1. A New Framework for Development Assistance

The existing framework for development cooperation between rich countries and poor has centred on a collection of programmes, bilateral and multilateral, loosely classified as "foreign aid". These programmes include grants, long term loans at subsidized rates of interest, short term export credits, distributions of surplus commodities, military assistance, technical assistance and much else. The ostensible purpose of these programmes was to promote the economic development of poor countries in the expectation that global poverty would thereby be eliminated. In practice however motives were mixed, the donors pursuing diplomatic, military and commercial interests as well as humanitarian ones.

The outcomes also were mixed. Indeed foreign aid has been a great disappointment to those who hoped that the end of the colonial era would usher in a new age of solidarity and cooperation between North and South. Looking back over four and a half decades of international economic assistance to the developing countries, it is clear that expectations of rapid and dramatic progress were too high, the economic analysis was faulty and the political assumptions were simplistic. In short, the supporters of foreign aid were embarrassingly naive. The end of the Cold War, however, provides an opportunity to think afresh and to build a new framework for development cooperation. But if the new framework is to be more solid than the old one, a clear understanding of what went wrong is essential.

The motives for foreign aid--as is true of most human activities--were complex and multi-faceted. Idealism, generosity, solidarity with the disadvantaged were combined with political expediency, commercial self-interest and ideological conflict between the West and the former USSR. It is hardly surprising that foreign assistance programmes resting on such motives often failed to be of much assistance to the recipient countries. Economists of course were able to produce reasoned justifications for the establishment, indefinite continuation and enlargement of bilateral and multilateral aid programmes, but the actual economic performance of the intended beneficiaries, viewed as a whole, often left much to be desired.

It is becoming widely recognized that the economic rationale for foreign aid was built on false premisses, namely, that underdevelopment is perpetuated by a lack of savings and access to Western technology.1  In fact savings rates in developing countries often are higher than in developed ones and despite the existence of patent and copyright protection, restricted access to technology in an age of instant communications cannot possibly be the major barrier to economic and human development. As regards savings, for example, in 1990 gross domestic savings were 28 per cent of GDP in the World Bank's "low-income economies", 24 per cent in the "middle-income economies" and 22 per cent in the "high-income economies"2  That is, savings rates were inversely correlated with per capita income, the opposite of a central premiss of aid programmes.

While there still is much debate about the economic consequences of foreign aid, four stylized facts about aid gradually are coming to be accepted. First, it is evident that foreign aid, contrary to original expectations, has not contributed to a noticeable acceleration of the rate of growth of developing countries. Some have argued that on occasion aid has actually retarded growth, but even the most ardent advocates of foreign aid have been hard pressed to show that conventional foreign assistance programmes have systematically achieved their primary objective of increasing the pace of development. In the case of Africa, by far the poorest continent, growth has ceased even to be an objective of foreign aid. Instead the objective is said to be "structural adjustment". Yet "structural adjustment has neither restored growth nor eased poverty in Africa." 3  Even the World Bank has been forced to admit the obvious and issued a paper entitled "Why Structural Adjustment Has Not Succeeded in Sub-Saharan Africa" in which it was argued that Africa hardly benefited from World Bank programmes. This report "was withdrawn, and reissued under a blander title."4

Second, where aid inflows are large in relation to the recipient's national product, relative prices are distorted in an anti-development direction. Large inflows of foreign aid tend to result in an appreciation of the exchange rate, thereby discouraging production for export or production intended to replace goods produced abroad. Large inflows of aid also tend to reduce real rates of interest in the recipient country, thereby discouraging savings, encouraging those who hold financial assets to place their holdings abroad and creating incentives for local investors to adopt techniques of production which are biased against the employment of labour. When aid takes the form of subsidized exports of commodities, the change in relative commodity prices can be devastating for local producers. The European community, for instance, subsidizes the production of beef in Europe. Some of the beef is then dumped in West Africa, namely in the Cote d' Ivoire and Ghana, where it is sold at half the price of beef produced in Mali, Burkina Faso, Niger and Chad that once was exported to Cote d' Ivoire and Ghana.5  That is, agricultural subsidies in Europe, disguised as foreign aid, are used to destroy the livelihood of some of the poorest people on earth.

Third, the availability of foreign aid has made it easier for the governments of recipient countries to increase unproductive current expenditure, to expand the military and to reduce taxation. Inflows of foreign aid, far from resulting in a rise in productive investment of equal magnitude, as once postulated, have in practice been used to finance higher expenditure on a wide range of activities, some of which contribute to development and some of which do not. Aid has been highly "fungible" and this has made it possible for those governments that wished to do so to divert aid from its intended purposes to purposes desired by the recipient government. The purposes of recipient governments, as revealed by their actual expenditure programmes, often had little to do with promoting development and eliminating poverty. Fungibility is an inescapable feature of all aid programmes and makes it difficult to tie foreign assistance to particular objectives, policies or expenditure programmes in recipient countries.

Indeed, fourth, there is no evidence apart from the occasional anecdote to suggest that either bilateral or multilateral aid programmes have succeeded in reaching the poor. On the contrary, most of the available evidence indicates that most of the benefits of foreign assistance programmes are captured by middle and upper income groups, i.e., by the élite. In some cases it is the international élite rather than the national one that captures the lion's share of the benefits. It has recently come to light, for example, that half of the expenditure of the European Bank for Reconstruction and Development was allocated to itself and the residual half to the countries the Bank was created to assist. The high salaries and perks of international aid agencies in general have in fact become a source of political outrage. Quite apart from this, however, even when aid-financed projects are aimed directly at the poor, they often miss their target. In the majority of cases the projects are not even aimed at the poor or at contributing to human development. That is to say, the impact of aid on reducing poverty and inequality in developing countries and on raising the level of human development has been negligible. This, again, is an inescapable fact of any aid programme.
 

Multilateral donors

These facts are damning and raise the question whether foreign aid as presently organized deserves to have a future. Perhaps the developing countries, and particularly the poor in developing countries, would be better off if the entire foreign aid enterprise were declared intellectually bankrupt and shut down. Before we call in the receiver, however, we should pause briefly to consider whether the criticisms that have been made apply equally to bilateral and multilateral aid programmes.

It could be argued that the muddle of motives and the consequent confusion of outcomes is more characteristic of bilateral aid programmes than of multilateral ones. Insofar as foreign aid programmes originated out of the ideological confrontation between the United States and the former Soviet Union, one might assume that this affected the character of the bilateral programmes of the United States and its allies, leaving the multilateral programmes largely untainted. The World Bank, the many regional development banks, the United Nations Development Programme and the specialized agencies of the United Nations, according to this view, were relatively free to pursue "pure" development objectives and were less constrained by political, diplomatic, commercial or ideological considerations.

This is certainly a plausible view, but unfortunately it probably is not a correct view. Let us consider the World Bank, the largest multilateral institution providing investment funds to developing countries.6  First, World Bank lending has been biased against the countries where the largest number of poor people live and this bias has been extraordinarily strong. Consider the ten countries with the largest number of poor people--India, China, Bangladesh, Indonesia, Pakistan, the Philippines, Brazil, Ethiopia, Myanmar and Thailand. These ten countries account for 72.4 per cent of the world's poor, but during the period 1989-91, World Bank lending commitments to these countries accounted for only 39.5 per cent of the total. The bilateral donors have an even worse record--only 26.8 per cent of their assistance was allocated to these ten countries--but this merely demonstrates that neither bilateral nor multilateral aid agencies have given highest priority in their assistance policies to reducing global poverty. There is of course a difference between poor countries and poor people, and some large but less poor countries none the less have large numbers of poor people. The view we shall take below, however, is that international aid should be directed to poor countries and that it is the responsibility of governments (whether in poor or less poor countries) to ensure that the needs of poor people are met.

Second, the World Bank has shown little interest in the overall allocation of resources in the countries it assists. No serious attempt has been made to combat "fungibility" or the use countries make of the totality of resources available to them. There is no evidence, for example, that the Bank discriminates against countries which devote an exceptionally high proportion of their resources to military expenditure. Despite its frequently announced concern for good management, and its willingness to tie programme lending to specific policies, the Bank appears to be indifferent as to whether a country squanders its resources. It is possible that because the World Bank concentrates on the microeconomics of projects, it has chosen to ignore the macroeconomics of resource use and to act as if there were no problem of fungibility, i.e., of aid being used indirectly to finance military expenditure or government bureaucracies or unproductive consumption expenditure. The Bank has persistently ducked this issue by claiming that ex post evaluations indicate that its projects, by and large, show pretty good rates of return. It has never satisfactorily answered the question why, if aid has been moderately generous and projects have been well selected, there is not a convincing positive relationship between aid and accelerated development.

Third, within individual countries, World Bank lending has been biased against poor people. That is, Bank lending intended to tackle poverty head on has been disproportionately low in the sense that the proportion of "poverty loans" in the Bank's country portfolio has usually been smaller than the proportion of people living in poverty. Included in the term "poverty loans" are Bank financed investments in basic education, primary health care, nutrition programmes, safe drinking water, sanitation and family planning. The importance of expenditure in these areas both for reducing poverty directly and for promoting long term growth has been underlined repeatedly in the debates in the 1970s on basic needs and, in more recent years, in the discussions on human development. Yet even as late as 1990 the World Bank allocated only 10.8 per cent of its resources to these activities. No doubt the Bank would justify its behaviour by claiming that all its projects benefit the poor indirectly even if not all are targeted directly on the poor. This defence need not be taken seriously however: it is an application of the old and long discredited thesis of "trickle down" development applied at the microeconomic or project level.

Finally, World Bank lending has been biased in favour of countries with a relatively high degree of inequality in the distribution of income and, as we saw above, there is little evidence within countries that the Bank used its influence to reduce poverty and inequality. In the 1960s about 22.2 per cent of the Bank's loans were allocated to countries where the degree of inequality was exceptionally high, namely, where the Gini coefficient was greater than 0.50. The proportion allocated to high inequality countries rose steadily in the 1970s and 1980s and by 1990 no less than 37.5 per cent of the Bank's loans were allocated to such countries. Thus the most that can be said about the Bank's policies toward inequality is that it is indifferent as to the distribution of income among individuals and households.

Taken together, these four points add up to a very strong condemnation of World Bank policies, at least when viewed from the perspective of reducing global poverty and promoting human development. Whether one considers the countries selected by the Bank to receive assistance, or the overall use of resources, or the project mix within the total portfolio or the policy commitment of recipient governments to reduce poverty--as indicated by Gini coefficients and the like--the Bank has been strongly biased against the poor.
 

Are development banks obsolete?

What, then, should be done about the World Bank and, by extension, to the other development banks? Some no doubt will argue for reform. We do not believe, however, that the World Bank (and similar institutions) can or should be reformed. As we shall explain below, the Bank should instead be wound up and the world should organize resource transfers from rich to poor countries by combining the features of an automatic and progressive income tax in rich countries with an automatic disbursement mechanism in poor countries, in which the resource transfer per head would vary inversely with per capita gross domestic product. Under such a scheme there would no longer be a need for the World Bank; it would have reached the end of its useful life.

We must make it clear, however, that it is not our intention to single out the World Bank for special criticism. The Bank is the largest, most prestigious and professionally the most competent of the international agencies (bilateral and multilateral) supplying long term capital to developing countries. If the World Bank does not deserve to have a future, one can be quite certain that the foreign aid enterprise as a whole, as presently constituted, does not deserve to survive to the twenty-first century.

Quite apart from possible reforms of the policies and procedures of international development banks, one must ask whether they continue to play a useful role. Are they obsolete or could they remain a pillar of a new framework for development cooperation?

When the World Bank was created it was intended to serve two purposes: to help finance the reconstruction of economies severely damaged during the Second World War and to supplement private capital in financing the long term development of economically backward countries. The first purpose was achieved several decades ago although some might argue that the needs today in Eastern Europe and Russia are fairly similar to those of Western Europe forty-five years ago. Even if one accepts this argument, however, it is not obvious that the needs of Eastern Europe and Russia should be met by the World Bank since a new institution, the European Bank for Reconstruction and Development, was created specifically to provide capital to the former Soviet bloc countries. Whether the EBRD should have been created is another matter.

The second purpose, that of supplying capital to the developing countries, no longer is as important as it once was. First, domestic savings rates in developing countries are much higher than they were, say, four decades ago and hence the developing countries are less dependent on foreign aid and foreign capital in general than in previous years. Second, private international capital markets have grown very rapidly and the original justification for publicly supplied international capital (by the World Bank and the regional development banks), namely the poor functioning of global capital markets, no longer applies. In fact, about two-thirds of the capital received by developing countries today originates in the private sector and only a third can be described as foreign aid. True, private foreign capital tends to by-pass the poorest countries, for reasons which are consistent with the logic of markets, but this can best be compensated not by providing subsidized loans from development banks but by giving untied grants to the poorest countries through a scheme such as that described below. Thus the role for international development banks is becoming insignificant and the time may have come to recognize this fact and gradually phase them out.
 

Aid to Eastern Europe and Russia

It is tempting to suggest that the need to assist the reform process in Eastern Europe and Russia will give a new lease on life to foreign aid, enabling it to abandon the constraints imposed by the Cold War, sort out the muddle of motives and design a truly effective aid programme. Aid to the former Soviet bloc countries, according to this view, could become a model for a new framework for development cooperation worldwide. Unfortunately, however, aid to Eastern Europe and Russia conforms more to the old model than to a new one and merely underlines the necessity of a fundamental reconsideration of the conventional approach to foreign aid.

First of all, the question of purposes and priorities remains unclear. In 1992 the developing countries as a whole, some 100 countries, received $70 billion in foreign aid in various forms. In 1993 one country, Russia, has been promised $43 billion. That is, Russia, smaller and more richly endowed than either India or China, will receive far more aid per head of the population than these two obviously poorer countries. Of course the promise to Russia may not be kept, commitments almost certainly will be smaller than the original promise and disbursements will be much smaller still.

Second, the promised aid is to come in many forms, not all of which are helpful. The precise composition of the $43 billion package is as unclear as the purposes and priorities of the aid. Some of the aid consists of lines of credit from the International Monetary Fund. These loans are conditional on IMF approval of further economic reform and it is quite possible, even likely, that some of the loans will not be drawn. Part of the aid consists of bilaterally supplied export credits, usually tied to the purchase of specified products produced in the donor country. The prices of these tied purchases are almost always above competitive world prices and hence the real value of the aid is significantly lower than the nominal value. Some of the aid consists of debt relief, including cancellation of debts which could not be repaid in the foreseeable future in any case. Finally, some consists of bilateral grants which are genuinely useful.

Third, the administration of the aid programme has been no less cumbersome than the administration of conventional aid programmes to developing countries. The IMF was expected to be the lead agency, but it has failed to fulfill that role efficiently and has been much criticized. In response, the IMF has created a "systemic transformation facility" which is supposed to channel soft funds to the former Soviet bloc countries. It remains to be seen whether this works and if so, how quickly aid flows out the pipeline.

Finally, the fundamental issue that must be addressed is whether a conventional aid package is the best way to help the process of economic reform in Eastern Europe and Russia. There are reasons for doubt, based largely on the disappointing record of foreign assistance in other parts of the world. In the case of Eastern Europe and Russia, the alternative to aid is trade. The OECD countries have been conspicuous in their unwillingness to grant access to their markets to exporters in Eastern and Central Europe and Russia. Agricultural products, manufactured goods and steel have been excluded from Western markets: a combination of domestic protection plus promises of foreign aid has been preferred to market liberalization and no aid. Yet the latter combination makes much more sense economically: it would enable the former Soviet bloc countries to earn the foreign exchange they desperately need; it would facilitate economic restructuring and allow them to exploit their comparative advantage, thereby increasing efficiency in resource allocation and the average level of income; it would integrate them into the global economy and help them to avoid long term dependence on foreign aid; and it would benefit consumers in the OECD countries by providing a wide range of commodities at lower prices.

Yet policy toward Russia and Eastern Europe has centred on aid, not trade. The approach is irrational and reflects the power of organized producer interests in the OECD countries. Everyone potentially could gain if aid were discontinued and were replaced by more liberal trade policies.7  Indeed the situation in Russia and Eastern Europe illustrates a major theme of this study, namely, that there are three sets of policies which in general are preferable to foreign aid: (i) macroeconomic policies which stimulate growth, (ii) trade policies which liberalize access to markets and (iii) immigration policies which liberalize the international labour market. With the possible exception of the very poorest countries, foreign aid is a fourth-best policy. The qualification, however, is important. Although the three sets of preferred policies can be mutually beneficial to all countries, there is no presumption that the gains from faster growth, freer trade and a more liberal global labour market will be evenly distributed. On the contrary, as we will show below, there is a high probability that if market forces are allowed to operate freely, global inequalities will increase. It is in this context that foreign aid can play a useful role, namely, to reduce inequality in the distribution of world income and provide a global safety net for the poorest countries.
 

Technical assistance

The discussion so far has concentrated on capital flows, whether provided by multilateral or bilateral aid agencies. Development cooperation, however, also takes the form of providing technical assistance to developing countries, and this too can be supplied either by multilateral or bilateral agencies. By far the largest provider of technical assistance is the United Nations Development Programme and our observations on conventional technical assistance programmes will centre on the role of UNDP.

The assistance to developing countries provided by UNDP is in the form of grants and because of this, in terms of net resource transfers, UNDP is one of the largest aid agencies in the world, substantially larger than the regional development banks and broadly comparable in size to the World Bank. It is no exaggeration to say that UNDP is one of the pillars of the current framework for development cooperation. The question we raise is whether it should remain a pillar within a new framework for cooperation.

Until quite recently UNDP did not provide funds directly to developing countries to pay for technical assistance. Instead it operated through the specialized agencies of the United Nations (UNESCO, UNIDO, ILO, WHO, FAO, etc.). In theory the governments of developing countries requested technical assistance from the specialized agencies and the UNDP covered the costs, including a generous contribution to cover the overhead costs of the specialized agencies. In practice, however, technical assistance projects originated in the specialized agencies and the agencies then sold the projects to the recipient countries. Selling activities often were rather intense as the specialized agencies came to depend on UNDP funds as a major source of revenues.

A number of problems arose from this method of supplying technical assistance. First, competition was created among the specialized agencies of the United Nations as each sought to maximize its revenues from UNDP. There was no coordination of technical assistance at the national level or a careful determination of priorities. In principle the Resident Representative of UNDP in each country could have played this role, but in practice each U.N. agency pursued its own objective. Because of the incentive structure unwittingly created by UNDP, each agency tried to maximize its income rather than assist the development of the recipient countries.

Second, within their area of competence, each specialized agency had a virtual monopoly in providing technical assistance. UNDP was so large that bilateral agencies and non-governmental organizations rarely were able to provide effective competition. The U.N. agencies had a clear field. The consequence, third, is that the quality of the technical assistance that was provided often was mediocre. The "experts" sometimes were unsuitable and the value of the advice given to the recipient country often was low. Even worse, finally, because the opportunity cost to the recipient country often was nearly zero, developing countries received technical assistance they neither wanted nor needed. The system was supply-side driven: the agencies supplied the experts, the UNDP paid the bills and the developing countries were the passive recipients.

The results, not surprisingly, were far from satisfactory. UNDP responded by improving the structure of incentives in two important ways. First, each country was assigned a notional budget for technical assistance projects. The effect of this was to introduce an opportunity cost at the country level: if the offer of project A from agency X is accepted, then the offer of project B from agency Y may have to be declined. Countries thus have to choose and think about their priorities.

Second, UNDP broke the monopoly of the specialized agencies in providing technical assistance. Henceforth agreed technical assistance projects are put out to tender and many organizations are encouraged to submit bids, including of course organizations in the private sector. Governments now have a greater choice of supplier; there is more competition and there is reason to hope that the cost of technical assistance projects will fall and the quality increase.

Although governments now are able to exercise wider choice within the area of technical assistance, their national allocation from UNDP still is restricted to technical assistance projects. The logical next step in the reform process would be to remove this restriction and allow governments to use UNDP funds for any developmental purpose, be it technical assistance, investment in human capital, investment in physical capital or investment in natural capital. UNDP would then become a grant giving agency that supplies untied funds to developing countries.

Such a transformation of UNDP would have profound implications for the framework of development cooperation. It would shift the balance of power in favour of developing countries. It would force the specialized agencies to become fully competitive in supplying technical assistance, since not only would they have to compete against other suppliers of technical assistance, they would also have to show that expenditure on technical assistance yields higher returns than, say, additional expenditure on secondary education. Such a transformation also would reduce dramatically the cost of delivering foreign aid since there would no longer be a need for a huge bureaucracy to administer tens of thousands of small technical assistance projects. Finally, provided UNDP grants were distributed on the basis of need, as reflected in average income per head of a country's population, and provided the governments of recipient countries used the aid wisely, foreign assistance would become more equitable and contribute to a genuine reduction in global poverty.

Growth as a substitute for aid

With the possible exception of some of the very poorest countries on the margin of the global economy, even a better administered and more generous flow of aid would be a poor substitute for faster growth in the donor countries. The reason for this is that the OECD countries still account for over 70 per cent of world output and hence growth in these countries provides a strong stimulus to growth elsewhere, and particularly in the developing countries. Indeed it has been estimated that "a one per cent increase in OECD growth may raise growth in developing countries by an average of 0.6-0.7 percentage points."8

Official development assistance in 1990 accounted for 0.36 per cent of the GNP of the donor countries. (It has since fallen to 0.33 per cent.) Seen from the perspective of the poorest developing countries, aid represented $12.65 per capita or 1.3 percent of their GNP. The inflows of foreign aid thus are relatively small. (See Table 4.2 on page 39.)

Let us compare two possible alternative policies that could be followed in the OECD countries. First, macroeconomic policies might be adopted which raise aggregate demand and domestic output. Assume these policies increase the rate of growth by one percentage point, a perfectly feasible outcome given the high unemployment and slow (and even negative) growth currently being experienced in the developed countries. Assume also that the acceleration of growth in the OECD countries does in fact raise the growth rate in the developing countries by 0.6 percentage points. Clearly both groups of countries would benefit, although rich countries would benefit more than poor and global inequalities consequently would increase.

Compare this with a second possible policy of increasing the amount of foreign aid supplied to developing countries. By how much would aid have to increase in order to raise the growth rate in developing countries by 0.6 percentage points? To answer this question let us assume heroically, contrary to experience, that all the additional aid is devoted to growth enhancing (capital) expenditure. Let us also assume that the incremental expenditure-output ratio is 3. It then follows that to achieve the growth objective, additional foreign aid equivalent to 1.8 per cent of the GNP of the developing countries would have to be provided by the donor countries. That is, the amount of aid would have to more than double and this of course would have to be financed by higher taxation (or lower public expenditure) in the donor countries. Such a policy is unlikely to be feasible either now or in the foreseeable future.

Faster growth is a feasible policy; more aid is not. Faster growth would benefit everyone, in rich countries and poor alike; more aid would require additional taxes or lower expenditure in rich countries and possibly increase marginally the hardships of some who already are suffering from the unsatisfactory performance of the economy. When all of this is combined with the fact that the record of foreign aid in promoting development is far from persuasive, it would seem sensible to build a new framework for development cooperation around policies for faster growth rather than on larger commitments of foreign aid.
 

A new basis for financing aid

There is widespread disappointment, to use the World Bank's own words, that "aid has done much less than might have been hoped to reduce poverty"9 and this prolonged disappointment has undermined the support of humane internationalists in the developed countries for foreign aid programmes. The basis for aid will have to be reconstructed if the support for aid is not to wither away. Yet the moral case for the people of rich countries to help those in poor countries remains intact.10  Those who are prosperous do have a moral obligation to assist those who live in poverty and thus there is a prima facie case for a transfer of resources from rich countries to poor. It does not follow from this, however, that there is a moral obligation to support official aid programmes as presently and historically organized. As Rogell Riddell, someone who has thought hard about these issues, correctly says, "the bedrock question about foreign aid" is whether it assists the poor.11  If it doesn't, and if it can't be reformed so that it does, the moral case for official aid programmes collapses.

The thrust of our argument is that enough years have passed to be able to assess the effectiveness of official development assistance and on the whole the assessment turns out to be rather negative. The time has come to think again. Let us be clear, however, that what needs rethinking is long term foreign assistance intended to alleviate poverty by promoting human development. We are not challenging the need for such things as emergency assistance during famines and periods of economic distress or aid for political refugees.12

One issue that needs to be resolved is the amount of aid provided and the way it is financed.13  At present official development assistance is financed by what can be described as a system of voluntary taxation of rich countries. In a rational world one would expect that the burden of aid "taxation" would fall most lightly on donor countries with the lowest per capita income and gradually rise as per capita income increases. That is, one would expect the "tax rates" to be progressive. Yet if one compares GNP per capita of the 18 major OECD donor countries with each country's aid burden as measured by the percentage of GNP allocated to official development assistance, one discovers that the aid burden is randomly distributed among donors: Spearman's coefficient of rank correlation is not significantly different from zero.

The United Nations aid target of 0.7 per cent of GNP implies a proportional tax rate. This evidently is inequitable since donor countries with very different levels of average income are none the less asked to contribute the same proportion of their income as aid. New Zealand, for example, with a per capita income in 1990 of less than $13,000 would be expected to give as much aid as, say, Norway with a per capita income of more than $23,000. A strictly proportional aid burden would be less inequitable than the present randomly distributed burden, but it would not correspond to what most people would regard as fair. The present system in which three very rich countries (the United States, Japan and Switzerland) contribute less than the average percentage is utterly indefensible, but it would be almost as difficult to justify a system that requires Ireland (with a per capita income of $9,550) to pay as much as Switzerland (with a per capita income of $32,680).

When creating a new framework for development cooperation, the objective should be to end the present system where aid contributions are voluntary, the aid burden is distributed randomly and inequitably, and the flows of aid are unpredictable because they are subject to annual appropriations by national parliaments. Instead the world should move to a system where contributions to the aid effort are obligatory, the burden is distributed progressively and the annual flows are predicable. In short, what is needed is a progressive international income tax on rich countries administered by an international authority such as the United Nations. A reorganized UNDP, for example, could be entrusted with this task. Let us call this new organization the UNDP. The idea of a progressive international income tax to finance foreign aid is not new14 and if development aid is to have a future and be more than marginal in size, the idea should be taken seriously.

The design of such a scheme could be very simple and easy to implement. We present one scheme to illustrate how it might work. The first issue that must be resolved is what countries would be liable to an international tax. We suggest that the cut-off point be a real GDP per capita in 1990 (expressed in US dollars of purchasing power parity) of $10,000.

Sources: UNDP, Human Development Report 1993, New York: Oxford University Press, 1993; and World Bank, World Development Report 1992, New York: Oxford University Press, 1992.

Countries with a real income higher than $10,000 would be liable to the tax; those with an income less than $10,000 per capita would be exempt. Using this somewhat arbitrary criterion, 31 countries would be liable, ranging from the United States at the top to Ireland at the bottom. The countries are listed in the first column of Table 1.1. We shall refer to these countries as Group A.

The second issue is the rate of taxation. The tax schedule should be progressive in order to take into account differences in real income among the donor countries and it also should be simple in order to avoid disputes over tax liabilities which turn on alternative estimates of a country's GDP. We suggest that there be only three tax rates and that the tax base be a country's gross domestic product as conventionally measured, i.e., not adjusted for purchasing power parity. Specifically, we suggest that a tax rate of 0.375 per cent be applied in countries with a real GDP per capita (in PPP terms) greater than $16,000. Sixteen countries, from Norway to the United States, fall into this category. Next, we suggest that a tax rate of 0.25 per cent be applied in countries with a real income between $12,000 and $16,000. This category includes eight countries, from New Zealand to Italy. Finally, we suggest that a tax rate of 0.20 per cent be applied in countries with a real income between $10,001 and $11,999. This category includes seven countries, from Ireland to Spain.

Application of these rates of taxation implies that some countries would contribute less foreign aid than they do at present, notably the Nordic countries, France and the Netherlands. Others would contribute more than they do at present, notably the United States and Austria. Still others would join the club of donors for the first time, e.g., Singapore, Hong Kong and the Bahamas. Those countries which wished to contribute more to the aid effort than required by their tax obligation would of course be free to do so. They could either make voluntary contributions to the international development fund (UNDP) or they could supplement the multilateral programme with a bilateral programme.

The scheme described in Table 1.1 produces a tax yield of nearly $56.8 billion, excluding Brunei for which data are not available. This is only slightly more than the $55.6 billion contributed by the OECD countries in 1990. Moreover, all of the aid under this scheme would be in the form of grants and none of the grants would be tied in any way. Hence the real value of the aid under an international tax financed scheme undoubtedly would be considerably higher to the recipients than the value of the present mixture of loans, export credits, surplus commodities, etc.

Under our suggested scheme, foreign aid would represent 0.35 per cent of the GDP of donor countries as conventionally measured. This can be compared with the actual situation in 1990 when foreign aid accounted for 0.36 per cent of the GNP in the OECD countries. By 1992 the proportion had fallen to 0.33 per cent.
 

Disbursements under a negative international income tax

Once the question is resolved of how to finance foreign aid under a new framework for development cooperation, attention can be turned to a second issue, namely, the criteria of eligibility to receive development assistance. If foreign aid is to enjoy the support of the public in donor countries, the criteria used to select recipient countries must be clear and fair. That is far from the situation at present. Indeed we now have an extreme case where Ireland is an aid donor and Israel a large aid recipient, yet Israel's per capita income is higher than Ireland's. Such a situation obviously could not be allowed to continue if aid were to be financed through a system of international taxation administered by the United Nations.15  Indeed in the tax scheme described in Table 1.1, Israel becomes a donor country.

What is needed is an internationally agreed cut-off point so that only those countries below the critical point would be eligible for foreign assistance. The dividing line could be either absolute or relative. One possibility would be to make a real per capita income (in PPP terms) of, say, $1,500 in 1990 the cut-off point. This corresponds to the average income in Honduras, Zimbabwe and Nicaragua and implies that 51 developing countries would be eligible for aid. The countries are listed in Table 1.2. We shall refer to these as the Group C countries.

An alternative possibility would be to declare eligible for aid those countries that fall, say, into the bottom quintile of the global distribution of income. Either criterion would permit the composition of eligible countries to change, some graduating and ceasing to be eligible when their incomes rise above a certain level and others becoming eligible as their circumstances deteriorate. In the example presented in Table 1.2, we use an absolute criterion of eligibility, namely, those with an average income no higher than $1,500.

The total population of the 51 countries eligible under this criterion (the Group C countries) is just over 1.5 billion persons. If the entire international development fund of $56.8 billion were divided equally among the eligible recipients, the average aid allocation per person would be about $37.87. We suggest, however, that it would be more equitable to allocate more than average to the poorest countries and less than average to those who are not quite so poor. To illustrate how this might work, in Table 1.2 we divide the eligible recipients into two groups.

The poorest 27 countries, with a real (PPP) income per head of less than $900, are allocated $50 per head of foreign aid. This group of countries, from Zaire to Bangladesh, contains 385.4 million people and merits special assistance. The second group contains 24 countries with a real income per capita between $900 and $1,500. This group, from the Gambia to Nicaragua, contains nearly 1,182 million people. The aid allocation to this group is $30 per head.

Note:* refers to 1991.
Sources:UNDP, Human Development Report 1993, New York: Oxford University Press, 1993; World Bank, World Development Report 1992, New York: Oxford University Press, 1992.

Such a system of grants, a negative international income tax, could become a powerful mechanism for ameliorating and ultimately eliminating world poverty. Much would depend however on the policies of the governments of recipient countries. Under the global transfer programme we recommend, aid would be concentrated where it is most needed. The amount of aid per head of the population would be large enough to have a significant impact on the well being of people. And the amount of aid received, expressed as a percentage of GDP, would make it possible, if it is used wisely, to accelerate markedly the rate of growth in the recipient countries. The two largest countries in our group of 51, Bangladesh and India, would benefit enormously under our aid disbursement scheme. Bangladesh, for example, currently receives aid equivalent to about 9.5 per cent of its GDP; under the negative income tax described in Table 1.2, aid inflows would be equivalent to 23.3 per cent of its GDP, and all of the aid would of course be in the form of grants. Aid to India is only about 0.6 per cent of its GDP; under our scheme it would rise to 10 per cent. Some countries would lose, e.g., Zambia and Ghana, because their entitlements under a negative income tax would be less than the amount of foreign aid they now receive. And of course those countries that now receive aid but are not included in our group of the 51 poorest countries would cease to receive foreign aid under our scheme. Countries such as Argentina, Malaysia, Tunisia and Botswana would not be eligible for assistance from the international development fund. The limited resources available for international transfer payments would be channelled to countries with the lowest real incomes.
 

Alternative allocation criteria

Disappointment with the results of foreign aid has led donors to move in the opposite direction from that we recommend and to consider applying even more conditions on aid recipients than exist at present. Specifically, disappointment has led donors to examine more critically the general development policies adopted in recipient countries and, more recently, to use foreign aid as an incentive to governments of recipient countries to enter into a "policy dialogue" with the International Monetary Fund, World Bank and the numerous bilateral development agencies. The World Bank now proposes to take this process a step further and link development assistance to what it calls "a country's commitment to pursue development programs geared to the reduction of poverty."16  Aid would thus be tied to "commitment" and presumably in due course to success in "the reduction of poverty". Given what we have said above about the Bank's own commitment to a reduction in global poverty, this pronouncement may seem rather presumptuous. Let us put this point to one side however.

It is not entirely clear how "commitment" would be judged or by whom, although it seems the World Bank has itself in mind, possibly in conjunction with bilateral donors and even non-governmental organizations. The developing countries, however, almost certainly would object to aid allocations under an internationally funded programme being determined by the views donors happen to have at a particular moment about what constitutes "a sound development strategy". Such a formula for aid allocations would replicate on a larger scale the antagonism that has existed for many years between the developing counties and the IMF over short term balance of payments support. Even if one regards the views of the IMF as "sound", the fact that the Fund is controlled by the rich countries and has often imposed its policy views on poor countries against their better judgment is cause for regret.

Moreover, if the Fund and the Bank begin to impose conditions on aid recipients in the form of a "commitment . . . to the reduction of poverty", there is nothing to prevent others from imposing their own, different, conditions. UNDP might be tempted to make aid conditional on the implementation of a human development strategy. The ILO and trades unions in developed countries might wish to link aid to labour standards. UNEP and the environmentalists might link it to measures to protect the environment. Still others might attempt to make foreign assistance conditional on respect for human rights, or holding free elections or introducing parliamentary democracy.17  There is no end to the number of good things that can be tied to aid. The danger is that once again donors will illustrate the truth of Oscar Wilde's aphorism: "Philanthropic people lose all sense of humanity. It is their distinguishing characteristic."

The course of wisdom is not to encumber the international aid programme with a host of extraneous conditions but to use it to transfer resources to countries which are most in need. A mechanical rule such as a negative income tax, which accepts the rough with the smooth, the good regimes and the bad, is perhaps the best way to do this.

A radically different approach to both a negative income tax and donor-imposed conditionality would be to allow the eligible recipient countries to determine the allocation of foreign aid among themselves. While superficially attractive, and successful in Europe after the Second World War in allocating aid under the Marshall Plan, this option has three serious disadvantages. First, it would not be acceptable to the donors, or to the taxpayers in donor countries who ultimately have to provide the resources. Second, the number of countries involved in taking decisions about aid allocations would be large and the decision making process probably would be cumbersome and unwieldy. Third, if decisions in such a group were to be taken on the basis of one country, one vote, there would be a strong built-in bias against poor and populous countries such as Bangladesh and India. Thus there would be a danger that one of the undesirable consequences of the present system of allocating aid would be perpetuated.

There are of course intermediate positions. Decisions about aid allocations could be taken by representatives from countries which are neither donors nor recipients, i.e., the Group B countries. Such persons would be likely to be disinterested and their views on a government's commitment to reducing poverty, promoting human development or whatever, and the actual effectiveness of the policies introduced, would perhaps command international respect and acceptance. A compromise solution would be to form a tripartite body consisting of representatives of donors (Group A) and of recipients (Group C) plus persons belonging to neither group (Group B). Under such a formula the disinterested members would in effect have the final word whenever donors and recipients failed to agree.

All of this presupposes that aid allocations should be discretionary and reflect internationally agreed development objectives. But it is not obvious that this is the best way to proceed. While it is desirable, it is not absolutely essential to have agreement on objectives, and it may be sensible for a variety of reasons for some countries to depart occasionally from generally agreed objectives. Even less is it necessary to reach agreement on the degree of commitment of every potential aid recipient to international objectives, such as the eradication of poverty, as the World Bank proposes. Indeed, it is doubtful that it is possible accurately to assess the strength of commitment of governments to particular objectives. Finally, there is also room for disagreement about the effectiveness of policies adopted in different countries to reduce poverty, promote human development or achieve some other objective. A case in point is the intense debate in Russia and Eastern Europe about the effectiveness of different policies in bringing about a transition to a market-guided economy.

In our view it would be far simpler--cheaper, less bureaucratic, ultimately more humane--to transfer automatically each year, by applying a previously agreed rule, all of the foreign aid available in an international development fund to all of the eligible countries. Ideally the distribution rule should contain an element of progressivity, so that countries with the lowest income per head would receive the most aid per head. Such an international negative income tax would be symmetrical to the progressive international income tax we have proposed be levied on donor countries.

It will be objected that if aid is distributed automatically to poor countries, without subjecting them to a performance test, there will be no guarantee that resources will not be misused. This is indeed a risk. In fact, it is a certainty that some resources will be wasted, misappropriated or used for purposes other than development. That is a fact of life. But we know that under present arrangements aid has been given and used in such a way that it has neither contributed to faster growth nor promoted human development nor significantly reduced poverty or inequality. It is hard to believe that the scheme we propose would actually result in a worse outcome. On the contrary, the outcome almost certainly would be better. The combination of a progressive international income tax levied on eligible donor countries and a negative international income tax applied to eligible recipients would produce a scheme for funding and disbursing foreign aid which is transparent, fair among countries, automatic, predictable and inexpensive to administer. The present system of foreign aid possesses none of these virtues and is ineffective as well.

1:  See, for example, Keith Griffin, "Foreign Aid After the Cold War," Development and Change, Vol. 22, 1991, pp. 645-685.

2:  World Bank, World Development Report 1992, New York: Oxford University Press, 1992, Table 9, pp. 234-5.

3:  "Nothing To Lose But Your Chains," The Economist, 1st May 1993, p. 44.

4:  Ibid.

5:  "Fat of the Land," The Economist, 29 May 1993, p. 16.

6:  The comments below are greatly influenced by Mahbub ul Haq, "Political Dimension of World Bank Policies," mimeo., August 1992.

7:  See Jim Rollo and Alasdair Smith, "The Political Economy of Eastern Europe's Trade with the European Community: Why So Sensitive?", Economic Policy, No. 16, April 1993.

8:  OECD, Development Assistance Committee, Development Co-operation, Paris: OECD, December 1992, p. 35.

9:  World Bank, World Development Report 1990, New York: Oxford University Press, 1990, p. 127.

10:  The moral case for providing foreign aid is discussed at length in Roger C. Riddell, Foreign Aid Reconsidered, Baltimore: Johns Hopkins University Press for the Overseas Development Institute, 1987, Part I, chapters 1-7.

11:  Ibid., p. 217.

12:  Short term emergency assistance and aid for political refugees urgently needs re-examination in order to increase its effectiveness. We believe that it should be independently funded and administered separately from the transfer scheme recommended below. Emergency assistance should be available to all countries needing it regardless of the level of per capita income and should be seen as a form of global insurance against disasters. The use to which the funds are put, however, should be much more closely linked to long term development objectives than has so far been customary.

13:  This paragraph and the following draw heavily on Keith Griffin and Azizur Rahman Khan, Globalization and the Developing World, Geneva: United Nations Research Institute for Social Development, 1992, pp. 33-6.

14:  See, for example, the Report of the Independent Commission on International Development Issues (Brandt Report), North-South: A Programme for Survival, London: Pan Books, 1980, ch. 15. Also see Arjun Sengupta, "Aid and Development Policy in the 1990s," UNDP, mimeo., 5 January 1993. The academic literature on the subject includes Paul Rosenstein-Rodan, "International Aid for Underdeveloped Countries," Review of Economics and Statistics, Vol. 43, 1961; Irving Kravis and Michael Davenport, "Political Arithmetic of International Burden Sharing," Journal of Political Economy, Vol. 71, 1963; Keith Griffin and Azizur Rahman Khan, op. cit., pp. 33-6. A progressive international income tax was endorsed in the UNDP, Human Development Report 1992, New York: Oxford University Press, 1992, P. 79.

15:  As mentioned in the text, countries would of course (whether official aid donors or not) remain free to raise resources through national taxation and to transfer those resources to other countries (whether eligible for internationally financed aid or not). Thus there would be nothing to prevent the United States from continuing to give money to Israel, but such a resource transfer would not count as foreign aid and would not reduce the obligation of the United States (or Israel) to carry its fair share of the international aid burden.

16:  World Bank, World Development Report 1990, loc. cit., p. iii; also see p. 4.

17:  The Economist notes (10 July 1993, pp. 11-12) that the "most effective aid pushes its recipients towards sound government. Increasingly . . . that has come to mean making aid conditional not only on economic reforms but also on free elections and respect for human rights." No evidence is given of the effectiveness of aid in promoting human rights, democracy or even successful economic reform.