freedom House index.
-Results from three additional tests are not reported in the tables for space reasons. First, aid’s impact on democratization might depend on the initial level of democracy. Democracy assistance from USAID tends to focus more on countries already experiencing some democratic openings, rather than on hard-core autocracies, reflecting beliefs about where such aid is likely to be most efficacious. Accordingly, an interaction term equal to the product of aid/ GNP and initial democracy level was added to the regressions. This interaction provided to be insignificant, indicating no general tendency for aid to be more efficacious in recipients with either low or high initial levels of democracy. Second, the insignificance of aid potentially could be attributed to multicollinearity, as aid/GNP is correlated with illiteracy at .56 and with initial per capita income at – .56. However, when either or both of those variables are deleted from the regression, aid remains unrelated to democratization. Third, aid remains unrelated to democratization when the dependent variable used is the end-of-period (2000) level of democracy, rather than the change over time from 1975 to 2000.
-Table 3. Alternative Specifications
Equation 1 2
Intercept 3.359 17.422
(4.498) (8.777)
Aid/GNP −0.004 −0.022
(0.038) (0.039)
Initial democracy index value −0.757nn −0.937nn
(0.110) (0.120)
Illiteracy change/initial illiteracy 0.179nn
(0.047)
Initial illiteracy −0.019 0.016
(0.016) (0.024)
Annual growth of per capita income 0.101
(0.217)
Initial (log) per capita income 0.394 0.494
(0.547) (0.689)
Latin America & Caribbean 2.722nn 0.064
(0.740) (1.038)
Africa 1.102 0.689
(0.822) (0.889)
Eastern Europe 3.551nn 0.186
(0.983) (1.084)
Urban share of pop. change/Initial urban pop. share 0.013
(0.036)
Initial urban pop. share 0.020
(0.025)
% Protestant 0.023
(0.023)
% Catholic –0.004
(0.012)
% Moslem –0.026
(0.014)
R2 0.46 0.65
Dependent variable is the change in the Freedom House index, 1975–2000. Standard errors are in parentheses. A n (nn) indicates significance at .05 (.01) level for 2-tailed test. Sample size is 105.
Conclusions
Both the Freedom House and Polity indexes indicate clear trends toward democratization over the 1975–2000 period.
The number of aid-recipient nations with improved democracy ratings far exceeds the number with reduced ratings.
Results from this study, however, indicate that little if any of this progress can be attributed to foreign aid.
This weak relationship between aid and democratization cannot be attributed merely to a lack of variation over time in the democracy indexes.
The median shift on the 2–14 point Freedom House scale over the 1975–2000 period was 2 points, and for the Polity index was 5 points.
Moreover, significant and strong relationships are found between changes in these indexes and several other variables, most notably the initial level of democracy.
Donor agencies find the Freedom House indexes sufficiently useful that many of them actually use these indexes in tracking progress toward democratization in recipient countries.
Findings from this analysis should be interpreted with caution, however. For example, results do not necessarily suggest that funding for democracy promotion should be curtailed.
The lack of any relationship between aid and improvements in democratic governance does not imply that none of the democracy-promoting projects sponsored by donors has any effects.
The available data on aid do not permit disaggregation of aid intended to promote democracy from aid intended to achieve other objectives.
The evidence presented here does suggest that either the favorable impacts of aid on democratization are minor, or they are roughly balanced by other democracy-undermining effects of aid dependence.
Cross- country, highly aggregated studies such as this one must be complemented by case study evidence that more closely examines the effectiveness of particular democracy-promoting programs.
Even were such studies to indicate that current approaches are mostly ineffective, that would not imply that no conceivable democracy-promoting efforts would be efficacious. Moreover, the developed democracies may be able to promote democracy through non-aid channels, including trade, foreign direct investment, and diplomacy.
Appendix
The Freedom House Democracy Indicators
Since the early 1970s, the Comparative Survey of Freedom, published by Freedom House, has provided cross-country indexes of "civil liberties" and "political freedoms."
Countries are assigned scores from 1 to 7, with smaller values assigned to countries with greater liberties. Values are based on judgments for each country on the following set of criteria.
Civil liberties
media/literature free of political censorship
open public discussion
freedom of assembly & demonstration
freedom of political organization
nondiscriminatory rule of law/independent judiciary
freedom from unjustified terror or imprisonment
free trade unions, peasant organizations
free business or cooperatives
free professional or other private organizations
free religious institutions
personal social rights: property, travel, residence, family
socioeconomic rights; freedom from dependency on landlords, etc.
freedom from gross socioeconomic inequality
freedom from gross government indifference or corruption
Political freedoms
chief authority recently elected by a meaningful process
legislature recently elected by a meaningful process
fair election laws, campaigning opportunity, polling and tabulation
fair reflection of voter preference in distribution of power
multiple political parties
recent shifts in power through elections
significant opposition vote
free of military or foreign control
major group or groups denied reasonable self-determination
decentralized political power
informal consensus; de facto opposition power
Data Appendix
Recipient Countries Freedom House (change) Freedom House (1975) Aid/GNP
Benin 10 2 10.4
Panama 10 3 1.1
Bulgaria 9 2 0.7
Hungary 9 4 0.2
Mali 9 2 17.9
Mongolia 9 2 6.6
Poland 9 4 0.7
Romania 9 3 0.3
Chile 8 4 0.2
Uruguay 8 6 0.3
Bolivia 7 5 8.0
Central Afr. Rep. 7 2 13.6
Cape Verde 7 6 16.7
Ghana 7 4 7.2
Malawi 7 3 19.7
Cyprus 6 8 1.7
Korea 6 6 0.2
South Africa 6 7 0.1
Albania 5 2 6.5
Bangladesh 5 4 5.6
Ecuador 5 5 1.2
Mozambique 5 4 24.2
Niger 5 3 14.1
Philippines 5 6 1.6
Jordan 4 4 15.0
Madagascar 4 6 9.6
Nepal 4 5 8.7
Peru 4 6 1.2
Argentina 3 10 0.1
Recipient Countries Freedom House (change) Freedom House (1975) Aid/GNP
Brazil 3 7 0.1
Ethiopia 3 3 9.0
Guinea Bissau 3 4 47.3
Guyana 3 9 16.5
Honduras 3 7 7.7
Indonesia 3 6 1.3
Moldova 3 7 2.3
Nigeria 3 5 0.5
Nicaragua 3 7 20.7
Paraguay 3 6 1.5
Senegal 3 6 12.3
Togo 3 3 11.6
Armenia 2 6 2.8
Burkina Faso 2 6 14.7
Dominican Rep. 2 10 1.4
Algeria 2 3 0.5
Croatia 2 9 0.3
India 2 9 0.8
Cambodia 2 2 4.8
Lithuania 2 11 1.2
Latvia 2 11 1.1
Mexico 2 9 0.1
Mauritius 2 11 2.6
Namibia 2 9 4.8
Slovenia 2 11 0.2
Chad 2 3 12.9
Burundi 1 3 16.2
Bahamas 1 13 0.1
Belize 1 13 6.4
Botswana 1 11 6.5
China 1 2 0.3
Congo 1 5 8.7
Haiti 1 4 9.8
Israel 1 11 3.7
Lesotho 1 7 11.4
Morocco 1 6 3.2
Mauritania 1 4 26.3
Zambia 1 6 17.2
Recipient Countries Freedom House (change) Freedom House (1975) Aid/GNP
Cote d’Ivoire 0 5 5.2
Eritrea 0 4 14.9
Guatemala 0 9 1.6
Iran 0 4 0.1
Malta 0 14 2.0
Papua New Guinea 0 11 11.7
El Salvador 0 11 5.2
Thailand 0 11 0.9
Trinidad 0 12 0.2
Tunisia 0 5 2.6
Yemen 0 5 6.1
Zimbabwe 0 5 3.9
Costa Rica −1 14 2.7
Recipient Countries Freedom House (change) Freedom House (19
This study investigates if foreign aid fosters democracy in recipient countries between 1975 and 2000. Despite the potential for aid to support democratization, the analysis,