Foreign aid and the potential effects on income inequality in recipient nations.

Van Rensselaer, Kristen N. ; Gordon, Bruce L. ; Barrett, J. Douglas 等

ABSTRACT

Foreign aid can be effective in promoting economic growth and reducing poverty but there are also cases where aid has been wasted by the recipient nation. This study examines the effectiveness of foreign aid on income inequality, as measured by the GINI coefficient. Several studies have documented that foreign aid has a positive effect on economic growth, but few studies have focused on the impact that aid has on income inequality. While economic growth should benefit the economy as a whole, economic growth does not imply that income inequality will improve. Using data for twenty-nine countries and controlling for a variety of factors, the results of the study suggest that increased aid may actually be associated with higher levels of income inequality in the recipient nations. However, the results are sensitive to model specification.

INTRODUCTION

"Foreign aid has at times been a spectacular success--and an unmitigated failure." "Assessing Aid: What Works, What Doesn't, and Why," World Bank Policy Paper (1998)

Foreign aid is a heavily researched topic. It is controversial from the standpoint that, when used correctly, aid has a tremendous impact on a developing economy. But, when used incorrectly, it is squandered, or worse, lines the pockets of a few individuals. Foreign aid levels reached their peak during the 1970s and 1980s and started to decline in the 1990s. Up until the 1990s, research indicates that foreign aid was given equally to countries with good and poor economic management. However during the 1990s, this changed due to the finding that foreign aid is more effective in countries with stable macroeconomic policies.

This study enhances the body of literature on aid effectiveness by investigating the influence that foreign aid has on changes in income inequality for a sample of 29 countries. After controlling for various factors, the results of the study suggest that foreign aid negatively impacts changes in income inequality. The results, however, are sensitive to model specification.

LITERATURE REVIEW

Although very little research could be found on the direct relationship between foreign aid and income inequality, there is a plethora of research on the effectiveness, if any, of aid on recipient countries. There is also a significant body of research on the determinants of aid policy in donor nations and the main focus from this area will be how these policies appear to have been changing in recent years.

Evidence that aid has begun to be allocated differently is provided by Dollar and Levin (2004), where they examined the allocation of foreign aid by 41 donor agencies. They found that there is a clear tendency among donor agencies to allocate more assistance to poor counties that have good economic governance. This contrasted with previous policies where aid was allocated on the basis of perceived need or other, possibly politically motivated, reasons, regardless of how corrupt or undeveloped the country's governance policies may have been. There is still an abundant amount of aid allocated based more on the purported "needs" of the donor country than the poverty levels of the recipient country. However, when aid is allocated based on poverty, the level of development of the country's institutions and policies is becoming increasingly important.

Ireland, McGregor, and Saltmarshe (2003) discuss some of the major changes occurring at donor agencies. Donor agencies are increasingly concerned with the effectiveness of their assistance programs and many have instituted "results-based" approaches to assistance. Ireland, McGregor, and Saltmarshe (2003) examine the policies of donor agencies and review the relevant literature and identify four common areas of concern that have emerged as agencies institute performance assessment; ownership, decentralization and leadership, accountability, and learning and complexity. They conclude that performance assessment takes place in a very complex environment, an environment often made more complex by the agencies themselves. Effective performance assessment is very difficult given the inherent complexity created by the large number of stakeholders and the additional agency-created complexity.

Alesina and Dollar (2000) found that foreign aid is "dictated as much by political and strategic considerations, as by the economic needs and policy performance of the recipients. Colonial past and political alliances are major determinants of foreign aid." However, they did find that countries that were more democratized received more aid, everything else being equal. They also looked at the flow of direct investment and found that "good policies" and the protection of private property rights were crucial to foreign investors.

Alesina and Weder (2002) found evidence that less corrupt governments do not receive more foreign aid than corrupt governments. This finding held for bilateral, multilateral, and individual donors, although evidence indicated that some individual donors, primarily Scandinavian countries, were more selective and gave less to more corrupt countries. In fact, their study showed that increased aid leads to increased corruption.

White (2004) reviewed trends in official flows to developing countries over three decades. He found that real aid levels declined in the 1990's and that the aid that was given has moved increasingly from the poorer countries, especially sub-Saharan Africa, and is driven more by political reasons. European donor countries appear to be donating more to European recipients, and aid in general is flowing more to areas where individual donors have a perceived interest, whereas multilateral donors still focus more on countries that appear to be more in need of assistance. Although the evidence is mixed, there does appear to be a growing recognition that need-based aid might be more effective in areas that are less corrupt and have more developed institutions. Many donors, especially multilateral donors like the World Bank are beginning to view corrupt, totally undeveloped countries as "rat holes" where money will simply disappear.

Additional evidence of this new "ideal" in the granting of aid is provided by the Key Principles of the Millennium Challenge Corporation (MCC). In January 2004, Congress passed a "new compact for global development" called for by President Bush which links greater contributions from developed nations to greater responsibility from developing nations. The President proposed a concrete mechanism to implement this compact--the Millennium Challenge Account (MCA)--in which development assistance would be provided to those countries that rule justly, invest in their people, and encourage economic freedom.

Key Principles

Reduce Poverty through Economic Growth: The MCC will focus specifically on promoting sustainable economic growth that reduces poverty through investments in areas such as agriculture, education, private sector development, and capacity building.

Reward Good Policy: Using objective indicators, countries will be selected to receive assistance based on their performance in governing justly, investing their citizens, and encouraging economic freedom.

Operate in Partnership: Working closely with the MCC, countries that receive MCA assistance will be responsible for identifying the greatest barriers to their own development, ensuring civil society participation, and developing an MCA program. MCA participation will require a high-level commitment from the host government. Each MCA country will enter into a public Compact with the MCC that includes a multi-year plan for achieving shared development objectives and identifies the responsibilities of each partner in achieving those objectives.

Focus on Results: MCA assistance will go to those countries that have developed well-designed programs with clear objectives, benchmarks to measure progress, procedures to ensure fiscal accountability for the use of MCA assistance, and a plan for effective monitoring and objective evaluation of results. Programs will be designed to enable progress to be sustained after the funding under the MCA Compact has ended (http://ww.mca.gov, 2005).

This new awareness of aid effectiveness and the new guidelines for need-based foreign aid have arisen primarily because of research that has shown that aid is more effective in countries with more developed institutions and policies. In a review of the effects of aid on recipient nations, Burnside and Dollar (1997, 2004) concluded that there is broad agreement that aid does not have the same effect on all recipients. They specifically investigated the role that the quality of state institutions and policies has on the effectiveness of aid. They present evidence that aid spurs growth much more strongly in countries where the quality and integrity of institutions and policies is more developed. They also concluded that in the past, aid has not systematically led to improvements in institutions and policies. They argue that this might change in the future if aid is allocated more to countries with more developed institutions and policies.

The work of Burnside and Dollar (1997) has not been well-received by some. Easterly (2003), Hansen and Tarp (2000), and Dalgaard and Hansen (2000) among others, have all been critical. Easterly (2003) criticized the methodology used by Burnside and Dollar (2000) and attempted to show that their results were somewhat contrived and did not hold up consistently. Dalgaard and Hansen (2000) find that while good policies spur growth they may at the same time lead to decreasing effectiveness of foreign aid. They also contend that the econometric results obtained in Burnside and Dollar (1997) are fragile and extremely dependent upon the data set used. Burnside and Dollar (2004) used the same methodology as their earlier work, but with a larger and improved data set, to refute the criticisms and illustrate the robustness of their results.

Durbarry, Gemmell, and Greenaway (1998) investigated the effects of foreign aid using a variety of econometric techniques and found that foreign aid inflows have a beneficial effect on the growth rates of economies in lesser developed countries, but only when the countries had stable macroeconomic policies. Their results also suggested that there was an optimal level of aid and that too little or too much aid could actually have little positive effect.

Research by Banerjee and Rondinelli (2003) was designed to discover whether or not increased foreign aid led to an increase in the privatization of state-owned enterprises in developing countries. The process of privatization is being increasingly viewed as a key factor in the economic growth in less-developed countries. Their research revealed no direct relationship between aid and privatization and concluded that it was fundamentally a political decision based on the assessment of local government leaders concerning the political benefits of such a change in policy. However, their research did find "robust evidence of foreign aid playing a constructive role in privatization in the presence of superior governance structures."

A very interesting result was obtained by Kosack (2003) in research aimed at determining if aid given to democratic countries was more effective in improving the quality of life of its citizens than aid given to autocracies. Kosack concludes that while aid in general has been shown to be ineffective, aid given to democratic countries does ameliorate the quality of life of its citizens. He attributes this to a general tendency of democracies to treat their people well, or at least better than autocracies. His results showed that economic aid can actually be harmful in autocracies in that the well being of citizens actually is worse off than if aid had not been given at all. A possible cause of this problem in autocracies is that aid money is spent to increase the advantage that the ruling parties have over the citizens of the country, an advantage which they feel is crucial if they are to remain in power.

Dollar and Kraay (2002) find that economic growth does not positively affect the poorest one-fifth of a country in a disproportionate manner. The poor benefit at the same rate as the rest of the country when economic growth occurs. The income of the poorest fifth of society increases with the average income of the nation. They also found that "a variety of pro-growth macroeconomic policies, such as low inflation, moderate size of government, sound financial development, respect for the rule of law, and openness to international trade, raise average incomes with little systematic effect on the distribution of income." They do not rule out the possibility that their methodology may not be able to capture the small marginal changes that some policies might have on the income of the poorest quintile, but in general conclude that at a minimum, the poorest citizens of a nation benefit just as much as everyone else when economic growth occurs. Thus, if economic aid can be structured to effectively help economic growth, then the poor should benefit in at least a proportional manner with the rest of the country.

DATA AND METHODOLGY

Gini coefficients are commonly used as a measure of income inequality. A Gini index value of zero indicates perfect income equality and, at the opposite extreme, a Gini index value of 100 indicates perfect income inequality. Since it is the goal of this research to explore the impact that financial aid has on the recipient nation's income inequality, we use as our dependent variable the change in the Gini index over a period of time. Ideally, the data used for this study would have consistent time frames for all countries included. However, Gini data is far from ideal. Deininger and Squire (1996) describe the problems with income inequality data and developed a "quality" data set that many researchers have utilized.

The income inequality data used in this study were taken from the database of Dollar and Kraay (2002), which draws upon four sources for the income inequality data including Deninger and Squire (1996), and the World Development Indicators (WDI) CD-ROM (2003). The countries selected had to have Gini indices spaced at least seven years apart. This would allow for a sufficient period of time to measure a change in income inequality. Also, the country had to be a recipient of Official Development Assistance (ODA). This was determined by the Aid Dependency Table (6.10) of the 2003 WDI Annual Report. The number of countries in the Dollar and Kraay (2002) data set with at least two Ginis with a seven year span and was a recipient of ODA was 66. These 66 countries were then analyzed based on their aid dependency. One measure that the World Bank uses as Aid Dependency is ODA as a percent of Gross National Income (GNI). This aid dependency ratio was then collected from the WDI CD-Rom over the 1980-2000 time period. The average over this time period was calculated and then ranked in descending order. The countries that had an average aid dependency ratio lower than 0.50% were omitted from the sample. The 0.50% cutoff was determined by Aid Dependency Table (6.10) of the 2003 World Development Indicators. In 1996 the average aid dependency ratio of low income countries was 2.5% and for middle income countries was 0.50%. Since the sample of countries used for this study included both low and middle income countries, the 0.50% was used as the cutoff. This left a sample of 44 countries. With this group of 44 countries, the next step was to start gathering other data that is used to control for certain aspects of the recipient nations' economy (population, GDP per capita, etc.) For these 44 countries, 15 countries did not have data going back to the 1980s. This led to the final sample of 29 countries.

Table 1 summarizes the data found for the 29 countries used in the study. It should be noted that the change in the Gini is actually the beginning Gini minus the ending Gini since this would lead to an intuitive interpretation. A lower Gini value indicates greater equality in income. For a country to show improvement in income inequality, the ending value of the Gini should have a lower value than the beginning Gini.

While foreign aid can take many forms, the aid measure used is Official Development Assistance. Official Development Assistance (ODA) is defined by the World Bank (WDI CD-ROM, 2003) as:

 Official development assistance and net official aid record the
 actual international transfer by the donor of financial resources
 or of goods or services valued at the cost to the donor, less any
 repayments of loan principal during the same period. Grants by
 official agencies of the members of the Development Assistance
 Committee are included, as are loans with a grant element of at
 least 25 percent, and technical cooperation and assistance.

Other measures of foreign aid have been used in the literature. Dalgaard and Hansen (2000) assessed the differences in alternative measures of foreign aid. They found that nominal ODA as a percent of GDP was highly correlated to alternatives such as real ODA (adjusted for purchasing power parity) and Effective Development Assistance of Chang, Fernandez-Arias, and Serven (1998). Dalgaard and Hansen (2000) found that, statistically speaking, there is little difference in alternative measures. Since we want to capture the effects of total aid dollars on income equality, ODA in current U.S. dollars is used (nominal ODA). The annual dollar amount of ODA is averaged from 1980-1989. The reason for measuring aid prior to the first Gini index is to allow for a time lag since aid will not have an immediate impact on income inequality. The time frame for the aid starts at a minimum of four years prior to the first Gini index.

The remaining variables used to explain the change in Gini and to control for various country specific conditions are provided in the appendix. This collection of potential explanatory variables stems from the past literature for income inequality and growth. For most variables with annual observations, an average is taken from 1985-1994. This ten-year window tends to correspond to the time period for which the change in the Gini is being measured. Table 2 shows the correlation matrix for all variables used in the study. There is very low correlation between the potential explanatory variables and the change in the Gini index. There are no correlations between any of the variables that are greater that 0.70. There are three correlations in the 0.60 to 0.70 range: population and ODA has a correlation of 0.69, fertility and enrollment in secondary school has a correlation of -0.63; and government consumption and trade has a correlation of 0.65.

A variety of statistical approaches have been used in the literature for income inequality. Most of the focus on foreign aid has centered around its potential effects on economic growth. Within the growth literature, there are various growth models and production functions which specify the inputs, and researchers have explored a wide variety of econometric models. For income equality, there is no exact theoretical model to dictate what variables or methodologies are correct. Without a theoretical model specifying the impact that foreign aid should have on income equality, the statistical approach is left to the researcher. Since the thrust of this research is to search for a causal relationship between foreign aid and changes in income inequality, ordinary least squares is used.

RESULTS

When hypothesizing the potential impact that foreign aid can have on the recipient nation's income inequality, there are three possible outcomes. First, foreign aid can lead to an improvement of income inequality. This would be indicated by a statistically significant positive coefficient for the aid variable. This is potentially the most desirable outcome from the standpoint that this would allow the conclusion that foreign aid benefits or improves the income inequality in recipient nations. This would also corroborate the Modernization Theory (Hoselitz, 1960), where assistance to developing countries expedites the development process.

Second, foreign aid has no impact on income inequality. This would be evidenced by a coefficient that was not statistically different from zero. While statistically this would mean no relationship between income inequality and foreign aid, this would not imply that foreign aid does not improve economic conditions or encourage growth of developing economies. Rather, that the benefits of aid are not directly related to improvement in income inequality.

The third possible outcome is that foreign aid has a negative impact on income inequality. If there is a statistically significant negative relationship between changes in income inequality and foreign aid, this would imply that recipients of aid were not made better off by receiving aid from an income inequality standpoint. We would not necessarily be able to say that the countries are in worse condition for having received aid but there would be evidence to indicate that foreign aid did not improve income inequality. Additionally, this outcome would lend support to the Dependency Theory (Amin, 1976; Frank, 1979) where the recipient nations become dependent on developed countries for trade and capital to the point where economic growth is impeded. If the Dependency Theory is correct, and foreign aid hurts economic growth it could also hamper efforts to improve income equality.

Table 3 reports the results of the regression analysis. This model encompasses all the potential explanatory variables. The variable that is statistically significant at the five percent level is arable land (average arable land per capita in hectares). ODA and government consumption are statistically significant at the ten percent level of significance. Arable land and ODA both have negative impacts on changes in income inequality and government consumption has a positive relationship. The adjusted R-squared (0.04) and F-statistic (1.09) for the regression model is very low since there are numerous variables included in the model that are not statistically significant. As far as the diagnostics for the regression results, the residuals are normally distributed, there is no hetroskedasticity, but there is evidence of serial correlation.

In order to find a more parsimonious model, the redundant variable test was used to determine if variables could be omitted from the regression. Using this approach, freedom index, foreign investment, per capita GDP, GDP growth, illiteracy, inflation, population, and enrollment in secondary school could all be omitted. As these variables were omitted, the adjusted R-squared and the F-statistic improved. The best fit model is presented in Table 4. The adjusted R-squared is 0.20 and the F-Statistic is 2.39. The residuals are normally distributed with no hetroskedasticity or serial correlation. In this model, arable land and government consumption are significant at the 5% level and ODA is significant at the 10% level. While fertility and trade are not significant at the 10% level, their role in the model is still important since without these two variables, the model's explanatory power and validity is reduced. The relationship among the variables is perhaps more complex than is being captured through the model.

Given the results in Table 4, there are some interesting findings. Arable land is statistically significant and has a negative coefficient. The more arable land a country has the more likely that income inequality will worsen over time in the presence of other variables being constant. This provides evidence for the theory that natural resources can be a curse. As explained in Sachs and Warner (2001), countries with abundant natural resources tend to grow more slowly than resource-poor countries. They conclude that resource-rich countries miss opportunities for export-led growth. Government consumption has a positive relationship with improvement of income inequality. This indicates that government spending does improve income inequality. Foreign aid has a negative impact on income inequality. As discussed earlier, a negative relationship indicates that, from an income inequality view, recipient nations are worse off from receiving aid. Again, it does not imply aid is not beneficial in other aspects, but given our results it lends some credence to the Dependency Theory. It is also interesting to note the variables that are not statistically significant in explaining changes in income inequality. For these 29 countries, measures for the countries' political policies (freedom index and corruption index), economic conditions (foreign investment, GDP per capita, GDP growth, and inflation), and human capital (illiteracy and enrollment in secondary school) were not able to explain changes in income inequality.

The conclusions of the study must be interpreted with caution since the statistical results appear to be somewhat fragile. As different variables were omitted from the regression, the significance but not the sign of ODA changed. To explore this issue further, two alternative specifications of the aid variable (ODA) were substituted in the above analysis. First, ODA as a ratio of each country's GDP was substituted in the model. This is the most common measure of aid utilized in the literature. The results did not show a significant statistical relationship (p-value = 0.1642) between aid and changes in income inequality. Second, instead of using a consistent time frame for measuring ODA across the 29 countries, we matched up the time frames with the Gini index. Under this specification, ODA in current U.S. dollars was still used but the average for ODA was calculated using ODA five years prior to the first Gini through five years prior to the second Gini. For this specification, the results did not show any statistical relationship between aid and changes in income inequality at the ten percent level (p-value = 0.1037). In two specifications of aid, we do not find a statistically negative relationship at the 10 percent level of significance but for our original specification, a negative relationship exits.

CONCLUSIONS

The results of the statistical analysis indicate that arable land, government consumption and foreign aid are all important variables in explaining changes in income inequality. Both arable land and foreign aid have a negative relationship with changes in income inequality, while government consumption has a positive relationship. Our results provide evidence of foreign aid ineffectiveness for recipient nations and that foreign aid negatively impacts changes in income equality over time. Also of interest are the variables that were not found to be significant predictors, such as political rights (freedom index), perceptions of the county's level of corruption (corruption index), economic factors (inflation and economic growth), and human capital (illiteracy and secondary school enrollment).

An area for future research would involve further investigation of the institutions and policies for the recipient nations. Burnside and Dollar (2004) found that foreign aid is beneficial for growth in recipient nations that have sound institutions and policies. While their results have been debated, the importance of quality institutions and policies should be explored further. In our attempts to measure this, we used corruption index, freedom index, and foreign investment could also be used as an indirect indicator of a country's management. Our measures of quality institutions and policies did not prove to be significant in explaining changes in income inequality. It may be that the indicators used are not adequate or that the relationship is too complex to capture in simple regression models. Also, more work needs to be conducted on the alternative measures of foreign aid since our results suggest that aid definitions may impact the overall outcome of the role that foreign aid has on income inequality in recipient nations.

APPENDIX: VARIABLE DEFINITIONS AND SOURCES

Variable Comments Source

Gini Index 0 indicates perfect Dollar and Kraay (2002)
 equality and 100 indicates and World Development
 perfect inequality Indicators CD-ROM 2003

Change Change in the Gini Index Calculated

Corruption Corruption Perception Index www.icgg.org
 (0=most corrupt and 10=
 corruption free) for 1996

Fertility Average Fertility Rates World Development
 (total births per woman) Indicators CD-ROM
 for 1985, 1987, 1990, (2003)
 and 1992

Foreign Average net inflows of World Development
Investment foreign direct investment Indicators CD-ROM
 as a percent of GDP from (2003)
 1985-1994

Freedom Average of Political Freedom House
 Rights Freedom Index (1=
 highest degree of freedom
 and 7=lowest) from 1985-1994

GDP Average GDP per capita in World Development
 constant 1995 dollars Indicators CD-ROM
 from 1985-1994 (2003)

GDP Growth Average growth rate in World Development
 GDP from 1985-1994 Indicators CD-ROM
 (2003)

Government Average government World Development
Consumption consumption as a percent Indicators CD-ROM
 of GDP from 1985-1994 (2003)

Illiteracy Average adult illiteracy World Development
 rate from 1985-1994 Indicators CD-ROM
 (2003)
Inflation Natural log of 1+average World Development
 inflation rate (CPI) from Indicators CD-ROM
 1985-1994 for some (2003)
 countries, the average is
 for the inflation that is
 available over this time
 period

Land Average arable land per World Development
 capita (hectares) from Indicators CD-ROM
 1985-1994 (2003)

ODA Average Official Development World Development
 Assistance in U.S. dollars Indicators CD-ROM
 from 1980-1989 (2003)

Population Average total population in World Development
 thousands from 1985-1994 Indicators CD-ROM
 (2003)

Secondary Ratio of total enrollment in World Development
School secondary school, regardless Indicators CD-ROM
 of age, to the population of (2003)
 the age group that officially
 corresponds to the level of
 education shown. The average
 is taken over 1990-1994.

Trade Sum of exports and imports of World Development
 goods and services measured Indicators CD-ROM
 as a percent of gross (2003)
 domestic product. Average
 over 1985-1994

REFERENCES

Anonymous (2004, May/June). Ranking the rich 2004, Foreign Policy, 46-56.

Alesina, A. and B. Weder (2002). Do corrupt governments receive less foreign aid? American Economic Review, 92(4), 1126-1137.

Alesina, A. and D. Dollar (2000). Who gives foreign aid to whom and why, Journal of Economic Growth, 5, 33-63.

Amin, S. (1976). Unequal development. New York: Monthly Review Press.

Banerjee, S. G. and D. A. Rondinelli (2003). Does foreign aid promote privatization? Empirical evidence from developing countries, World Development, 31(9), 1527-1548.

Boone, P. (1995). Politics and the effectiveness of foreign aid. Centre for Economic Performance Discussion Paper 72.

Burnside, C. and D. Dollar (1997). Aid, policies and growth. World Bank Policy Research Working Paper 1777.

Burnside, C. and D. Dollar (2000). Aid, policies, and growth. American Economic Review, 90(4), 847-868.

Burnside, C. and D. Dollar (2004). Aid, policies, and growth: Revisiting the evidence. World Bank Policy Research Working Paper 3251.

Chang, C., E. Fernandez-Arias, and L. Serven (1998). Measuring aid flows: A new approach, World Bank Policy Research Working Paper 2050.

Dalgaard, C. and H. Hansen (2000). On aid, growth, and good policies. Centre for Research in Economic Development and International Trade Working Paper, 00/17.

Deininger, K. and L. Squire (1996). A new data set measuring income inequality, World Bank Economic Review, 10(3), 565-591.

Dollar, D. and A. Kraay (2002). Growth is good for the poor. Journal of Economic Growth, 7(3), 195-225.

Dollar, D. and V. Levin (2004). The increasing selectivity of foreign aid, 1984-2002. World Bank Policy Research Working Paper 3299.

Durbarry, R., N. Gennell, and D. Greenaway (1998). New evidence on the impact of foreign aid on economic growth. Centre for Research in Economic Development and International Trade, Working Paper 98/8.

Easterly, W. (2003). Can foreign aid buy growth? Journal of Economic Perspectives, 17(3), 23-48.

Frank, A. G. (1979), Dependent accumulation and underdevelopment. New York: Monthly Review Press.

Hansen, H. and F. Tarp (2000). Aid and growth regressions. Centre for Research in Economic Development and International Trade, Working Paper 00/7.

Hansen, H. and F. Tarp (2000). Aid effectiveness disputed. Journal of International Development, 12(3), 375-398.

Hoselitz, B. (1960). Sociological aspects of growth. New York: Free Press.

Ireland, M., J. A. McGregor and D. Saltmarshe (2003). Challenges for donor agency country-level performance assessment: A review. Public Administration & Development, 23(5), 419-431.

Kosack, S. (2002). Effective aid: How democracy allows development aid to improve the quality of life. World Development, 31(1), 1-22.

Mosley, P., J. Hudson, and S. Horrell (1987). Aid, the public sector and the market in less developed countries. The Economic Journal, 97(387), 616-641.

Sachs, J. D. and A. M. Warner (2001), The curse of natural resources, European Economic Review, 45(4), 827-838.

White, H. (2004). Trends in the volume and allocation of official flows from donor countries. International Review of Economics and Finance, 13, 233-244.

World Bank (2003). World Development Indicators. CD-ROM.

World Bank (1998). Assessing aid: What works, what doesn't, and why. Oxford: Oxford University Press.

Kristen N. Van Rensselaer, University of North Alabama Bruce L. Gordon, University of North Alabama J. Douglas Barrett, University of North Alabama

Table 1
Summary of Income Inequality Data

 Dates Beginning Gini Ending Gini

Algeria 88/95 38.73 35.3
Bangladesh 85/95 26.92 33.63
Bolivia 90/99 42.04 44.70
Botswana 85/93 54.21 63.00
Cote d'Ivoire 85/95 41.21 38.00
Dom. Rep. 89/96 50.46 48.71
Ecuador 88/95 43.91 43.73
Egypt 91/99 32.00 34.4
El Salvador 89/98 48.96 50.8
Ghana 87/97 35.35 32.70
Honduras 86/96 54.94 53.72
India 86/96 32.22 32.86
Indonesia 87/96 32.01 36.48
Jamaica 88/96 43.16 36.43
Jordan 86/97 36.06 36.42
Morocco 90/99 39.20 39.50
Mauritania 88/95 42.53 38.90
Nepal 84/95 30.06 38.78
Nigeria 85/93 38.68 37.47
Pakistan 85/96 32.44 31.20
Panama 89/97 56.47 48.53
Peru 85/94 45.72 44.87
Philippines 85/94 46.08 45.07
Paraguay 91/98 39.74 57.70
Sri Lanka 85/95 32.47 34.36
Thailand 86/96 47.40 40.45
Tunisia 85/95 43.00 41.7
Uganda 89/96 33.00 37.4
Zambia 91/98 43.51 52.6

 Change Percent Change

Algeria 3.43 8.86%
Bangladesh -6.71 -24.93%
Bolivia -2.66 -6.33%
Botswana -8.79 -16.21%
Cote d'Ivoire 3.21 7.79%
Dom. Rep. 1.75 3.47%
Ecuador 0.18 0.41%
Egypt -2.40 -7.50%
El Salvador -1.84 -3.76%
Ghana 2.65 7.50%
Honduras 1.22 2.22%
India -0.64 -1.99%
Indonesia -4.47 -13.96%
Jamaica 6.73 15.59%
Jordan -0.36 -1.00%
Morocco -0.30 -0.77%
Mauritania 3.63 8.54%
Nepal -8.72 -29.01%
Nigeria 1.21 3.13%
Pakistan 1.24 3.82%
Panama 7.94 14.06%
Peru 0.85 1.86%
Philippines 1.01 2.19%
Paraguay -17.96 -45.19%
Sri Lanka -1.89 -5.82%
Thailand 6.95 14.66%
Tunisia 1.3 3.02%
Uganda -4.4 -13.33%
Zambia -9.09 -20.89%

Dates: The year for the beginning Gini index and ending
Gini index

Change: Calculated as the beginning Gini index minus the ending
Gini index (positive number indicates an
improvement in the country's income inequality)

Percent Change: Change in the Gini divided by the beginning Gini


Table 2

Correlations

 Change Corrup Fertility
 in Gini

Corruption -0.013
Fertility -0.267 -0.118
Foreign Inv. -0.028 -0.19 0.026
Freedom 0.088 -0.1 0.399
GDP 0.184 0.276 -0.502
GDP Growth -0.118 0 -0.210
Govt. Cons. 0.162 0.202 0.343
Illiteracy -0.059 0.03 0.572
Inflation -0.125 0.08 0.100
Land -0.360 -0.119 0.370
ODA -0.098 0.041 -0.151
Population -0.004 -0.066 -0.119
Sec. School 0.227 0.244 -0.634
Trade 0.15 0.148 -0.020

 For. Free GDP
 Inv.

Corruption
Fertility
Foreign Inv.
Freedom -0.005
GDP -0.044 -0.293
GDP Growth 0.045 -0.138 0.000
Govt. Cons. -0.016 0.287 0.283
Illiteracy -0.134 0.412 -0.575
Inflation 0.205 -0.167 0.069
Land 0.294 0.149 0.048
ODA -0.138 0.000 -0.373
Population -0.155 -0.096 -0.294
Sec. School 0.088 -0.254 0.503
Trade 0.236 0.060 0.469

 GDP Govt. Ill.
 Gro. Cons.

Corruption
Fertility
Foreign Inv.
Freedom
GDP
GDP Growth
Govt. Cons. -0.090
Illiteracy 0.029 0.196
Inflation -0.371 -0.204 -0.247
Land -0.090 0.254 0.012
ODA 0.286 -0.138 0.312
Population 0.268 -0.136 0.191
Sec. School -0.115 -0.026 -0.512
Trade -0.05 0.650 -0.278

 Inf. Land ODA

Corruption
Fertility
Foreign Inv.
Freedom
GDP
GDP Growth
Govt. Cons.
Illiteracy
Inflation
Land 0.191
ODA -0.220 -0.349
Population -0.133 -0.101 0.691
Sec. School 0.066 -0.386 0.096
Trade -0.273 0.096 -0.406

 Pop. Sec.

Corruption
Fertility
Foreign Inv.
Freedom
GDP
GDP Growth
Govt. Cons.
Illiteracy
Inflation
Land
ODA
Population
Sec. School 0.020
Trade -0.41 0.127

Table 3
Regression Results
Model with all explanatory variables (Dependent
Variable: Change in Gini)

 Regression Standard Error
 Coefficient

Constant 22.23574 13.451
Corruption -0.168028 3.007
Fertility -2.790028 1.656
Foreign Investment 2.483198 1.782
Freedom 0.621509 0.809
GDP 0.000159 0.003
GDP Growth -0.765483 0.674
Government Cons. 0.900801 0.448
Illiteracy -0.035686 0.099
Inflation -3.171945 5.386
Land -28.81735 10.995
ODA -7.25E-09 0.000
Population 1.03E-08 0.000
Secondary School -0.119644 0.106
Trade -0.133668 0.098

 Regression Statistics:

R-squared 0.521
Adjusted R-squared 0.043
S.E. of regression 5.385

 t-Statistic P-Value

Constant 1.653 0.121
Corruption -0.056 0.956
Fertility -1.684 0.114
Foreign Investment 1.393 0.185
Freedom 0.768 0.455
GDP 0.059 0.954
GDP Growth -1.135 0.275
Government Cons. 2.008 0.064
Illiteracy -0.36 0.725
Inflation -0.589 0.565
Land -2.621 0.02
ODA -1.885 0.08
Population 1.035 0.318
Secondary School -1.126 0.279
Trade -1.363 0.195

 Regression Statistics:

R-squared F-statistic 1.089
Adjusted R-squared Prob (F-statistic 0.438
S.E. of regression Durbin-Watson stat 2.379

Table 4
Regression Results
Dependent Variable: Change in Gini

 Regression Standard Error
 Coefficient

Constant 8.529097 5.237
Fertility -1.539249 0.902
Government Cons. 0.606070 0.289
Land Use -18.09975 7.522
ODA 4.41E-09 0.000
Trade -0.070857 0.059

 Regression Statistics:

R-squared 0.342
Adjusted R-squared 0.199
S.E. of regression 4.926

 t-Statistic P-Value

Constant 1.629 0.117
Fertility -1.706 0.102
Government Cons. 2.096 0.047
Land Use -2.406 0.025
ODA -1.768 0.090
Trade -1.201 0.242

 Regression Statistics:

R-squared F-statistic 2.392
Adjusted R-squared Prob (F-statistic) 0.069
S.E. of regression Durbin-Watson stat 2.299