Capital flows to developing countries: Essays on the behavior of aid donors and official creditors
MetadataShow full item record
- Department of Economics 
The capital flows from bilateral and multilateral donors to developing countries have increased considerably since the start of foreign aid in the 1950s. During the same period, the policies of aid allocation have changed, the number of aid donors has increased, and multilateral aid organizations have taken a more dominant position in the global aid systems. Thus, how donor countries distribute their aid budgets, which is paid by their citizens, has received great attention by researchers. The lack of results on economic development and welfare in the recipient countries has also contributed to the interest in this field. This thesis consists of an introductory chapter and three essays on the aid policies of donor countries and official lending to developing countries. The main objective has been to investigate further how bilateral and multilateral donors behave, focusing on their policies on the allocation of loans, concessional or non-concessional, and grants as well as the motivation for bilateral donors when delegating the responsibility of aid policies to multilateral aid organizations. Aid donors choose whether to disburse aid bilaterally, where the donor government controls the allocation and implementation of aid projects, or to delegate this responsibility to an agent, typically a multilateral aid organization. In the first essay, “Poverty aversion and delegation of aid policies”, I address the question naturally arising from this behavior: Why do donors delegate the responsibility for aid allocation to multilaterals? Using panel data on aid disbursements from 23 Development Assistance Committee (DAC) donor countries for the period from 1987 to 2011, I test a dynamic model for the decision to delegate. Focusing on the predictions from theories on the Samaritan’s Dilemma and time inconsistency in aid allocation, I analyze how the relative poverty aversion of bilateral and multilateral aid agencies affects the share of total aid budgets delegated. The choices donors make when allocating aid across countries and deciding whether to delegate the responsibility for aid allocation to an agent both reflect the donors’ motivations for aid and influence the efficiency of aid, with respect to economic development in the recipient countries. The results show that the share of multilateral aid is negatively related to average income and population size, and positively related to the degree of openness. An increase in the level of corruption, reflecting the quality of institutions in the donor country, also increases the share of aid budgets delegated to multilaterals. While there is some support for the prediction that delegation is a convex function of the relative poverty aversion of bilateral and multilateral aid agencies, the result does not hold once donor interests are controlled for. Thus, the results indicate that the characteristics of the donor country are more important when donors decide whether or not to delegate, and not the possibility to alleviate the commitment problem when donors have a strong or weak aversion to poverty. In the second essay, “Partner country ownership: Does better governance and commitment to development attract general budget support?”, I exploit disaggregated data on official development assistance (ODA) commitment from the Creditor Reporting System (CRS) to test whether better governed countries and countries with stronger commitment to development are more likely to receive general budget support. The data used in the analysis cover 23 DAC donor countries and 115 recipient countries from 1995 to 2009. Comparing the results using disaggregated and aggregated data, I confirm that the results are sensitive to the data used. As expected, I find that donors are selective when looking at the allocation of general budget support GBS, while the effect of the quality of governance and commitment to development is not significant at conventional levels when using data on total program aid. The results are in line with existing empirical evidence suggesting that the use of aggregate data on aid flows gives an inaccurate picture of the degree of selectivity among donors, and shows that donors do follow the recommendation of being more selective when allocating budget support than with other types of aid. Still, variables indicating political and historical ties between the donor and recipient countries have a strong effect on both the probability of receiving GBS and the volume received. The third essay, “Lending to developing countries: How do official creditors respond to defaults?”, which is co-authored with Cathrin N. Fløgstad, is related to the literature on aid allocation as well as the empirical literature on reputational costs of sovereign defaults. The focus in the existing literature on reputational costs is on defaults and exclusion from international capital markets. However, most developing countries, and especially low-income countries, are not considered to be creditworthy by private creditors and therefore mainly rely on grants and loans from official sources. Thus, for countries that only rarely have access to international capital markets, the effects of sovereign defaults on disbursements of new loans from official creditors are more important. We also discuss whether countries defaulting on their sovereign debt can turn to official creditors for capital. Using data on 118 low- and lower middle-income countries for the period from 1972 to 2011, we analyze the effect of sovereign defaults on disbursements of concessional and non-concessional loans from bilateral and multilateral creditors. Separating bilateral and multilateral, and concessional and nonconcessional lending, we find that disbursements of new concessional loans and bilateral non-concessional loans are negatively related to an increase in arrears on principal and/or interest, on average. The effect is robust and significant at conventional levels. There is also a negative relationship between arrears and disbursements of multilateral non-concessional loans, but the statistical significance of this effect depends on how the arrears are measured. While the existing literature has found that countries are excluded from international capital markets following sovereign defaults, our results show that access to capital from official creditors is also reduced. Thus, countries cannot simply turn to official creditors for loans after a default.