The Roles of Aid in Politics Putting China in Perspective

Foreign development aid since the end of World War II has increased in significance within international relations as a way of developing strategic interests and improving the international environment. Foreign aid is an important policy tool for expanding nations’ influence upon international affairs and maintaining local and international economic stability, and has played a special role within China’s foreign relations. Aid from China has grown in the recent years along with its economic development, and the corresponding change in the objectives, functions and manner of Chinese aid have given the country a higher standing within the sphere of worldwide development aid. Research into development aid policies and practices in developed countries is of help to China in establishing a more mature foreign development aid strategy. […]

The proportion of a country’s gross national income expended on foreign development assistance is subject to different factors of influence. On the basis of data from the 21-member Development Assistance Committee from 1990 to 2005, we used a quantitative model to estimate the effects of different factors that determine the level of a country’s relative ODA. This analysis shows direct relationships among ODA as a proportion of GNI and a country’s economic position, public social expenditure as a proportion of GDP and level of economic openness, and an inverse relationship between ODA and the level of GDP per capita and fiscal revenue. This means that a country’s expenditure on foreign assistance increases as it becomes more prominent within the global economy, and as the country’s electorate becomes more concerned with the objectives of welfare policies. But it means at the same time that wealthy countries are not necessarily more predisposed towards philanthropy than poor countries, and that countries with large fiscal revenues are not likely to provide more assistance than countries with limited fiscal income. An extent of empirical support for the explanations of both the realist theory of international politics and those based on domestic political factors is thus found. This implies that principles guiding foreign development assistance policies since the Cold War have become pluralized.

As regards interpreting the results of the model: as GDP per capita increases $1000, ODA as a proportion of GNI decreases 0.0038 percentage points, and as a country’s GDP as a proportion of world GDP increases 1 percentage point, ODA as a proportion of GNI increases 0.096 percentage points. As fiscal revenue as a proportion of GDP increases 1 percentage point, ODA as a proportion of GNI decreases 0.013 percentage points; as expenditures on public welfare as a proportion of GDP increase 1 percentage point, ODA as a proportion of GNI increases 0.011 percentage points; as economic openness increases 1 percentage point, ODA as a proportion of GNI increases 0.0045 percentage points.

Implications for China

Foreign development assistance is a main component of China’s strategy for sustainable, peaceful development and a tool that can be used in its economic diplomacy. China must therefore determine an appropriate set of principles and procedures for administering development assistance. The experience of developed countries, especially the DAC Member States of the OECD, is valuable in this respect.

It follows from our preliminary analysis of development assistance provided by the DAC states that the ideal level of development assistance any given state provides is determined on the basis of specific domestic conditions. China, as a rapidly developing power whose economy accounts for an ever-larger portion of the overall world economy, and which is relatively open, has a rising international responsibility. Its expenditure on foreign development assistance, therefore, should increase accordingly. Proposed objectives such as a harmonious society and a harmonious world should also bring about simultaneous increases in domestic public social expenditure and overseas development assistance. Such is the case because the precondition for a country’s concern about the welfare of people in other nations is that of its concern for the welfare of its own citizens. As the Chinese people become more concerned about welfare objectives, China’s political leaders should correspondingly become more concerned about similar matters within the wider international arena.

China’s assistance to other countries in recent years has risen steadily. As of the end of 2006, the country had provided assistance to more than 100 countries, regions and organizations, and assisted aid recipients in approximately 2000 different projects. China has provided more than 100 development loans on preferential terms to more than 60 countries, and large-scale assistance in the form of goods and food to more than 110 countries. It also provided technical and managerial training to more than 23000 people in over 100 countries. The Chinese government has moreover signed debt forgiveness agreements with 46 countries in Asia, Africa, Latin America and the South Pacific, thereby erasing a portion of what they owe China.²⁶ But taking into consideration China’s level of openness to the rest of the world and its rate of economic expansion, the international environment might degrade during the process of China’s economic rise. This being the case, it is imperative that China increase its level of foreign development aid.

First, China should consider streamlining its foreign assistance. At present, the forms of assistance China provides include: donations, interest-free loans, development loans at preferential interest rates, preferential export credit to purchasers of its goods and funds for the support of joint ventures. As many types of aid are provided, its administration is disorganized and consequently difficult. Generally speaking, China lacks a systematic programme with respect to foreign development assistance. This article recommends that the various channels of foreign development assistance be streamlined, that the former classification scheme be dropped, and that a new scheme comprising two types of aid: monetary aid and assistance loans, be implemented, preferential rates for the latter computed according to international standards.

Second, foreign development aid objectives should be set at each stage of China’s economic development. As China continues to open up to the world and its economic position to improve, it may be anticipated that its ODA will increase. The government, therefore, may make specific plans to increase ODA in proportion to its growing gross national income.

Third, as development assistance is influenced by fiscal income as a proportion of GDP and public social expenditure as a proportion of GDP, development aid should not be blindly expanded; increases should be conditional upon national fiscal revenue and public social expenditure.

Headey, Derek (2005), Foreign Aid and Foreign Policy: How Donors Undermine the Effectiveness of Overseas Development Assistance, CEPA Working Paper Series No. 05/2005, Australia: School of Economics, University of Queensland.

This paper has presented evidence which supports the hypothesis that donors’ political motivations - and not just recipient-side conditions such as good macroeconomic policies - also have a significant and plausible impact on the effectiveness of foreign aid. We first showed that previous studies have generally not used appropriate econometric methods, the two key elements of which are an improved measure of foreign aid which roughly gauges “aid intended to promote growth”, and a dynamic growth specification which simultaneously renders aid exogenous (without relying on untested assumptions regarding the equal effectiveness of strategic and non-strategic aid) and identifies a plausible dynamic relationship between the timing of aid inflows and the eventual returns to those flows. We then considered one of the easiest and, in policy terms, the most important means by which one can decompose aggregate aid flows into strategic and non-strategic (or less strategic) influences, the multilateral-bilateral division. As expected, multilateral aid flows are not well explained by strategic aid variables, whereas bilateral flows are, such that the latter constitutes a reasonable proxy for aid flows that are relatively strategic in orientation. Testing multilateral and bilateral flows produced the predicted result that the multilateral aid produces growth effects roughly twice the size of bilateral flows. Moreover, the construction of motivational indices and their interaction with both aggregate and bilateral aid flows also suggests that a great deal of the variation in aid outcomes is explained by political strategic factors. In fact, these results show that bilateral aid can be quite effective when it is not politically motivated, while multilateral aid tends to be more effective on average.

In this final section we address some possible objections to these findings. The most important of these is that the cross-country growth regression is a highly flawed means of testing the effects of a wide range of potential determinants of growth. Testing the effects of foreign aid on growth is arguably an especially hazardous exercise: aid is likely to be endogenous, it is difficult to measure accurately, and, most importantly of all, it presumably affects growth through several quite distinctive transmission mechanisms (Hudson and Mosley 2001; Gomanee, Girma et al. 2002): investment and consumption levels, macroeconomic outcomes and policies, and institutions, to name the most important. We are entirely in agreement with assessment. In our view, growth regressions cannot be used to make statements such as “aid has been a success”. Their strength, of course, is their ability to control for so many other determinants of growth so that researchers can make statements such as “controlling for initial incomes, policies, geography, and institutions, aid has had a positive effect on growth when it has not been used as a mere instrument of geopolitics”. But this strength is also a weakness: another way to interpret these results is to conclude that “aid has increased growth net of its effects on other determinants of growth”, especially policies and institutions. It is very difficult to look around the world and find countries which have experienced aid-induced “miracles”, and likewise easy to find aid dependent “disasters”.

But whether this is because aid does indeed worsen policies and/or institutions, however, is by no means clear. Whether or not aid improves or deteriorates governance, for example, has been shown to depend on which instruments one uses (Harms and Lutz 2004). Moreover, when we update Knack’s (2001) regressions of governance change against aid, we find that lagging aid reverses Knack’s result (see Appendix B). With regard to aid’s effects on policies, there is further ambiguity. While there is little doubt that conditionalities have often failed and, even when they have worked, have worked extremely sluggishly, the last twenty years also appears to have been a period of unprecedented reform. It therefore seems difficult to reconcile statements by Dollar and colleagues (Burnside and Dollar 2000; Collier and Dollar 2002) that “Aid has not systematically affected . . . policies during the 1970-93 period” with statements by Rodrik (2003) that “countries such as Mexico, Argentina, Brazil, Colombia, Bolivia, and Peru did more liberalization, deregulation and privatization in the course of a few years than East Asian countries have done in four decades”. It seems hard to believe that donors – the Washington institutions in particular - did not play any role in influencing these regimes. Moreover, several of the most stubborn African reformers – Ethiopia, Madagascar, Malawi, Kenya, Tanzania – were finally relenting to World Bank pressure at almost precisely at the time of Burnside and Dollar’s initial publications.

Our view is that whether one regards aid as a success or a failure very much depends on what one expects aid to achieve. If institutions are primal, is it reasonable to expect aid to be an engine of growth? Probably not. If institutions are rigid and bad institutions are especially difficult to break down under the best of circumstances, is it reasonably to expect aid and aid donors to substantially improve them? At best, only very slowly. Aid operates in a Second (or Third or Fourth) Best universe. And while the effectiveness of aid flows can doubtlessly be improved along a number of dimensions, it is unrealistic to expect aid to easily produce the deep institutional changes necessary for growth. Foreign aid is essentially a proximate determinant of growth, but the evidence here suggests that, measured against a more realistic institutionalist yardstick, it is a reasonably effective one, especially when its motivations are, in a developmental sense, relatively pure. Moreover, donors can make aid substantially more effective simply by reducing the strategic biases of their aid allocations.

Headey, Derek (2008), “Geopolitics and the Effect of Foreign Aid on Economic Growth: 1970-2001,” in Journal of International Development, Vol. 20, No. 2, pp. 161-180.

While improving the techniques used to test aid effectiveness generally shows that aggregate aid is more effective than was previously thought (by some, at least), these improvements also indicate that multilateral aid has typically been more effective than bilateral aid in promoting growth. Moreover, a potent explanation of bilateral aid’s ineffectiveness over the last 3 decades is an apparent Cold War effect. In this section I address some possible objections to these conclusions, before touching on the methodological and policy-related implications of the results.

In draft versions of this paper, some readers were sceptical of the idea that multilateral aid is really less strategic than bilateral aid, pointing, for example, to a plethora of recent results which show that US national interests seem to influence the allocation of IMF funds, World Bank aid and in one paper, even Asian Development Bank aid.³⁰ The typical approach in these studies is to first explain the aid allocations of the IMF, The World Bank or the ADB in terms of purely developmental variables, and then test whether indicators of US strategic motivations are significant, ceteris paribus. But whilst the general conclusion that such strategic variables are significant does indeed suggest that multilateral aid is not truly apolitical, the vast majority of World Bank aid flows are explained by developmental factors (the IMF is a relatively small aid donor, accounting for just 2 per cent of total aid flows). In fact, the bilateral aid allocations which share the closest similarity to World Bank aid are those of the ‘good’ but small donors, such as Denmark and the Netherlands.³¹ So what these studies report are ‘t-value results’ (rather than ‘R-squared results’)³² which do not materially influence the conclusion that World Bank aid as a whole is much less geopolitically motivated than the aid allocations of the larger bilateral donors.

On a more methodological front, this study has raised some interesting issues in the empirics used in aid effectiveness studies. In particular, I conjectured that improving the techniques in aid-growth regressions can significantly alter results, and Section 2 provided some evidence that this is the case. Easterly et al. (2003) have already revealed that the significant and positive coefficient on Burnside and Dollar’s aid-policy interaction term disappears with the updating of Burnside and Dollar’s sample. But in a sense my results are more powerful, since I show that even if Burnside and Dollar had only changed their specification in one dimension (lagging aid) then they ought to have concluded that aid does typically have large unconditional marginal effects on growth. In addition, a recent IMF working paper (Rajan and Subramanian, 2005) uses purely geopolitical factors to instrument for foreign aid. Since the results reported here suggest that geopolitical aid may be less effective than developmental aid, Rajan and Subramanian’s estimate of geopolitical aid is likely to give a more pessimistic appraisal of aid effectiveness relative to other approaches in the literature.33 See Murray (2006) for a discussion of instrumentation problems associated with heterogeneity in second stage effects.

Another contributing factor to the more pessimistic results of earlier studies—such as Burnside and Dollar and Boone (1996)—is that their timeframes almost entirely encapsulated the Cold War era, when bilateral aid appears to have been ineffective. Essentially, then, we have the ironic result that these writers were very vocally concluding that aid was ineffective at precisely the time when the main culprit in this conclusion— bilateral aid—was just starting to have a sizeable positive impact on growth rates.³⁴

Nevertheless, the results reported in this paper are not in entire disagreement with the main thrust of Burnside and Dollar’s work, which also emphasises geopolitical biases as the principal reason for the inefficient allocation of bilateral aid. The argument in this paper is somewhat broader in scope, and not necessarily confined to the allocative inefficiencies associated with geopolitics, but both lines of reasoning emphasise the importance of the aid allocation literature.³⁵ By at least partially elucidating the geopolitical motivations of donors, this literature can draw attention to the efficiency implications of politically biased aid allocations as well as their ethical implications. Donors should not deceive themselves with the assumption that aid flows intended to purchase geopolitical objectives will also achieve substantial developmental outcomes, mutatis mutandis. The evidence presented here suggests that assumptions such as these have been very costly in the past.

This paper has sought to sketch out a preliminary assessment of Aid for Trade, placing a particular emphasis on the special circumstances on small states. To the extent that Aid for Trade is concerned with compensating preference-dependent countries for the loss of income due to preference erosion, it has argued that small states are particularly worthy of attention due to their high dependence on international trade in general and preferential trade in particular. On the other hand, estimates of the welfare losses associated with preference erosion remain disputed – as is the question of the overall importance of compensating preference-dependent countries vis-à-vis other development priorities in the DDA. In an important sense, much may depend on whether Aid for Trade in respect of the DDA is conceived of in instrumental terms – that is, an attempt to ‘buy’ support for ambitious MFN tariffs cuts – or more as a genuine attempt to integrate the world’s poorest countries into the world economy. Ironically, it may actually serve the interests of small states better if the former is the case, since the fear of the loss of preference margins is seen as one of the main obstacles in the way of securing a meaningful round of tariff reductions (Francois et. al 2005). The other unknown factors which feed into the above centre on how much money Aid for Trade will ultimately be worth and what the overall significance of its inclusion inside of the DDA will turn out to be. From this perspective, we might conclude that, at this stage at least, that the overall significance of Aid for Trade is as uncertain as the outcome of the Doha Round as a whole.

Historically, foreign aid has served a multitude of objectives. For some of the smaller donors, the allocation and quality of aid flows have been largely, but not wholly, shaped by a concern for the development needs of the recipient community. By contrast, the foreign aid of several of the large donors has been firmly established as a foreign and commercial policy tool, designed to achieve a range of political, strategic, economic, but also genuinely humanitarian objectives. This assertion is born by observing the historical origins of foreign aid and is supported by the empirical literature on aid allocation. Indeed, it seems that donor self-interest has been an enduring feature of donor-recipient relations since the 1950s, a relationship that has otherwise been characterized by many changes in terms of volume, composition, types and objectives of aid flows. Yet, while this particular characteristic of aid flows may well have impaired the effectiveness of aid, by nature, there is no automatic contradiction between donor and recipient objectives.

Perhaps the most important of recent changes in the aid picture is the clear reversal after 1992 of the historic upward trend in aid volumes. This may not be a big problem when declining aid flows are compensated by higher private flows, as has happened in several developing countries. Yet it may be a considerable problem in low-income countries without access to private capital and which continue to rely heavily on aid for financial resources. Clearly, the underlying premises of donor-recipient co-operation are very different when aid resources become more limited rather than more abundant, especially when debt service is still a factor of significance. Indeed, this is one of the rationales behind writing this book, and many of the themes touched upon in this chapter will be pursued in later chapters.

Hoadley, J. Stephen, (1980), “Small States as Aid Donors,” in International Organization, Vol. 34, No. 1, pp. 121-147.

This brief deployment of data around propositions on small state donors’ foreign aid behavior lends credence to the presumption that small states do indeed behave differently from large states. The data confirm that small states give aid to a narrower range of recipients but give aid more generously; that small states channel more of their aid through multilateral agencies and give more of their bilateral aid to the poorest countries; and that small states conform more closely to international targets on aid volume and ease of terms. There is also evidence to suggest that small states avoid giving aid to the enemies of their large allies, with many exceptions, and that small states tend to give aid in a manner less directly self-interested, politically or economically, and more oriented towards recipient need. And it is confirmed that the decision to place Canada in the category of small state donors was a correct one, for Canada is found to fit the small donor aid performance averages far better than the large donor averages.

We end with a note of caution: size itself does not determine a state’s behavior directly; it is merely a quality which must be taken into account by decision-makers along with many other qualities and conditions of the international environment. There is no attempt in this essay to elevate size to the status of a causal variable, for to do so would be misplaced concreteness. Rather, the attempt here has been to demonstrate that size, measured by population, stands as a useful surrogate for a number of qualities of a state which structure the perceptions and policies of international actors in consistent ways and that it is especially useful in quantitative and comparative studies of patterns of policy outcomes such as aid allocations.