Svensson, Jakob (1999), “Aid, Growth and Democracy,” in Economics and Politics, Vol. 11, No. 3, pp. 275-97

Does aid work? Using the growth impact of aid as a yardstick, our results suggest it does, but only in certain environments. In particular we have found that the long-run growth impact of aid is conditional on the degree of political and civil liberties. Aid has a positive impact on growth in countries with an institutionalized check on government power; that is, in more democratic countries. The data suggest, however, if this is not the case, aid will be used to satisfy the government’s own non-productive demands. These results suggest that promoting democracy in the developing world may not only have a value in itself, it may also increase the long-run growth impact of foreign assistance. We also found that there are large differences across donors with respect to the allocation of foreign aid. While a few donors, conditional on other motives, have reward recipients with better than average records of civil and political liberties with more aid since the 1970s, this correlation is much weaker in the 1990s. This result points to the importance of creating incentives within the donor organizations to channel aid towards environments where it can be productive (see Svensson, 1997). A contribution of this paper is that we have taken a step towards identifying such an environment.

We have shown that foreign aid and windfalls are associated with higher corruption in countries more likely to suffer from powerful competing social groups. We believe this result is supportive of the theory. The model we have laid out is built around a standard rent-seeking specification. Admittedly, this is a black box approach to policy formation. It should be viewed as a reduced form of a more structural model in which organized social groups can capture a large share of government income, either by means of direct appropriation, or by manipulating the political system to implement favorable transfers, regulations and other redistributive policies.

In the empirical section we use corruption as a proxy of rent-seeking. We believe that corruption is likely to be highly correlated with other forms of discretionary redistribution, and therefore able to capture more than the empirical relationship between aid, corruption and the political equilibrium. This assertion also finds support in the data – the empirical results are robust to other proxies of wasteful rent-seeking.

The key insights we want to capture in the model are that “economically irrational” responses to windfalls that has been noted in the literature may be “politically rational”, and that foreign aid may affect the outcome (and the political equilibrium) through a less tangible mechanism.

These results rely on four general assumptions. First, economic policy is determined jointly by a number of powerful social groups. In the long run, the groups are better off if they cooperate than if the act noncooperatively. There is a large literature both in economics (see, e.g., Easterly and Levine, 1997; Rodrik, 1998) and in political science that links interest / social groups with redistributive policies in developing countries. Problems of coordination and cooperation are at the heart of this literature. Thus, we believe that our reduced form model captures an important aspect of reality.

Second, we assume that the deviating group can capture the entire government budget. An objection to this is that a slight increase in rent-seeking by one group when all others refrain yields a very large payoff. While technically correct, this critique takes the model’s structure too much at face value. Rent-seeking is a composite variable of both direct cost of redistribution (e.g., bribes), and indirect costs of political competition (e.g., resources employed to seize or attempt to seize power). In these dimensions, deviating from a cooperative code of conduct is likely to yield high short-run payoffs.

Third, the larger the government budget, the larger the incentives to deviate. An objection to this assumption is that it implies that the richer the economy, the more rent-seeking, and the type of discretionary redistribution analyzed in the model is not associated with policies in many rich developed countries. In response, it is important to make clear that the focus in the paper is on the relationship between rent-seeking, windfalls and foreign aid, and we have purposely assumed away other incentives to engage in rent-seeking.¹⁵ An intuitive way to think about the setup is that government income takes two forms, a constant flow and a stochastic flow. y(Ѳ_t) is the stochastic part, and there are pre-existing institutional arrangements determining the distribution of the constant flow. The constant part could vary between countries, implying that rich countries are not necessarily more prone to rent-seeking. The focus of this paper is the conflict arising when a country receives income above the level that its pre-existing institutional arrangements can handle, i.e. windfalls, and how expectations of foreign aid influence this response.

Finally, we assume that the donor (partly) cares about the recipient’s welfare. There is plenty of empirical support for this assertion.

The present model has abstracted from a number of issues influencing public policy in developing countries. The analysis may therefore be biased and it would be inappropriate to draw any definite conclusions. Nevertheless, some important insights emerge from the analysis. First, we have shown that the provision of public goods does not need to increase with government income, thus providing a political-economy rationale for why large windfall gains in revenue, or large inflows of foreign aid, do not necessarily result in general welfare gains. Second, we have shown that expectations of aid in the future may suffice to increase rent dissipation and reduce the expected level of public goods provision.

From a policy perspective, there are four main implications of these findings. First, the model points to the importance of studying the interaction between the political process shaping public policy and foreign aid. Second, concessional assistance may influence policy in the recipient country even without any resources actually being disbursed, implying that evaluations of project and sector assistance may overestimate the total impact of foreign aid. Third, the analysis stresses the important issue of commitment in foreign aid policy. If the donor community can enter into a binding policy commitment, aid may mitigate the incentives for social groups to engage in rent-seeking activities. However, such a regime shift would involve an aid policy that in the short run provides more assistance to countries in less need, and less assistance to those in most need. Enforcing such a regime shift may be difficult (Svensson, 1997). Finally, the fact that democracies seem to be less subjective to the perverse effect of aid on corruption suggests that political liberalization should have an important priority in the donors’ policy agenda.

We provide some empirical evidence supporting the mechanism we propose. Foreign aid and windfalls are associated with increased corruption in countries more likely to suffer from competing social groups. We find a weakly robust negative relationship between aid and corruption in countries where these conditions are less likely, while there is no evidence that the donors systematically allocate aid to countries with less corruption. These results are robust to a number of statistical problems.

The present model has abstracted from a number of issues influencing the game between the donor and the recipient. The analysis may therefore be biased and it would be inappropriate to draw definite conclusions, let alone to make final policy recommendations. Nevertheless, some important insights emerge from the analysis. First, it is shown that one reason for the poor aggregate record of foreign aid may be a moral hazard problem that adversely affects the aid recipients’ incentives to undertake structural reforms. In principle, conditionality could partly solve the problem, but this requires a strong commitment ability by the altruistic donor. Contrary to conventional wisdom in the aid literature, we show that without such a commitment technology, delegation of part of the aid budget to an international agency with less aversion to poverty as well as tied project aid may improve welfare for all parties.

Recent empirical evidence suggests that aid has a positive impact on growth under certain conditions, but that foreign aid has not been systematically channeled to countries where those conditions prevail. We argue that this finding is partly driven by a common feature in the donor agencies’ incentive system: the low opportunity costs of committed funds. As a result, there is a strong bias towards “always” disbursing aid to the ex ante designated recipient, irrespective of that recipient’s performance and (irrespective of) the conditions in other potential aid recipient countries. This assertion finds strong support in the data.

In this paper, we have analyzed a simple reform that may improve the outcome from the donor’s perspective. Instead of committing a fixed amount of aid to each recipient ex ante, and making aid conditional on reform or outcome, the donor would commit the aggregate amount to be given to a group of countries, but where the actual amount disbursed to each individual country would depend on relative performance. Explicitly linking the allocation and disbursement decisions has two important advantages as compared to the present practices. First, by creating a conflict of interest between the beneficiaries of foreign assistance, the opportunity cost of aid is internalized, thereby giving the donor (or country department) stronger incentives ex post to reward good policies. Second, competition among recipients allows the donor to make inferences about common shocks, which otherwise conceal the recipients’ choice of action. This enables the donor to give aid more efficiently. A recent World Bank report (Collier and Dollar, 1998) estimates that if aid is redirected towards poor countries with good policies, more than 80 million people could be lifted out of poverty for the same aggregate level of foreign aid. Consequently, there are potentially large gains of reforming current aid practices, and this paper has studied one such reform.

We believe that the argument for reforming program funding also applies to project aid. Thus, under certain conditions, it might be optimal to pool the budget for different (but similar) projects instead of having separate budgets for each individual project.

Four objections against linking the allocation and disbursement decisions are worth stressing. First, it could be argued that competition between recipients introduces uncertainty about financial flows, which renders planning more difficult and makes fiscal spending too volatile. This may be true if making comparisons with how the aid system presently seems to work; i.e., commitments are always disbursed. However, this is not true if we compare it with the conditionality outcome as it is supposed to work. In fact, if the shocks facing the recipients are (highly) correlated, the uncertainties will be reduced by having the recipients compete in an aid tournament.

Second, it could be argued that the degree of reform implementation depends on domestic political economy forces, rather than on conditional aid. In fact, recent evidence suggests this to be the case (Burnside and Dollar, 2000; Dollar and Svensson, 1998). However, these studies analyze the impact of conditional aid (as it seems to work), not as it was meant to work. Therefore, as stressed in the paper, we should not expect any significant correlation between aid flows and policy reform. More important, the model primarily concerns the incentive structure within the donor organization. Even if the degree of policy reform is solely determined by domestic political economy forces; i.e., is independent of foreign assistance, linking the allocation and disbursement decisions will still be useful since this provides incentives for the donor to allocate/disburse aid to where it can be effective.

Third, the time-inconsistency problem analyzed above could be dampened in a dynamic setting. If the donor–recipient game is repeated an indefinite number of times, this might provide incentives for the manager not to pay out all funds if he observes a negative signal, in order to build a reputation and give the proper incentives for countries to undertake reform. But, typically managers in aid agencies regularly switch positions (to other country desks). Thus, building a reputation would require strong internal control and coordination over time in the donor agency, a presumption the data does not support. In fact, the disbursement pattern does not change over time.

Finally, collusion among recipients undermines the equilibria described in Section 4. An important assumption is thus that the recipients act non-cooperatively. We believe this to be an accurate starting point for analyzing problems in the current aid system. Collusion may be a more important issue when it comes to linking projects within a country, however. This is an important topic for future research.

A question partly left unanswered is, why is it that if the linking of the allocation/disbursement decisions improves outcome, the donor community does not explicitly link these decisions? One answer is of course that the potential cost of tournament type aid schemes is perceived as being very high (for example the cost related to the political risk of creating competition between countries). However, the extent of competition between countries, and thus the potential cost, can be controlled by varying the share of aid disbursed through a tournament-type aid scheme. This also seems like a less important concern for project aid. A more plausible explanation is related to the change in the existing power structure within the donor agency/donor community implied by such a regime shift. In essence, the reform would reduce the discretionary power of many managers mainly in charge of the disbursement decisions. Moreover, by making the opportunity cost explicit in the decision process, the management would be required to make “tougher” choices. Recipient-specific interest groups (e.g., domestic firms, NGOs), and potentially the recipient government, may also oppose an institutional change that would imply aid flows conditional on performance, rather than ex post unconditional disbursements.

Taffet, Jeffrey (2007), Foreign Aid as Foreign Policy: The Alliance for Progress in Latin America, New York and London: Routledge Taylor & Francis Group.

Conclusion: Aid to Latin America in Context (pp. 195-197)

At the start of the Kennedy administration, the Alliance for Progress was the great hope of U.S. policymakers as the means to counter the rise of Communism in Latin America. Rather than use force or coercion to influence political change, the program would encourage leaders to pursue reform. Aid from the United States would be a way to help Latin American leaders help themselves create lasting economic growth and stability, and in doing so develop new, more cooperative and positive inter-American relationships. As Richard Nixon had found in 1958, and as Castro’s success illustrated, many in the region resented U.S. power and arrogance, but Kennedy hoped to change all this and usher in a new age of collaboration and mutual respect.

The Alliance for Progress did not achieve these goals. In an effort to establish the program, the United States instructed Latin Americans on how to pursue reform. As the case studies demonstrate, they were rewarded when they cooperated and punished when they did not. The program therefore did not represent a partnership, but reinforced Latin American ideas about the overbearing United States. In developing the program, policymakers in the United States did not imagine, comprehensively or rationally, exactly how it would work; the mechanisms they developed were unwieldy, impractical, inefficient, and ignored. While there was an initial desire to allow Latin Americans an important role in determining how money was spent, blunt political considerations made that impossible. This meant that the Alliance for Progress devolved into a U.S. foreign aid program that retained little of its dramatic revolutionary content. It became simply a way to help friends, hurt enemies, and promote a set of theories about how to best create economic stability. It was not an alliance, and it was not even always about economic progress.

Initially, Kennedy stressed that the Alliance for Progress would reject the notion of imperial hubris – the idea that the United States, as the strong and wealthy power, and with great ability to influence change, knew what was best for Latin America. Kennedy wanted Latin Americans to believe that the United States wanted to help because it was the right thing to do and that he cared about their problems. In terms of rhetoric, Kennedy was successful. Latin Americans saw him as a great leader, dedicated to progressive change. The reality is that, however unfortunately, his policies and those of his successors did not live up to the initial ideals. Faced with political and economic instability and the threat of anti-American politicians taking power, U.S. policymakers felt they had little choice but to use the tools available, including the Alliance for Progress, to create conditions favorable to their own interests. To expect them to have acted otherwise ignores the overwhelming and larger context of the Cold War, which informed their understandings of the world and the reality of their power relative to Latin Americans.

The application of foreign programs as a way to manipulate foreign nations was at one level necessary and obvious, but on a second level clearly counter to the best traditions of American democracy. In a remarkable address at the University of Denver in August 1966, Johnson explained that “the overriding rule” for U.S. foreign policy was that it “must always be an extension of … domestic policy.” He argued, “our safest guide to what we do abroad is always to take a good look at what we are doing at home.” One application of this “rule” in Johnson’s mind was that “in the United States we do not like being told what to do. We like even less being told what to think.” Thus, in international relations, “The United States has no mandate to interfere wherever government falls short of our specifications.” Unfortunately, these dictums did not guide policy. Rather, U.S. policy was the exact opposite of what Johnson professed.¹

Had the policy actually worked effectively it would be possible to make an argument that interference was a good idea, but U.S. political successes were few. In Brazil, U.S. foreign aid programs had minimal effect on the political orientation of the Goulart government, and were unable to influence the military government in important ways. In Chile, the massive aid in the pre-1964 era had little impact on Frei’s victory, and thereafter served mostly as an irritant that drove Chilean leaders toward anti-American positions. Most dramatically, nine and one-half years after Kennedy announced his ten-year commitment to Latin America, Salvador Allende won his country’s presidential election. In the Dominican Republic the ledger may be significantly more positive. It is unlikely that U.S. aid programs had much impact on political changes during the first half of the decade, but they did create stability in the postintervention era. Finally, in Colombia, aid may have had a marginal effect in strengthening the governments of the National Front; it allowed them to avoid tough decisions, and to spend, essentially, beyond their means.

Understanding the political logic of the Alliance for Progress helps in developing a perspective about U.S.-Latin American relations during the 1960s. Throughout the decade, U.S. policymakers continually looked to the program as a catchall solution to the problems they faced. This should not suggest that the Alliance for Progress was central to every piece of the relationships, only that it was a way of manipulating them to U.S. advantage. With the notable exception of policy toward Cuba, decisions about economic aid programs were part of every U.S. action in the region. The U.S. government was willing to use the CIA to engage in covert actions, to use Marine to invade the Dominican Republic, to encourage military leaders to overthrow civilians, and to use its considerable power to help U.S. businesses. But in each one of these cases there was a connection to economic aid in an attempt to create stability, reward friends, or keep threats from emerging.

In broader terms, the Alliance for Progress also helps clarify how policy toward Latin America in the 1960s was consistent with earlier and later periods in inter-American relations. There is little debate in the scholarly community that the major theme in the history of U.S.-Latin American relations is the U.S. desire and ability to dominate the region.² This interest came from aspiration about increasing U.S. power, economic and otherwise, and it was usually justified by assumptions about Latin American cultural inferiority. The Alliance for Progress demonstrates that U.S. policy in the 1960s was, though different rhetorically from other eras, essentially similar in application. Kennedy talked about the Alliance for Progress as a partnership of equals and suggested that his goals were more moral and cooperative than his predecessors. The reality was more complicated. Though U.S. policymakers did hope to implement Kennedy’s vision, the discrepancy in economic power, U.S. global interests, and assumptions about U.S. superiority meant that U.S. policy was not a repudiation of the past, but a continuation of it.

“It’s easy to forget that, when this war began, we were united, bound together by the fresh memory of a horrific attack and by the determination to defend our homeland and the values we hold dear. I refuse to accept the notion that we cannot summon that unity again. I believe with every fiber of my being that we, as Americans, can still come together behind a common purpose, for our values are not simply words written into parchment. They are a creed that calls us together and that has carried us through the darkest of storms as one nation, as one people.”

—President Barack Obama, West Point, New York, December 2, 2009

—

This strategy calls for a comprehensive range of national actions, and a broad conception of what con­stitutes our national security. Above all, it is about renewing our leadership by calling upon what is best about America—our innovation and capacity; our openness and moral imagination.

To effectively craft and implement a sustainable, results-oriented national security strategy, there must be effective cooperation between the branches of government. This Administration believes that we are strong when we act in line with our laws, as the Constitution itself demands. This Administration is also committed to active consultation with Congress, and welcomes robust and effective oversight of its national security policies. We welcome Congress as a full partner in forging durable solutions to tough challenges, looking beyond the headlines to take a long view of America’s interests. And we encour­age Congress to pursue oversight in line with the reforms that have been enacted through legislation, particularly in the years since 9/11.

The executive branch must do its part by developing integrated plans and approaches that leverage the capabilities across its departments and agencies to deal with the issues we confront. Collaboration across the government—and with our partners at the state, local, and tribal levels of government, in industry, and abroad—must guide our actions.

This kind of effective cooperation will depend upon broad and bipartisan cooperation. Throughout the Cold War, even as there were intense disagreements about certain courses of action, there remained a belief that America’s political leaders shared common goals, even if they differed about how to reach them. In today’s political environment, due to the actions of both parties that sense of common purpose is at times lacking in our national security dialogue. This division places the United States at a strategic disadvantage. It sets back our ability to deal with difficult challenges and injects a sense of anxiety and polarization into our politics that can affect our policies and our posture around the world. It must be replaced by a renewed sense of civility and a commitment to embrace our common purpose as Americans.

Americans are by nature a confident and optimistic people. We would not have achieved our position of leadership in the world without the extraordinary strength of our founding documents and the capability and courage of generations of Americans who gave life to those values—through their service, through their sacrifices, through their aspirations, and through their pursuit of a more perfect union. We see those same qualities today, particularly in our young men and women in uniform who have served tour after tour of duty to defend our nation in harm’s way, and their civilian counterparts.

This responsibility cannot be theirs alone. And there is no question that we, as a nation, can meet our responsibility as Americans once more. Even in a world of enormous challenges, no threat is bigger than the American peoples’ capacity to meet it, and no opportunity exceeds our reach. We continue to draw strength from those founding documents that established the creed that binds us together. We, too, can demonstrate the capability and courage to pursue a more perfect union and—in doing so—renew American leadership in the world.