International aid and political economy
- The Incentives of Donors and Aid Agencies
- The Interaction between Donors and Country Recipients
- The Delivery of Aid in the Recipient Countries
- The Market for Aid
International donors provided more than U.S. $2.4 trillion in the formof foreign aid from1960 to 2004. In many African countries, for example, the size of the official development assistance (ODA) divided by gross domestic product (GDP) exceeded 10 percent during that period. Despite this large amount of financial aid, the economic performance of the recipient countries has been disappointing. The poor macroeconomic impact of foreign aid raises questions about its efficiency. Political economy considerations can provide explanations for the scarce effectiveness of foreign aid. By the term political economywemean the allocation ofmarkets and other economic institutions subject to the political environment in which these institutions work. In particular, the process of decision and delivery of foreign aid present multiple agency problems, derived from the existence of asymmetric information (the donor and recipient do not have access to the same information) and moral hazard (receipt of aid causes the recipient to exert less effort to solve its own problems), and inappropriate institutional designs.
Most of the academic research finds that international aid is ineffective in fostering economic growth. For instance, Boone (1996) finds that aid does not significantly increase investment or any human development indicator, but it increases the size of the government. Burnside and Dollar (2000) find that aid works, but only when there are good fiscal, monetary, and trade policies. Easterly, Levine, and Roodman (2003) dispute these results, however, since they are sensitive to changes in the sample period or filling in the missing observations.
One of the basic reasons that development aid does not seem to work is the presence of perverse incentives and faulty institutional designs. The political economy perspective treats foreign aid as the result of the incentives and motivations of the donors, the response of country recipients, and the interaction between donors and recipients.
The Incentives of Donors and Aid Agencies
Efficient public programs are supported by the existence of sufficient (perfect) information and the accountability of the representatives. Households in developed countries are, at the same time, clients and citizens. As clients, they have information about the programs intended for their benefit and as citizens in democratic countries they can hold politicians accountable for their poor performance. In the case of foreign aid this feedback loop is absent: taxpayers in donor countries (citizens) are not the beneficiaries of the aid, and they do not have information about its effectiveness; and the intended beneficiaries (clients) donothave the right to vote in donors’ countries and, therefore, cannot affect the politicians in charge of approving the aid programs. We can describe this situation as a broken information feedback loop (Martens et al. 2002). In the absence of effective feedback from the intended beneficiaries of aid, the interests of consultants and suppliers of goods shape many decisions of aid agencies: these groups are the direct beneficiaries (control the information about the programs) and have direct influence on the decision- makers of donor countries.
Donors and aid agencies also have multiple objectives and represent multiple principals that often do not share the same objectives. For these reasons the institutional design of official aid agencies is difficult even if donors have good intentions. First, the multiplicity of objectives and the difficulty of performance measurement make it complicated to link the salaries of workers in aid agencies to their performance. Therefore, aid bureaucracies tend to define their output in terms of money disbursed (or volume of loans) instead of the rate of poverty reduction or economic development of recipient countries (Easterly 2003). Second, aid bureaucracies tend to develop their own procedures and increase the complexity of tenders and contracts (see Easterly 2006). Finally, aid agencies will tend to choose projects that minimize the risk of bad publicity, not the ones that maximize, for instance, the reduction of poverty.
Donor countries may also have objectives that undermine the effectiveness of foreign aid. In fact, almost half of the foreign aid provided by the members of the Organisation for Economic Co-operation and Development does not have as basic objectives poverty alleviation and economic development. Alesina and Dollar (2000) find that while some donors have the right incentives (raising income level, poverty reduction, institutional improvement), many others are driven by political and strategic considerations. The wrong incentives include targeting aid to inefficient and nondemocratic former colonies or developing countries that vote in favor of the proposals of the donor at the UN. The strategic interest in the Middle East explains why Egypt and Israel receive so much economic support from Western countries.
Another possible reason for the ineffectiveness of foreign aid is the so-called warm glow effect. Donor citizens may be interested only in the fact of giving itself and not in its effects on the recipient countries. Therefore, the utility of donors does not depend on thewell-being of the citizen of recipient countries but rather on the satisfaction the donors derive from providing aid to developing countries. This interpretation is consistent with the evaluation of policies based on inputs (money disbursed) rather than on outcomes (effect of the aid on poverty and economic development). It is also consistent with voters who are worried about the total amount of money that is dedicated toforeign aid (say0.7 percentofGDP) and not about the results that aid produces in recipient countries.
Finally, since there are multiple principals (donors and agencies), the effectiveness of foreign aid depends also on the degree of coordination among them. The need for coordination has increased over time because the level of fractionalization, or diversity, of donors (many donors with a small share of aid) has increased from0.25 in 1960 to 0.68 in 2003. Fractionalization is measured as 1 minus the sum of the squares of the shares of different donors. Empirical evidence suggests that higher donor fragmentation is associated with a reduction in the bureaucratic quality in recipient countries.
The Interaction between Donors and Country Recipients
The relationship between donors and country recipients is usually subject to principalagent problems and moral hazard. The relationship between a principal (the donor) and an agent (the country recipient) can lead to problems when the interests of the principal and the agent do not coincide and the principal has less than perfect information about the actions of the agent. The usual game theoretical solution to these types of problems is to write a contract, or design an incentive mechanism, to force the agent to perform as the principal wants. This incentive mechanism usually takes the form of aid conditionality: the requests associated with the concession of some types of foreign aid. The credibility of the threat to retire aid if the conditions of development of the projects are not satisfied is critical for the effectiveness of the conditionality, however. Recipient countries realize that these threats are not credible because donors do not have strong commitment ability. Donors may want to punish countries that do not fulfill their promises but, at the same time, they are pressed by their own constituencies to continue helping poor countries. The problem is reinforced by the fact that the status of themanagers within the aid agencies is determined by the size of their budget and the disbursement of funds. Empirical studies show this lack of commitment technology. For instance, Burnside andDollar (2000) do not find evidence of a significant positive relationship between aid flows (ODA) and good policies. Other studies find that there is no relationship between the developing countries’ score on quality of governance (corruption, democracy, service delivery, etc.) and who receives adjustment loans, for instance, fromthe World Bank.
Some authors have proposed alternative institutional designs to overcome the lack of credibility of donor threats. Svensson (2000) shows that, when there is no commitment technology available for donors, the delegation of the aid budget to an institution with less aversion to poverty (underdevelopment) than the donor will improve the welfare of the citizens of the recipient country since itmakes the threat credible. This theoretical result has yet to be implemented, however. If the aid game implies a donor and several recipients’ countries, or a pool of projects that are disbursed to each individual country depending on its relative performance, it is possible to design an aid tournament in which potential aid recipients compete for funds. In a bilateral relationship the recipient country has more bargaining power and, therefore, the credibility of the donor’s threats is weak. If the donor can choose between different recipients, however, then recipient countries have to commit to a high level of reformeffort if they want to win the aid tournament. In fact, donors can impose a cost on the recipients (for instance, they may ask for a set of reforms even to get considered in the tournament). This is the traditional ‘‘stick and carrot’’ type of strategy. In addition, having alternative countries to receive the aid of a donor reduces the ability of one particular recipient country to use holdup strategies.
Another problem associated with the moral hazard present in the relationship between donors and developing countries is the so-called Samaritan’s dilemma (Gibson et al. 2005), which can be represented as a two-person game. The donor (Samaritan) wants to help a poor country to get out of poverty. Therefore, it gets a higher payoff from helping, or giving a large amount of help, than from not helping (or providing less aid). The recipient has to decide how much effort to exert. Obviously, the donor prefers the big-effort action. The recipient country, however, prefers to exert little effort. If the game is repeated, as happens in real interactions between donors and developing countries, then we get little effort and high levels of foreign aid. There are many examples of this situation. Donors give large amounts of food aid to developing countries with food crises. Recipient countries know this and, therefore, they have incentives to relocate their own resources away from agriculture. This action will increase the likelihood of food shortages in the recipient country, which will generate constant food aid from the donor. Then the recipient country produces less food than necessary, and the developed country sends food aid on a regular basis.
The solution to this principal-agent problem is to monitor the recipient country. However, the lack of information and the absence of accountability of the bureaucracies of recipient countries make monitoring a nonviable solution. Another possible solution for this type of problem is to threaten to withdrawthe aid in the future. If this threat is credible it can support the best outcome (high level of aid and high level of effort, which economists describe as a Pareto efficient equilibrium). In game theoretical terms the donor threatens the recipient country with ending in the worst situation for both agents (low level of aid and low level of effort, which economists call, in this case, Nash equilibrium). However, donors are usually unable to make this type of credible threat.
The Delivery of Aid in the Recipient Countries
The distribution of aid at the operational level is also subject to a principal-agent relationship and to a moral hazard problem. The moral hazard problem exists because the benefit of the contract to the donor depends on the actions taken by the recipient, and the recipientmay not have incentives to behave properly after the donor has provided the funds.The recipient will take the action most beneficial to his own preferences. Therefore, the terms of the contractmust be modified to provide the recipient enoughmotivation to behave properly. Bureaucracies are even more important in recipients’ countries than in donors’ agencies. The civil servants of the public bureaucracies of developing countries have a long-term contract with a low wage. Public regulation makes it difficult to fire government workers even if their performance is not appropriate. One particularly important consequence of this lack of incentives in the provision of education and health in developing countries is the so-called missing doctors and teachers. For example, Banerjee and Duflo (2006) report that, in India, the absence rate for teachers is more than 24 percent and for health providers ismore than 40 percent. This is the usual shirking behavior predicted by the principal-agent theory.
In addition, since salaries forworkers in the public bureaucracies of developing countries are low, they tend to run privatemoneymaking activities or extract side payments from citizens (extort payment for free health services, demand money for a license, etc.). Corruption and rent-seeking behavior are often the result of severe moral hazard problems combined with the absence of clearly defined property rights. One important explanation for the lack of effect of foreign aid on growth and social development is the capture of aid by government officials and politicians. The measurement of the extent of corruption in the allocation of foreign aid in developing countries is difficult because accounting systems work poorly and local accountability is weak. The World Bank has used Public Expenditure Tracing Surveys (PETS) to calculate how much money is diverted from the time the central government receives the funds until they reach the final beneficiaries. The proportion of funds that reach the intended beneficiaries tends to be very low. For instance, Reinikka and Svensson (2004) analyze a large public educational program (a capitation grant to cover schools’ nonwage expenditures) in Uganda financed by the central government using district offices for its distribution. They find that many schools received nothing. On average, schools got only 13 percent of the government spending in the program. In Tanzania the analysis of the Primary Education Development Plan launched in 2002 by the World Bank and bilateral donors shows that, on average, only 20 percent of the funds disbursed at the central level finally reached the schools.
Rent-seeking activitiesmay even produce negative consequences in the recipient countries. Djankov, Montalvo, and Reynal-Querol (2006) provide empirical evidence that a suddenwindfall of resources in the form of foreign aid can damage the political institutions of the receiving country by reducing checks and balances in government and democratic rules. They label this effect the curse of aid. Other authors provide evidence that Somalia’s civil war was caused by the desire of different factions to control the large amount of food aid that the country received.
The Market for Aid
Some authors have argued that a possible solution to the agency problems in the aid industry is to increase competition among aid agencies (Klein and Harford 2005) and recipient countries (Svensson 2003) and use competition to make the market work better. Aid agencies should compete to fund the best projects, enabling donors to channel funds through the agencies that produce the best results; and developing countries should compete to receive official aid. This increase in competition, if accompanied with improvedmonitoring and benchmarking of aid agencies and proper evaluation of outcomes for the projects, may alleviate some of the incentive problems originated by the multiple layers of delegation in the delivery of foreign aid.
Notice that this does not contradict the need for coordination among the donors. Competition is good in the sense that recipient countries will choose the most efficient project, and therefore will end up with the best donor for them. So, the existence of many donors competing for recipients does not imply thatmany donors will be deliveringmoney to the same project, but only the most efficient one. See also aid, international