Foreign Aid and Democratization in Post-conflict Societies

Foreign Aid and Democratization in Post-conflict Societies John Ishiyama University of North Texas Kathryn Sanders University of Missouri, St. Louis Marijke Breuning University of North Texas What is the relationship between development aid and the emergence of democracy in post-conflict societies? In this study we examine twenty six post-conflict countries that experienced civil wars ending after 1980. The dependent variable is measured using data from the Polity IV database (discussed below). The principal independent variables are the amount of aid provided in the periods following the conflict settlement, the timing of aid as well as ethno linguistic homogeneity/heterogeneity, economic performance prior to the conflict, the extent to which the state was democratic prior to the onset of the conflict, and the length of the conflict. We find that none of these variables affect the emergence of democracy eight years after the conclusion of the conflict, nor movement towards democracy in that period, except for the length of the conflict in years. Aid and the timing of aid have little or no effect on emergence of democracy in postconflict countries. Introduction What is the relationship between development aid and the emergence of democracy in post-conflict societies? Although there has been a considerable amount of literature that empirically investigates the relationship between development aid and corruption (Tavares. 2003; Alessina and Weder 2002; Knack 2000; Rimmer 2000;Svensson 1998; Ijaz 1996) aid and the quality of governance (Knack 2001) aid and ethnic conflict (Esman and Herring, 2003; Herring, 2001) and foreign aid and post-conflict economic growth (Collier and Hoeffler 2004; 2002; Hamburg 2002; Casella and Eichengreen 1994), no study of which we are aware has examined the direct effects of both the quantity and timing of development aid on promoting political democracy in post-conflict societies. Given the often publicly declared purpose of aid as promoting peace, stability and presumably in states torn by civil conflict (Adelman 2003; Carrothers 2003; Carapico 2002) it is important to empirically test the asserted relationship between foreign aid provision and democratic development and assess its effectiveness as a policy tool. Historically much of the literature on foreign aid has concentrated on the alleviation of poverty—if there were political effects these were considered secondary. To a large extent, it was assumed (consistent with the modernization paradigm familiar in the study of comparative political development) that political Midsouth Political Science Review, 2007-2008, Vol. 9 Ishiyama, Sanders, and Breuning | 20 development would naturally grow out of economic expansion. Thus, scholars such as Jeffrey Sachs (2005) focus on foreign aid primarily as a means to promote economic growth-- rather than examine the relationship between aid and democracy, such works have analyzed the effectiveness of foreign aid and the promotion of growth (see also Arvin and Barillas 2002). On the other hand, there are those who argue that foreign aid only creates dependence and economic inefficiency, with little or no evidence of economic growth (Kraay and Raddatz 2007; Frank and Baird 1975). More recent works have examined the effect of aid on political development, including the promotion of political and institutional stability (Hamburg 2002; Hartzell, Hoddie and Rothchild 2001). Others have examined the relationship between aid and corruption, and the effects of aid on “political ineffectiveness” (Djankov, Mondalvo and Reynol-Querol 2006; Werlin 2005; Knack 2001; Tornell and Lane 1999). There have also been a few studies that have examined the relationship between aid and democracy, such as Stephen Knack who in 2004 asserted that there is little evidence to support the proposition that aid promotes democracy (Knack 2004; see also Djankov et al 2006). These studies, however, focused on the broad group of lesser developed nations as opposed to post-conflict societies. However, it is upon these very post-conflict countries, from Bosnia to Rwanda that much of the world’s attention has been focused, and aid has been used specifically and explicitly to promote democracy and stability. Indeed, some literature suggests that foreign aid in post-conflict societies should be a far more effective tool than in developing countries generally. Indeed, Collier and Hoeffler assert that they are more responsive to the development assistance provided by affluent countries. “Aid is considerably more effective in augmenting growth in post-conflict situations than in other situations…aid volumes should be approximately double those in other situations” (Collier and Hoeffler 2002, p.13). Not only do Collier and Hoeffler contend that aid leads to economic growth in postconflict societies, but also that political stability and democracy will result. Thus, the truest test of the proposition that aid has a direct relationship to democracy is to test it using cases emerging from prolonged civil conflict. In this study we examine twenty six post-conflict countries that experienced civil wars ending after 1980. The dependent variable democracy-autocracy is measured using data from the Polity IV database (discussed below). The principal independent variables are the amount of aid provided in the periods following the conflict settlement, the timing of aid as well as ethno linguistic homogeneity/heterogeneity, economic performance prior to the conflict, the extent to which the state was democratic prior to the onset of the conflict, and the length of the conflict itself. 21 | Foreign Aid and Democratization Literature Review Most of the literature on foreign aid has focused on aid and economic development. For many scholars, large donations of aid are viewed as the primary way to alleviate the poverty-stricken countries of the globe. Aid can promote growth through the raising of living standards and breaking the “poverty trap”. Sachs, in The End of Poverty, illustrates the concept of the “poverty trap” of the Third World and claims that the only way to overcome it is with foreign assistance and Official Development Aid (ODA). We start with a household that is impoverished. All of its income goes to consumption, just to stay alive…The result is a fall in capital per person and a negative growth rate of per capita income…The solution is, where foreign help, in the form of official development assistance (ODA), helps to jump-start the process of capital accumulation, economic growth, and rising household incomes (Sachs 2005, p. 246). On the other hand, critics like Kraay and Raddatz (2007, p. 321) disagree with the “jump-start” theory asserting that aid does not necessarily lead to sustainable economic growth. Indeed, they are skeptical of the claim that “sufficiently large increases in aid will have disproportionate effects on economic growth in low income countries…[we] do not find evidence of threshold effects based on these leading explanations for poverty traps whereby sufficiently high levels of aid are necessary to “jump-start” a sustainable growth process” (Kraay and Raddatz 2007, p. 321). Some scholars also claim that the aid creates economic dependency and ultimately detracts from real economic growth (Chilcote 1978; Vengroff 1977). Lesser developed countries begin to rely on the aid for government spending and as a way to pay off debts. Thus if there is growth it is “façade growth”-- the recipient nation may not be able to sustain that growth without international funding. Some scholars such as Stephen Kosack examined the impact of aid on the quality of life (Kosack 2003). Although he found that aid does not affect quality of life generally, when combined with democracy, foreign aid is more likely to improve the quality of life in a given country. He concludes the democratization would make aid more effective in promoting growth and improvements in the quality of life. Although there have been studies that examine the impact of aid on economic expansion in post-conflict studies, there has been considerably less attention paid to the relationship between aid and the development of political structures in postconflict societies. Nonetheless in the last decade or so there has been a growing interest in the impact of aid on democracy and stability (Hartzell et al 2001; Casella and Eichengreen 2004). Gordon Crawford has examined the impact of conditionality on democracy from the perspective of four donor states (Crawford 2001). Stephen Ishiyama, Sanders, and Breuning | 22 Brown (2005) argues that international actors can, through development aid, play a very important role in promoting (or preventing) democratization. However, many strategic errors have been made particularly in promoting a rapid transition to a multiparty system without paying attention to the development of political infrastructure and stability. This rush to democracy can in fact impede further democratization. Knack (2004) notes that foreign aid can promote democracy in several key ways. Aid can provide for technical assistance in building and strengthening political institutions. Aid can also indirectly promote democracy by improving education and increasing per capita incomes (see also Almond and Powell 1965; Lipset 1959). Others, however, argue that increasing the amount of aid in lesser developed countries (LDCs) actually lowers the quality of governance (Knack 2001; Esman and Herring 2001; Tornell and Lane 1999). For example, Milton Friedman (1958) argues that since foreign aid goes directly to government this tends to strengthen the state sector relative to the private sector, and as a result, democracy is less likely to emerge since in aid-dependent countries most economic activity is controlled by the state. Similarly, aid levels may also reinforce executive dominance at the expense of other institutions in new democracies (Brautigam 2000). In addition, as Robert Herring and Milton Esman (2001, p. 11) argue “aid often has the potential to create the political substitution effect: meaning that the regime substitutes external props for domestic support. Politicians come to believe that support of the international financial institutions is more important than, and can substitute for, building domestic coalitions for governance.” Thus, the leaders of the lesser developed countries, because of the need for funding, begin to form policies based on what the donors want, and are far less responsive to the wants and desires of the country’s population. This form of aid “rentierism” thus has the great potential to detract from the development of democracy in post-conflict societies. In fact, foreign assistance may encourage political instability by making control of the government and aid receipts a more valuable prize (Grossman 1992). More development aid could also lower the quality of governance and democracy by adding fuel to corrupt government practices (Nye 1967). As several scholars have noted, donors tend to tolerate the corruption, as long as the nation puts into place policies that reflect the values of the donors (Hanlon 2004; Alesina and Weder 2002). Joseph Hanlon (2004, p.750) argues “that the donor community is prepared to tolerate quite blatant corruption if the elite rapidly puts into place “market friendly” policy changes…Donors like formal democracy, although they do not seem too concerned about how well it works”. Even though some donors may claim that good democratic governance is being instilled in these lesser developed nations, this usually means “market democracy”, that is defined by trade liberalization and free market values as opposed to political democracy itself. 23 | Foreign Aid and Democratization Some recent studies have examined the effects of aid and the spread of democracy within “lesser developed countries” (LDCs). Djankov et al.(2006, p.2) demonstrated that “foreign aid actually has a negative impact on the democratic stance of developing countries and on economic growth by reducing investment and increasing governmental consumption.” Knack in 2004 also studied democracy and aid, asserting that the relationship between the two was actually quite weak: “The evidence presented here does suggest that either the favorable impacts of aid on democratization are minor, or they are roughly balanced by other democracyundermining effects of aid dependence” (Knack 2004, p. 262). Nonetheless, both Djankov et al (2006) and Knack (2004) focused on LDCs generally as opposed to post-conflict societies per se. On the other hand, some scholars have argued that aid can mitigate conflict, and, presumably, promote democracy and stability. This is particularly true of countries torn by ethnic conflict. Milton Esman (2001, p. 237) notes, for instance, that “there is nothing predetermined about the interethnic effects of development assistance… Development assistance projects may contribute to aggravating conflict among already mobilized and politicized ethnic communities…[However] there is convincing evidence that development assistance can be managed in ways that avert or mitigate interethnic conflict”. In order to analyze the effects of foreign aid, two additional variables must be considered--- the timing of the aid and the type of political regime that was present before the conflict began. The timing of the level of allotment has shown to be significantly related to the stability (but not necessarily democracy) in a post-conflict nation (Patrick 2000; Collier and Hoeffler 2002). Casella and Eichengreen (1996) claim that aid provided right after the conflict gives the region more political and economic stability, whereas aid that is given later tends to have the opposite effect. Collier and Hoeffler (2002) go a step further asserting that aid is more effective with a “gradual rise during the first four years [after conflict], then gradually taper back to normal levels by the end of the first post-conflict decade”. Consequently the timing of the distribution of aid will also be analyzed in this study to see the relationship with the democratic development. In addition to the timing of aid, it has also been demonstrated that the previous type of regime will also have an effect on the potential development of democracy. Hartzell et al (2001) assert that if the regime prior to the conflict was a democracy then peace after conflict is a more likely result and is more resilient. Finally, another variable that may affect political developments after the end of the civil conflict is the length of the conflict itself. Indeed many scholars in the conflict resolution literature have focused on the importance of “conflict fatigue” which is the point after many years of the stalemated struggle where the populations of both opposing groups simply tire of the conflict, and looks for means, often less than perfect, to resolve it. At this point, there is often a pragmatic leadership on both Ishiyama, Sanders, and Breuning | 24 sides that emerges to forge a consensus that contends that conflict cessation, and perhaps a potential normalization of relations, has more benefits than continuing with the struggle (Touval and Zartman 1985). In terms of democracy, this rough balance of representation of different interests may provide the seeds of accommodation of different interests, which ironically may bode better for democracy than short intense (and unresolved) conflicts, especially those that are stopped by outside intervention. Design and Methodology In this paper, we are primarily concerned with testing the impact and timing of foreign aid on the establishment and growth of democracy in post-conflict societies. For this study twenty-six post-conflict societies will be examined: Angola, Armenia, Azerbaijan, Bosnia-Herzegovina, Burundi, Chad, Croatia, The Democratic Republic of Congo, Republic of Congo, El Salvador, Ethiopia, Georgia, Guatemala, Indonesia, Iran, Morocco, Mozambique, Nicaragua, Nigeria, Peru, Philippines, Russia, Rwanda, Somalia, Tajikistan, and Uganda. These nations are derived from an adjusted list from Collier and Hoeffler (2002) and all have had major civil conflicts (defined as resulting in at least 500 combat deaths and either resulting from wars of contention for national power (as in Congo-Kinshasa) and wars of separation (such as East Timor in Indonesia and Chechnya in Russia).1 Low level insurgencies were not included in this data set. Also the data set was limited to only the first conflicts after 1980. To be sure, many of these conflicts were only temporarily resolved, but they resumed later (as in Angola and Chechnya in Russia). However we concentrated only on what happened eight years after the first resolution (thus if these countries descended again into significant conflict this was generally reflected in the Polity scores) but we focused on what happened eight years after the first resolution. Further we did not include conflicts where eight years have not yet passed since the conflict was resolved (such as in Sudan between the Khartoum government and the Southern separatists – settled in 2005, or the recently resolved Nepalese conflict) or on continuing conflicts (as in Sri Lanka. Finally we examined only foreign aid recipients (not net donors, so neither the United Kingdom nor Spain were included in this study). Using these decision rules, 26 countries were included in this study. Again, cases of post-conflict countries were chosen because they are likely to better reflect the potential political changes caused by foreign assistance compared to other developing countries. In order to evaluate the relationship between aid and democracy, it is first necessary to identify the relevant variables. For this study, the focus is on the extent 1 Azeribaijan and Armenia were also included- although the war between these two states appeared to be an interstate war, the primary conflict was a war of separation in Nagorno Karabakh, that sought amalgamation with Armenia. Hence, we (like Collier and Hoeffler) included this as a case of civil conflict. 25 | Foreign Aid and Democratization to which aid is provided for the country that has just suffered some internal conflict. As noted above, many scholars claim an inverse relationship between the amount of aid and the quality of governance (Knack 2004; Djankov et al 2006). To measure the amount of aid received, one of the primary independent variables is the amount of aid provided per capita in 2000 constant dollars in the 8 years following conflict. The 8 year rule was used because generally this meant that at least one election (either presidential, legislative, or a referendum or plebiscite) could be held after the end of the conflict—given that elections are critical to any definition of democracy, this would be a reasonable cutoff point in terms of time passed. In order to capture some of the dynamism of the aid and democracy relationship, we also examined the average aid received per year per capita across the eight years from the end of the conflict. The data for the aid variables are derived from the World Bank”s World Development Indicators. In addition, the timing of the aid is also important. With the findings of Collier and Hoeffler (2002) one would expect that if the aid were to be phased in, with the larger sums arriving at least 4 years after the end of the conflict, then democracy would flourish. In order to measure the timing of the aid, a dummy variable is employed. If most of the aid was received in the first four years (years 1 through 4) after the conflict it will be coded a “0”. When a country received most of its aid in the second four year period (years 5-8 after the conflict) then it will be coded as a “1”. Based on the previous literature, one would expect that if most of the aid were to come in the second four year period that the country would be more democratic. Another variable to consider is the extent to which the country was divided along ethnic lines. Indeed, if a post-conflict society is deeply divided along ethnic lines it may negatively affect democratic development. In order to measure the extent to which countries are ethnically divided we employ the Ethnolinguistic Fractionalization Index (ELF). The ELF was developed by Russian demographers in the 1960s and later modified by Easterly and Devine (1998) as a way to measure the effects of ethnicity on the economic development of African nations. We use the data provided by Easterly and Devine. The economic and political situation of the country before the conflict is also important to the effectiveness of foreign aid on democracy. Thus, we examine two additional independent variables are employed-- estimated pre conflict GNP per capita and pre conflict regime type. In order to measure pre conflict wealth, the estimated average GNP per capita for the decade before the end of the conflict will be used. To measure the next two variables (the final independent variable and the dependent variable), we employ the often used Polity IV data base. The Polity IV measure was developed at the University of Maryland College Park in order to demonstrate the degree to which a political system is autocratic or democratic. It uses different measures of constraints on the chief executive, the openness and competitiveness of executive recruitment and the competitiveness of political Ishiyama, Sanders, and Breuning | 26 participation. These scores are then combined onto a scale ranging from -10 to +10, with -10 being the most autocratic and +10 being the most democratic. In order to measure the past regime type, the Polity IV score will be used for the year before the start of the conflict within that nation. As mentioned above the pre-conflict regime type is also an important independent variable. As noted above, democracies are more likely to recover than non democracies. Put another way, regimes that were less autocratic are more likely to redemocratize than regimes that were autocratic prior to the conflict. If the state before the conflict was autocratic, democracy building is far more difficult then regimes that “return” to democracy. In order to measure the type of regime the state had before the conflict we recode the combined polity score in the year just prior to the onset of the conflict. For countries that scored less than zero these were wholly autocratic and scored a “1”. For those countries that scored zero and higher, these were relatively less autocratic and scored a “0”. The measure differentiates between autocratic previous regimes, and less autocratic regimes.2 In order to fully study the effects of aid on democracy in post-conflict societies we also measure the level of democracy after the conflict. This dependent variable is measured through the use of the combined Polity IV score for 8th year after the end of the conflict. The reason why we use resulting score after eight years, as opposed to a yearly measure, is that we are most interested in whether democracy results (or not). To merely examine “first differences” or changes over time would weight a score moving from -10 to -7 equally to a score that moves from +1 to a +4. Thus, countries that remain thoroughly authoritarian would remain thoroughly authoritarian, but would have a score equal to a country that is becoming clearly democratic. Hence we focus on outcome as opposed to change in terms of the polity score in this paper. In sum, Table 1 on the following page reports the data for this project. Results Table 2 reports the results of an Ordinal Logistic regression analysis (ordered logit) for the dependent variable of the combined polity IV score for 2003. Since the dependent variable is ordinally measured (as is the Polity IV measure for democracy/autocracy which ranges from -10 to +10), the ordered logit procedure is most appropriate for this situation. 2 Although one might consider keeping the ordinal measure as opposed to dummying up these variables, we have opted to use dummies, largely because of the critique many scholars have made regarding the use of endogenous ordinal variables in regression equations (for a summary see Fox, 1991). Essentially there is some question of interpreting the coefficient sizes associated with ordinal variables, and what they mean in regression analysis. A preferred technique is thus using dummy variables (given that what is interpreted generally is the significance and sign of the coefficient). 27 | Foreign Aid and Democratization Table 1. Data for Independent and Dependent Variables Years of Conflict 75-91 88-94 88-94 91-94 88 80-88 91-95 Aid Provided Timing of Aid Dummy 1 0 1 0 1 0 1 Ethno Linguistic Fractionalization Index 0.78 0.13 0.31 0.67 0.31 0.86 0.42 1 Country Angola 28.00 Armenia 60.41 Azerbaijan 21.56 Bosnia 203.33 Burundi 40.25 Chad 39.22 Croatia 18.44 Dem Rep of 96-97 22.22 1 0.90 Congo Rep of 97 32.63 0 0.68 Congo El Salvador 79-92 49.67 0 0.16 Ethiopia 74-91 16.75 0 0.77 Georgia 91-93 44.89 0 0.49 Guatemala 78-84 20.89 1 0.76 Indonesia 75-82 6.56 1 0.76 Iran 81-82 1.00 1 0.75 Morocco 75-89 29.11 0 0.40 Mozambiqu 76-92 65.44 0 0.70 e Nicaragua 82-90 129.89 1 0.39 Nigeria 80-84 1.78 1 0.86 Peru 82-96 17.33 1 0.51 Philippines 72-96 8.44 0 0.86 Russia 94-96 2.32 0 0.33 Rwanda 90-94 63.89 0 0.26 Somalia 88-92 . . 0.03 Tajikistan 92-97 19.56 1 0.55 Uganda 80-88 33.22 1 0.92 1. (pc-- 2000 constant US$ avg per year for 8 years following conflict 2. 8th year after conflict ended Estimated GNP/pc prior to conflict 1875.00 3309.00 2617.00 10065.18 840.98 765.36 10065.18 Autocracy dummy 0 0 0 0 0 1 0 Postconflict Polity Score2 -3 5 -7 0 11 -2 7 958.60 1 0 1244.67 1 -4 5095.73 866.00 4499.75 3633.57 1100.00 4897.99 2395.00 0 0 1 0 0 0 0 7 1 5 8 7 2 -6 694.00 0 6 4555.98 935.25 5512.66 3543.00 7576.47 1175.82 0.00 2096.92 892.00 0 1 1 0 1 0 0 0 0 8 4 9 8 7 -3 0 -3 -4 Table 2. Coefficient Estimates and Colinearity Statistics, Ordered Logit Procedure Average Aid Per GDP provided in post-conflict period Timing of Aid Dummy ELF Pre conflict GDP/capita Pre Conflict Polity Dummy Length of Conflict in years Coefficient Std. Error VIF .002 .01 1.28 1.46 1.08 1.11 .00 .00 1.94 -3.57 1.86 .28*** 1.99 1.17 .10 1.73 1.10 1.20 Pseudo R-Square = .12; * = p < .10; ** = p < .05; *** = p < .01 Ishiyama, Sanders, and Breuning | 28 As indicated in Table 2, neither the amount of aid per capita on average provided for the first eight years after the conflict ended (coefficient =.002 and standard error = .01), nor the timing of aid (coefficient =1.46 and the standard error = 1.08) was statistically significant in relationship with the dependent variable of the postconflict democracy score (i.e. neither had a p value of less than or equal to .05). This generally supports the findings of scholars like Knack (2004) who found similar results regarding democracy scores and aid for LDCs in general. To be sure, it should be noted that these findings do not necessarily refute the previous findings by scholars like Collier and Hoeffler (2002) who argued that aid promotes growth (but perhaps not democracy) or Casella and Eichengreen (1996) or Patrick (2000) who argued that aid promotes stability. Nonetheless, there is little evidence that foreign aid in post-conflict countries supports the growth of democracy. In addition, the results do not lend support to the assertion made by Hartzell et al (2001) that the regime type prior to the conflict has any effect on the emergence of democracy after the conflict (again where the coefficient for the pre conflict dummy variable for democracy was 1.86 and the standard error was 1.17) Further the alterative contextual variables (the ELF and the GNP per capita prior to the conflict) also had little in the way of an effect on the post-conflict development of democracy (neither coefficient was statistically significant). However, there is the possibility that the significance of the coefficients are distorted by problems with multicolinearity. Multicolinearity results when the independent variables have some significant interdependence, which can result in numerically unstable estimates of the regression coefficients (small changes in X can result in large changes to the estimated regression coefficients). The variance inflation factor (VIF) can be used to detect if multicolinearity is problem – generally if the VIF score exceeds “4” than it is a potential problem. However as indicated for the VIF scores for each of the independent variables in the model there is little in the way of a multicolinearity problem with the model (VIF scores are all less than 2). The one independent variable that does exhibit a positive and statistically significant relationship is the length in years the conflict took place (coefficient = .28 and standard error = .10). Interestingly the positive sign of the coefficient indicates that the longer the conflict took place, the more likely the regime scored higher on the Polity IV combined polity score in the eighth year following the end of the conflict. Although beyond the scope of this project, this is a curious result. Perhaps the longer the conflict the more exhausted the combatant parties become, and the more willing they are to seek an enduring accommodation. It could also indicate that the longer the conflict, the more like the sides are equally matched, suggesting that a post-conflict settlement requires some power sharing agreement that guarantees representation of government and opposition. Whatever the case, fully investigating the relationship between the length of the conflict and the development of democracy is currently beyond the scope of this project. 29 | Foreign Aid and Democratization Perhaps the amount and timing of aid affects the progress towards democracy, i.e. that although aid may not produce democracies as an outcome, they are more likely to promote movement towards democracy? To test this proposition, we operationalized the dependent variable as the difference between the polity IV autocracy-democracy score, comparing the value in the year the conflict ended and eight years after the end of the conflict. This variable was coded as a binary result, 0 if no positive movement was recorded, and “1” if there was at least positive movement (meaning a move from a lower value on the polity scale, to a higher value). This provides us with some sense of movement over time, and tests the proposition that the provision of development aid may not result in democracy, but at least promotes some movement towards democracy. The results in Table 3 are remarkably similar to the results reported in Table 2-again neither the amount of aid per capita on average provided for the first eight years after the conflict ended (coefficient =-.01 and standard error = .02), nor the timing of aid (coefficient =.55 and the standard error = 1.26) was statistically significant in relationship with the dependent variable of the post-conflict democracy score . The pre conflict GDP/capita and Democracy dummy variables were also not significant, which is consistent with the findings in Table 2. Further again, the longer the conflict, the more likely the move towards democracy (but the relationship is somewhat less robust). The only difference in terms of the change model as opposed to the outcome model in Table 2 is the ELF score—in Table 3, the greater the degree of ethnolinguistic fractionalization, the more likely the country would move towards democracy (this is an interesting finding, and may have to do with the greater attention to settlement issues and issues of representation in countries that emerge from ethnically based conflicts). Table 3. Coefficient Estimates and Colinearity Statistics, Binary Logit Procedure Coefficient Std. Error VIF Average Aid Per GDP provided in post-conflict period Timing of Aid Dummy ELF Pre conflict GDP/capita Pre Conflict Polity Dummy Length of Conflict in years -.01 .02 1.28 .55 1.26 1.11 9.40** 4.60 1.94 .00 .04 .26* .21 1.78 .15 1.73 1.10 1.20 Pseudo R-Square = .39; * = p < .10; ** = p < .05; *** = p < .01 Ishiyama, Sanders, and Breuning | 30 Conclusion What is the relationship between development aid and the development of democracy in post-conflict societies? The above results indicate that there is no evidence to support the notion that aid (the amount or the timing) has a positive effect on the development of democracy in post-conflict societies either in terms of outcome nor in terms of movement towards democracy once the conflict has ended. This supports the findings of some scholars (like Knack and Djankov et al) who argued that aid has no effect on democracy generally among developing countries. Thus, even when testing under the more rigorous conditions of the post-conflict societies, there was no relationship between neither the timing nor the amount of aid on the subsequent level of democracy. This finding also does not support some other scholars who contended that aid may have a positive effect on democracy. Although it may be the case that the ultimate effects of foreign aid can only be discerned after many years have passed, the initial track record of aid provision and the timing of aid has not been particularly promising when it comes to building democracy in post-conflict countries. One very interesting finding, although this is beyond the scope of the current study, is the relationship between the length of the conflict and the development of democracy-- the longer the conflict the more democratic the country. This might support the literature from conflict studies (like Touval and Zartman, 1985) that suggests that the longer the conflict the more likely a military stalemate emerges, thus compelling the different sides to seek a lasting settlement and accommodation. This bodes better for democracy than short intense (and presumably less likely to be fully resolved) conflicts. Although these findings are somewhat preliminary, the above results suggest further questions for future investigation. First, what is the precise relationship between the length of the conflict and the provision, and timing, of foreign aid. Perhaps the provision of aid works best under conditions of short and quick conflicts, or vice versa. Second, perhaps it is not the amount of aid, nor the timing, that makes a difference in promoting political democracy, but the type. In other words, perhaps aid that is designed to promote civil society institutions may be more effective and providing voice to potentially disaffected populations and hence better promote political democracy. Although the evidence thus far cannot yet answer these questions, the findings above indicate that further investigation into political consequences of foreign aid in post-conflict societies is warranted, and represents a potentially promising avenue for future inquiry. 31 | Foreign Aid and Democratization References Adelman, Carol (2003) “The Privatization of Foreign Aid: Reassessing National Largesse,” Foreign Affairs 82: 9-14. Alesina, Alberto, and Beatice Weder (2002) “Do Corrupt Governments Receive Less Foreign Aid?” The American Economic Review 92: 1126-1137. Almond, Gabriel and G. Bingham Powell (1965) Comparative Politics: A Developmental Approach Boston: Little, Brown. Arvin Mark and Francisco Barillas (2002) “ Foreign Aid, Poverty Reduction, and Democracy”, Applied Economics 34:2151-56. Brautigam, D. (2000) Aid Dependence and Governance. Stockholm: Almqvist & Wiksell International Brown, Stephen (2005) “Foreign Aid and Democracy Promotion: Lessons from Africa” TheEuropean Journal of Development Research 17:179-198. Carapico, Sheila (2002) “Foreign aid for promoting democracy in the Arab world” The Middle East Journal 56: 379-395. Carothers, Thomas (2003) “Promoting Democracy and Fighting Terror” Foreign Affairs 82: 84-97 Casella, Alessandra, and Barry Eichengreen (1996) “Can Foreign Aid Accelerate Stabilization?” The Economic Journal 106: 605-619. Chilcote, Ronald H. (1978) “A Question of Dependency”. Latin American Research Review 13: 55-68. Collier, Paul, and Anke Hoeffler (2002) “Aid, Policy and Growth in Post-Conflict Societies”. World Bank Policy Research Working Paper 2902 October Crawford, Gordon (2001) Foreign Aid and Political Reform: A Comparative Analysis of Democracy Assistance and Political Conditionality. Basingstoke: Palgrave. Djankov, Simeon, Jose Montalvo, and Marta Reynal-Querol. (2006) “Does Foreign Aid Help?” Cato Journal. 26:1-28. Easterly W. and R. Devine (1998) “Africa”s growth tragedy: policies and ethnic. Divisions”, Quarterly Journal of Economics 112: 1203–1250. Ishiyama, Sanders, and Breuning | 32 Esman, Milton J. (2001) “Policy Dimensions: What can Development Assistance Do?” In Carrots, Sticks and Ethnic Conflict: Rethinking Development Assistance, ed. Milton J. Esman and Ronald J. Herring. Ann Arbor: University of Michigan Press Esman Milton J. and Ronald J. Herring (2003) Carrots, Sticks, and Ethnic Conflict : Rethinking Development Assistance Ann Arbor: University of Michigan Press Fox, John (1991) Regression Diagnostics. Englewood Cliffs: Sage. Frank, Charles R., and Mary Baird (1975) “Foreign Aid: Its Speckled Past and Future Prospects”. International Organization 29 (Winter): 133-167. Friedman, Milton (1958) “Foreign Economic Aid: Means and Objectives”. Yale Review 47: 500–516. Grossman, H.I (1992) “Foreign Aid and Insurrection”. Defense Economics 3: 275– 288. Hamburg, David A. (2002) No More Killing Fields: Preventing Deadly Conflict. Lanham, MD: Rowman & Littlefield Publishers. Hanlon, Joseph (2004) “Do Donors Promote Corruption?: The Case of Mozambique”. Third World Quarterly 24:747-763 Hartzell, Caroline, Matthew Hoddie, and Donald Rothchild (2001) “Stabilizing the Peace after Civil War: An Investigation of Some Key Variables”. International Organization 55 (Winter): 183-208 Herring, Ronald J. and Milton J. Esman (2001) “Projects and Policies, Politics and Ethnicities”. In Carrots, Sticks and Ethnic Conflict: Rethinking Development Assistance, ed. Milton J. Esman and Ronald J. Herring. Ann Arbor: University of Michigan Press Ijaz Mansoor (1996) “A Primer on the Perils of Foreign Aid,” Wall Street Journal, October 2. Knack, Stephen (2001) “Aid Dependence and the Quality of Governance: CrossCountry Empirical Tests”. Southern Economic Journal 68: 310-330. ----. (2004) “Does Foreign Aid Promote Democracy?” International Studies Quarterly 48:251-266. 33 | Foreign Aid and Democratization Kosack, Stephen (2003) “Effective Aid: How Democracy Allows Development Aid to Improve the Quality of Life”. World Development 31:1-22. Kraay, Aart, and Claudio Raddatz. (2007) “Poverty Traps, Aid and Growth”. Journal of Development Economics. 82: 315-347. Lipset, Seymour M. (1959) “Some Social Requisites of Democracy: Economic Development and Political Legitimacy”. American Political Science Review 53: 69–105. Nye, Joseph S. (1967) “Corruption and Political Development: A Cost-Benefit Analysis”. The American Political Science Review 61: 417-427. Patrick, Stewart (2000) “The Check is in the Mail: Improving the Delivery and Coordination of Post-Conflict Assistance”. Global Governance 6 (Jan-Mar): 6193. Rimmer, Douglas (2000) “Review Article: Aid and Corruption”, African Affairs, 99(392): 121-128 Sachs, Jeffrey, J.W. McArthur, G. Schmidt-Traub, M. Kruck, C. Bahadur, M. Faye, and G. McCord. (2004) “Ending Africa”s Poverty Trap”. Brookings Papers on Economic Activity, No 1: 117-240 Sachs, Jeffrey (2005) The End of Poverty. New York: The Penguin Press Svensson, Jakob (1998) “Foreign Aid and Rent Seeking”. World Bank Policy Working Paper 1880 February. Tavares, Jose (2003) “Does Foreign Aid Corrupt?” Economics Letters 79 (2): 99– 106 Tornell, Aaron, and Philip Lane (1999) “The Voracity Effect”. The American Economic Review 89: 22-46. Touval, Saadia and I. William Zartman. (1985) International Mediation in Theory and Practice. Boulder, Colorado: Westview Press/Foreign Policy Institute Vengroff, Richard. (1977) “Dependency and Underdevelopment in Black Africa: An Empirical Test”. Journal of Modern African Studies 15: 613-630. World Bank (2006) A Decade Measuring the Quality of Governance. Washington D.C.: The World Bank. Ishiyama, Sanders, and Breuning | 34 Werlin, Herbert (2005) “Corruption in a Third World County”. World Affairs 168: 79-84. View publication stats