Journal Issue

The Allocation of Aid by the World Bank Group

International Monetary Fund. External Relations Dept.
Published Date:
September 1972
  • ShareShare
Show Summary Details

Andrew M. Kamarck

Anyone examining the flow of aid to less developed countries (LDCs) over the past 10 years or so will be struck by a conspicuous trend—the increase in importance of the multilateral agencies in this flow. Not only has there been a striking increase in the actual amounts of funds disbursed by them, but they have made a greater relative contribution to the development process, and finally, have acquired the central leadership role in the whole process.

The net flow of funds (i.e., net of amortization -repayments) to developing countries from multilateral agencies has been increasing much more rapidly than that from any other source. From 1960 to 1970 total net flows from all sources, both public and private, reaching the LDCs rose by 80 per cent (from $8.1 billion to $14.7 billion), while from multilateral sources they increased by 400 per cent (from $0.3 billion to $1.5 billion). In the total of Official Development Assistance made available by the noncommunist industrialized countries (i.e., the “DAC countries” or the members of the Development Advisory Committee of the Organization for Economic Cooperation and Development (OECD)), the proportion of contributions to multilateral organizations has grown rapidly. The total Official Development Assistance of the DAC countries made available bilaterally rose from $4.1 billion in 1960 to $5.7 billion in 1970 or an increase of 37 per cent, while their contributions to multilateral agencies rose from $0.5 billion to $1.1 billion or an increase of 110 per cent.

The flow of funds from the multilateral agencies in the early 1960s was roughly equal to that provided by the communist countries, but by the end of the decade, the rate of flow from the communist countries had remained about the same and the multilateral flow was three times larger.1

By 1970, the World Bank Group of agencies had become the largest single financier for the less developed countries—bigger even than the U. S. Agency for International Development (AID). In 1970/71 the World Bank Group’s commitments of new loans, credits, and investments totaled $2.6 billion and in 1971/72 the total was expected to reach about $2.8 billion or twice as much as all the other multilateral agencies put together. The Inter-American Development Bank is lending at a rate of about $650 million a year; the United Nations Development Program (UNDP) spends about $280 million a year; the European Economic Community’s European Development Fund (EDF) commits about $220 million a year; the Asian Development Bank, $250 million; and the African Development Bank, about $25 million.

Measurement Problems

A great deal of progress has been made in recent years in improving the data on flows of funds to the LDCs, and more improvement is currently in preparation. The principal statistics on the flow of official funds to the LDCs from the industrialized countries have been compiled by the OECD. A new joint reporting system of the World Bank Group and the OECD based on detailed reports from the donor countries on individual capital transactions with the LDCs is still in the process of “running in.” This will supplement the World Bank’s debt reporting system from the recipient countries which is also being broadened in scope and deepened in detail. Consequently, in a few years considerably better data will be available to register and to make possible better analyses of various aspects of the flow of official funds to the less developed countries.

As it is, the pioneer work of various economists, such as John A. Pincus, Goran Ohlin, and Wilson E. Schmidt, in developing ways to reduce various financial terms to a common measure (e.g., “grant equivalent”)2 has made it possible for the OECD to produce a considerable amount of useful analytical data on the basis of which the aid performance of the various donors can be compared.

“By 1970, the World Bank Group of agencies had become the largest single financier for the less developed countries…

These measures have proven themselves useful. But, it must also be said that for our purposes some other data—which are only now beginning to be collected—are also needed. The presently published statistics are all statistics that focus on the “input”; what are missing are statistics on “output.” The available data on flows all concentrate on presenting the amounts and the terms of the funds made available to the LDCs. The information tells us what the actual and grant equivalent terms are of loans or credits made for a particular project or to a particular country. This is needed and useful information. But, until recently there has been no information systematically collected or compiled on what the rates of return are on the funds after they have been invested, or even on the returns that are forecast when the investment is made.

Obviously this information is extremely difficult to collect, but it would do more than satisfy idle curiosity. Such information, in fact, concerns the very heart of the development process—how great a contribution capital makes to the economic growth of a country. Existing DAC data concentrate on measuring and comparing the amounts and the conditions of the aid made available by donors and on how these amounts are distributed among the LDCs; the very important other side of the picture has been blank. No data have been collected to measure the productive contribution of the funds made available and to compare what happens to the investments once the funds cross the borders of the LDCs.

The grant equivalent is an important bit of information about a particular amount of aid made available to an LDC, but the real measure of aid being extended by the particular transaction is derived from the difference between the discounted total of the returns on the investment and the amounts of repayment due. This difference might be called the real aid equivalent of a particular investment. For example, a 100 per cent grant that finances a white elephant in an LDC makes less of a development contribution than a high interest rate loan that results in developing a high income resource. It is clear that the U. S. and the British 100 per cent grants to the Libyan Arab Republic in the early 1950s provided much less help to the Libyan Arab Republic’s economic development than the oil companies’ investments even if the latter had paid a return to the companies at the equivalent interest rates of, say, 30 per cent a year.

The conventional measure, grant equivalence, can to some extent be regarded as a measure of the sacrifices made by donors in such loans. Occasionally, the proposal is made that the grant equivalent should be calculated at the opportunity cost of such funds to the borrower—say, at the alternative rates available to them in foreign capital markets; it is argued that if such calculations were feasible this would measure the benefits received by an LDC in receiving such loans. If, as was true of international loans before World War II and is occasionally true today, the foreign exchange were made freely available to the borrower, this procedure would indeed give an indication of the benefits. But most foreign loans, credits, and grants today are tied in one respect or another—they are as a minimum project-specific. It has been argued that all investment resources of a government are fungible and, therefore, even if a loan is tied to a project, it merely frees funds to make possible the marginal investment elsewhere thus playing the same economic role as the international loans of the 1920s. In actual fact, however, this is not true of most governments—if a particular entity or department is able to get a loan or grant abroad, the allocation of investment funds in the public sector is affected to a greater or lesser degree accordingly. Consequently, the way in which investment funds from abroad are made available does have an impact on the pattern in which investment takes place in the country. (To be completely accurate in measuring the real aid equivalent of an investment from abroad, the return would have to be weighted by the proportion of the investment in the project in question that the loan really made possible at this time and the marginal investment made possible elsewhere. But these refinements can wait until there is some systematic information available on returns.)

In this connection, not only is the grant equivalent an inadequate indication of the output worth of economic aid, but so is the net flow concept. Again, to measure the input contribution of a donor, the net amount it is lending abroad after receipt of amortization is relevant. But, an LDC can get considerable help from a new investment of funds making possible a highly productive project even if this investment is offset by an equal outflow of amortization payments on older investments. (This is on the fairly realistic assumption that the LDC does not have its own capital market that can mobilize savings and rechannel them to the most productive uses.)

This problem of measurement has direct relevance to the subject of this article: the World Bank Group has from its initiation emphasized the project approach in the granting of loans and credits; that is, it has considered loans only from the standpoint that an individual investment has to show a good prospect for a satisfactory rate of return before it will be considered for finance. The bilateral donors have emphasized this much less and in some instances, not at all. While no comparative study has been made of the subject, it is highly probable that the rate of return on World Bank Group projects has been higher than that of the investments financed by most or all bilateral donors. (For the 90 per cent of the projects financed by the World Bank Group in 1970/71 where it was possible to prepare estimates of the rate of return, the average was 18 per cent.) But whether these investments have done relatively better or not, the point is that what is relevant for the LDCs is not only the size and allocation of the World Bank Group’s IDA funds, which are about 84 per cent grant equivalent, but also the allocation of the hard Bank loans and IFC investments. Of total World Bank Group investment in 1970/71 of $2.6 billion, hard loans by the International Bank for Reconstruction and Development (IBRD) amounted to $1,896 million; International Development Association (IDA) credits amounted to $584 million; and International Finance Corporation (IFC) equity investments and loans at negotiated rates amounted to $101 million.

Allocation Criteria

One necessary economic condition for any World Bank Group 3 investment in a country is a project condition: a reasonable assurance that the proposed project to be financed will be carried out and operated successfully, that it will fit into the development strategy of the country, and that it will provide a satisfactory return on the total investment. A set of several conditions applies to the country. For an IBRD loan the country must meet the test of positive credit-worthiness, i.e., that in the Bank’s judgment a proposed loan is within the limits of the country’s capacity and willingness to service external debt, taking account not only of its existing and prospective debts to the Bank but also to all other sources of external finance. For an IDA credit, the test is negative—the country must be judged to have insufficient capacity to service debt equivalent to its longer-term capital import requirements.

For both the Bank and IDA, another country condition is the suitable economic performance of the country. This is a complicated subject.4 But briefly, the character, magnitude, and timing of Bank Group lending to a particular country is affected by the commitment and the effectiveness of the effort (in relation to its capacity to make an effective effort) the country is making to try to cope with the developmental obstacles it confronts, as limited by the exogenous developmental opportunities facing the country. The greater the government’s commitment in trying to cope with the obstacles and the greater the effective effort (and these two are not necessarily directly proportional), the greater the willingness of the Bank Group to try to provide help and the greater the opportunities that open up for effective investment. But the latter is also, in the short run and medium run, limited by exogenous factors—natural resources, world market conditions, etc.

There are also two important boundary conditions relating to per capita gross national product (GNP) that affect both the Bank and IDA: compared to Bank funds, IDA funds are more scarce (being derived from government contributions). As a first step to cut the demand for IDA funds, an arbitrary limit has been placed on the kind of countries eligible to be considered for IDA credits. Originally, only countries with a per capita GNP of $250 or under were eligible, but since 1968 this limit has been raised to $300. This is in part because the very poor countries have a much longer way to go than a middle-income country before they reach a tolerable level of development and, therefore, their creditworthiness is likely to be less for hard loans. (If a country will need to import capital net over a period of, say, 50 years and has to depend on hard loans at high interest rates and with maturities of 25 years or less, the need to borrow to repay interest and amortization on old debt as well as to get a net capital flow will soon cause an explosive growth of external debt.) Another argument to justify this policy is that one of the reasons a country is very poor is that the rate of return on investment in the country is not likely to be as high as in other countries. In any case, allowing the very poor country to keep most of the benefit of any return on an investment financed from abroad (i.e., maximizing the aid equivalent of the loan in the terminology suggested above) should help the country in its attempt to catch up with the rest of the world.

Within this per capita GNP limit, there is a further limit. As some of the governments of the richer countries5 that make periodic contributions (replenishments) to IDA would find it more difficult to do so if the investment of IDA funds were too concentrated in a few large countries, it is understood that the amounts going to India and Pakistan (former boundaries) and Indonesia have been restricted to somewhat over half of the total.

While IDA credits are restricted to countries where GNP per capita is $300 or less, Bank loans are not restricted to countries above $300. Bank loans are made to countries with GNP per capita below $300 if the creditworthiness prospects are good enough. For many of these countries Bank loans and IDA credits are blended to provide terms harder than IDA but softer than Bank terms. The character of the blend is determined by the capacity of the country to carry additional external debt as governed by its expected future growth path for net capital imports. It is obvious that this cannot be a precise determination since, among other reasons, the variables are by no means independently determined.

Bank lending differs from IDA credits in that there is no precise upper limit of per capita GNP that acts as a cutoff. In general, and as a matter of practice, countries with a per capita GNP above $1,000 in 1970 usually do not attempt to borrow from the Bank. There are exceptions to this practice. For example, Finland has been a borrower from the Bank largely because its special geographical position impedes the access to private capital markets that would otherwise be regarded as normal. Again, a country confronting a special structural problem may need an inflow of capital considerably greater than the private capital markets are willing to provide while it is making its structural changes; an example is New Zealand that could be severely hurt by losing markets for its principal agricultural exports through the United Kingdom’s joining the European Economic Community.

“… the capability of the developing countries to make productive use of resources has increased considerably.”

On the other hand, Japan, because it was able to grow at a very high rate even without the help of Bank loans, was no longer considered as a potential borrower when its per capita GNP was approaching, but had not yet passed, the $1,000 mark.

The Pearson Commission in its report recommended that the “IDA should formulate explicit principles and criteria for the allocation of concessional development finance and seek in its policies to offset the larger inequities in aid distribution.”6 The Commission argued that for political and historical reasons the distribution of official bilateral aid has been extremely uneven and bears little relation to economic factors and that there has been a tendency for large countries to receive less assistance per head than smaller ones. This statement of the Commission is fairly well documented and can be accepted as true.

The World Bank Group considered the Pearson Commission recommendation carefully. It was clear that the IDA could hardly escape giving consideration to the amount of assistance that prospective borrowers may be receiving from other sources of finance. IDA allocations could not be made without reference to the policies pursued by other donors. It was also clear that it would be impracticable for the IDA to adopt as the principal criterion for all of its lending that countries judged to be in the same situation should receive equal amounts of per capita aid. It is quite possible that the distribution of bilateral aid is such that if IDA were to try to rectify inequities, then all IDA funds might have to flow to a very small group of countries—perhaps to only one country. For example, in the period 1968-70, French overseas dependencies received $141-183 per capita in official aid receipts and Australian Trust Territories received $57 per capita, while India received under $2 per capita and Pakistan received under $4 per capita. If an attempt were to be made to use the IDA to bring India or Pakistan up to the level of Papua and New Guinea, not to speak of the French overseas territories, either India or Pakistan alone would absorb the whole of IDA credits several times over.

The conclusion was that IDA should accept as an additional country criterion for the allocation of IDA credits that of securing greater equity in the overall distribution of aid, but this could not be accepted as the overriding criterion. Thus the fact that a country was judged to be receiving unduly low amounts of assistance from other sources would be one, but only one, factor to be weighed in determining its share of IDA financing.

A parallel criterion has also been applied to IBRD loans for some time. That is, in deciding the size of the Bank lending program in a country, an attempt was made to evaluate the total requirement of the country for external capital during a 5-year period as well as to forecast the amounts likely to be forthcoming from other sources. The decision as to the size of Bank lending was related consequently to the remaining gap to be financed.

This section started out with stating that a necessary condition for World Bank Group financing was that it be for an economic project as defined above. In the early history of the Bank this was also an operative country allocation device that was quite restrictive in holding down the volume of Bank lending, particularly to the least developed countries among the LDCs, i.e., the African countries and countries like Nepal and Bolivia. As many of these would in any case have been ruled out or severely limited by the creditworthiness boundary condition, the project condition was not always the operative limitation. However, after IDA came into existence in 1960, the project limitation was often the operative one.

A large part of Bank Group policy since 1960 has been directed to moving this boundary farther out so that country factors rather than the scarcity of acceptable projects in a country would be the ruling criteria in the allocation of Bank Group resources. Among the measures taken during the presidency of George Woods (1962-68) were the setting up of jointly financed programs with the United Nations Educational, Scientific, and Cultural Organization and the Food and Agriculture Organization whereby these two organizations created special staffs, in large part financed by the Bank, to help countries identify and prepare suitable projects in agriculture and education. Regional offices in East and West Africa were set up to help the African countries prepare projects. At the same time, the Bank Group began to work very closely with the newly created UN Development Program (initially called UN Special Fund) in the preinvestment work carried out by the UNDP. This process was continued by the Bank Group President, Robert S. McNamara. Arrangements were worked out in 1970 to secure better cooperation with the other UN specialized agencies, such as the United Nations Industrial Development Organization, the International Labor Organization, and a jointly financed program set up with the World Health Organization. The Bank Group country economic missions with the help of the UNDP and other specialized agencies began to make a survey of the needs for preinvestment work in each country so as to move the project boundary still further out.

Whereas prior to 1960, the rate of lending of the World Bank Group in most countries was primarily restricted by the scarcity of economic projects, the World Bank Group was able to state in September 1965:

… the capability of the developing countries to make productive use of resources has increasd considerably. A preliminary Bank inquiry, carried out country by country and based on the judgment and experience of the Bank’s country specialists and area economists, suggests that the developing countries could effectively use, on the average over the next five years, some $3 billion to $4 billion more of external capital per year than has been provided in the recent past.7

It was as a result of the work just outlined, as well as the general improvement in the capability of the LDCs from their own efforts and the help of bilateral agencies, that it was possible for the lending of the World Bank Group to continue to grow rapidly throughout the years.

There are still, however, some countries where the availability of economic projects is still the ruling constraint. In some cases this is due to the fact that for the kinds of projects that would be economic in the particular country, there is presently a shortage of technical people available in the country or in the aid agencies to work on them. Another related difficulty is that in some countries the obstacles to development (scarcity of natural resources, difficult climate, extremely low level of past development) are so great that the finding and preparing of suitable investment projects is extremely difficult. For most countries today, however, the ruling conditions in the allocation of Bank Group investment are country and not project considerations.

Whatever else may have been accomplished in this article, I hope that I have succeeded in conveying a clearer idea of the complex nature of the decisions involved in allocating resources to the developing countries through multilateral institutions. It is a process to which no magic formula or simple set of decision rules can be applied. No single objective function can be used to specify the goals. On the whole, however, I believe Alexander Pope’s comment would not be inapplicable to the set of decisions that result:

  • “Whoever thinks a faultless piece to see, Thinks what ne’er was, nor is, nor e’er shall be. In every work regard the writer’s End, Since none can compass more than they intend; And if the means be just, the conduct true, Applause, in spite of trivial faults, is due…”

This article is based on a chapter included in “The Grants Economy in International Perspective: Some Lessons for Foreign Aid Policy,” edited by Kenneth E. Boulding, James Horvath, and Martin Pfaff; Belmont, California, Wadsworth Publishers Company, 1972.


Development Assistance: Efforts and Policies of the Development Assistance Committee, 1970 Review, Report by E. M. Martin, Chairman of DAC, Organization for Economic Cooperation and Development, Paris, December 1970, Tables I and IV. The picture would not be complete without a reference to the important, if indirect, contribution made by the IMF. This subject has recently been covered in three articles in Finance and Development: in March 1972, in June 1972, and the final article in the series in this issue.


“Grant equivalent” is defined as the difference between the face value of a loan and the discounted cost to the recipient of the flow of future service repayments (amortization and interest). It has become a convention to use 10 per cent as the discount rate, i.e., it is assumed that 10 per cent is the normal return on capital in the lending countries. Here is an explanation of how this works: assume that normal return on capital is 10 per cent. Mr. A needs a loan and Mr. Banker agrees to lend the money. Mr. A agrees to repay $100 a year from now and Mr. Banker lends him $100 now on the basis that he is not going to charge Mr. A any interest since, after all, Mr. A is Mr. Banker’s favorite nephew. Mr. S, a stranger, also wants to borrow money and he also agrees to repay $100 a year from now. Mr. Banker therefore lends him $90.91 now since that is the present value of $100 a year from now when the interest rate is 10 per cent (that is, 10 per cent of $90.91 is $9.09, and $90.91 plus $9.09 equals $100). In other words, when Mr. Banker’s nephew received the loan of $100 for one year he was getting the equivalent of a regular loan of $90.91 plus a grant equivalent of $9.09.


Although what is said here is based on the experience of the World Bank Group, other multilateral agencies—the Inter-American Development Bank, the Asian Development Bank, the African Development Bank, and the EDF—use similar criteria, as do the various aid-coordination groups such as the India Consortium.


See A. M. Kamarck, “The Appraisal of Country Economic Performance,” Economic Development and Cultural Change, Vol. 18, No. 2, January 1970, pp. 153–165.


The 19 so-called “Part I” countries: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Federal Republic of Germany, Iceland, Italy, Japan, Kuwait, Luxembourg, Netherlands, Norway, South Africa, Sweden, the United Kingdom, and the United States. In addition, in the current replenishment Ireland, Spain, and Yugoslavia, who are Part II members of IDA, are making contributions; Switzerland, who is a non-member, is making a loan to IDA on very favorable terms. The World Bank itself allocates a substantial portion ($110 million from fiscal year 1971) of its profits to IDA.


Commission on International Development, Partners in Development, Praeger Publishers, New York, Washington, London, 1969, p. 230.


World Bank and International Development Association, Annual Report 1964–65, Washington, D. C., p. 62.

Other Resources Citing This Publication