¹ Official bilateral aid is defined as flows of finance at concessional rates of interest and repayment which is not explicitly used for commercial gain or military purpose. It is an aggregation of grants, loans, official export credits, and other long-term capital. It excludes military assistance, private investment, and a number of other means of assistance (such as preferential entry into markets of high-income countries for the exports of low-income countries).

² The analysis of aid now seems to have achieved the status of a distinct field of study, and consequently there exist a large number of textbooks. For some particularly useful ones, see Bhagwati, Jagdish and Eckhaus, Richard S., eds., Foreign Aid (Harmondsworth, England: Penguin 1970)Google Scholar; Little, Ian M. D. and Clifford, J. M., International Aid (London: Allen and Unwin 1965)Google Scholar; Mikesell, Raymond F., Economics of Foreign Aid (London: Weidenfeld and Nicholson 1968);Google ScholarPincus, John, Economic Aid and International Cost Sharing (Baltimore: The Johns Hopkins Press 1965)Google Scholar; Wall, David, The Charity of Nations (New York: Basic Books 1973)Google Scholar; White, John, The Politics of Foreign Aid (London: Bodley Head 1974)Google Scholar.

³ We have explicitly tested a humanitarian model of U.S. aid allocation. The model consists of the variables of per capita gross domestic product (G.D.P.) measuring the level of relative assistance, and two sets of variables measuring welfare needs (per capita G.D.P., per capita calorie consumption, and number of doctors per 100,000 population), and economic performance (size of international liquidity as a percentage of imports, the rate of growth of real per capita G.D.P., and gross domestic fixed capital formation as a percentage of G.D.P.). The hypothesis derived from the model stipulates that there is a negative relationship between the amount of assistance and the levels of welfare needs and economic performance. However, the regressions of per capita U.S. aid receipts onto the six independent variables do not produce a significant equation for any of the years i960–1970. Although these results do not indicate that aid is allocated in an explicitly antihumanitarian manner, they do clearly show that humanitarian criteria do not underlie and explain the distribution of U.S. aid.

⁴ For examples of studies that reject the humanitarian view of aid allocation and suggest a foreign policy basis, see Griffin, Keith B. and Enos, John L., “Foreign Assistance: Objectives and Consequences,” Economic Development and Cultural Change, XVIII (April 1976)Google Scholar; Hayter, Teresa, Aid as Imperialism (Harmondsworth, England: Penguin 1970)Google Scholar; Jalee, Pierre, The Third World in World Economy (New York: Monthly Review Press 1969)Google Scholar; Kaplan, Jacob J., The Challenge of Foreign Aid (New York: Praeger 1976)Google Scholar; Knorr, Klaus, Power and Wealth (New York and London: Macmillan 1973)CrossRefGoogle Scholar; Magdoff, Harry, The Age of Imperialism (New York: Monthly Review Press 1969)Google Scholar; Mason, Edward S., Foreign Aid and Foreign Policy (New York: Harper and Row 1964)Google Scholar; Mende, Tibor, From Aid to Recolonization (London: Harrap 1973)Google Scholar; Morgenthau, Hans J., “Preface to a Political Theory of Foreign Aid,” in Goldwin, Robert A., Why Foreign Aid? (Chicago: Rand McNally 1962)Google Scholar; Nelson, Joan M., Aid, Influence and Foreign Policy (New York: MacMillan 1968)Google Scholar; Payer, Cheryl, The Debt Trap (Harmondsworth, England: Penguin 1974)Google Scholar; Weissman, Steve and ot hers, The Trojan Horse (San Francisco: Ramparts Press 1974)Google Scholar; White (fn. 1).

⁵ There have been arguments that aid does not in fact provide a means of economic assistance. The standard economic models examining the impact of aid on the economic performance of low-income countries have perceived aid as assisting development by virtue of increasing investment, which, with a given incremental capital-output ratio, will then increase growth. See, for example, Chenery, Hollis B. and Strout, Alan, “Foreign Assistance and Economic Development,” American Economic Review, Vol. 56 (September 1966)Google Scholar. The arguments that aid does not provide economic assistance have been made on the grounds that aid reduces savings, creates debt-service problems, and reduces the capital-output ratio. See, for example, Bauer, Peter, Dissent on Development (London: Weidenfeld and Nicholson 1971)Google Scholar; Griffin, Keith B., “Foreign Capital, Domestic Savings, and Economic Development,” Bulletin of the Oxford University Institute of Economics and Statistics, XXXII (May 1970)Google Scholar; Weisskopf, Thomas E., “The Impact of Foreign Capital Inflow on Domestic Savings in Underdeveloped Countries,” Journal of International Economics, 11 (No. 1, 1972)Google Scholar. These criticisms have been well answered by Papaniek, Gustav F., “The Effect of Aid and Other Resource Transfers on Savings and Growth in Less Developed Countries,” Economic Journal, Vol. 82 (September 1972)Google Scholar. Although the critiques are useful in pointing out that the acceleration effect of aid on development is a function of a large number of factors, we can in general accept that aid can provide economic assistance.

⁶ The foreign policy interpretation of the allocation of aid is not new. From the inception of modern aid programs in the Marshall Plan, the capacity of aid to provide economic assistance to the recipient and to satisfy certain foreign policy interests of the donor was recognized. The assistance provided to low-income countries in the early fifties primarily took the form of military aid and was consequently interpreted largely in terms of foreign policy. However, starting in the late fifties, with the establishment of economic assistance programs separate from the military assistance programs, the economic orientation of aid came to the fore. More recently, the foreign policy interpretation has become dominant as it has been realized that economic assistance can be used to promote the foreign policy interests of the donor.

⁷ These observations are substantiated by an examination of work by Nelson and Knorr, who have both made important contributions to the foreign policy view of aid. Nelson (fn. 4) enumerates a series of purposes, allocation criteria, and ends that can be achieved by use of the influence inherent in aid. However, there is little justification or support for the selection of the purposes, criteria, and ends; it is very difficult to distinguish between these, and they are not related in any systematic fashion. Similarly, Knorr (fn. 4) lists a number of rationales, coercive uses, and instrumental purposes of aid. Comparable problems occur in his discussion. There is confusion and overlap between the categories, and little systematic justification for the various elements. There have, however, been some studies of U.S. aid from perspectives that attempt to assess the importance of criteria employed by decision makers. See, for example: Kato, M., “A Model of U.S. Foreign Aid Allocation,” in Mueller, John E., ed., Approaches to Measurement in International Relations (New York: Appleton-Century-Crofts 1969)Google Scholar; Wittkopf, Eugene R., “Western Bilateral Aid Allocations,” Sage International Studies Series, 11 (No. 5, 1972)Google Scholar. Although these studies resemble the present one to the extent that an attempt is made to identify the criteria underlying the distribution of aid by regressing aid onto an array of independent variables, they differ in that: (1) a smaller population is used; (2) a more restrictive time period is used; (3) a smaller and more restrictive array of independent variables is used; and (4) little attempt is made to evaluate the main findings in more general terms. While these studies are more precise than the work of writers such as Nelson and Knorr, they are less comprehensive in their scope and evaluation.

⁸ The failure to develop a systematic foreign policy model of aid is largely a function of the neglect, except on the part of Marxists, of the study of international political economy. Though economists have widely recognized the foreign policy context of aid, they have considered an examination of this aspect of aid to be outside their sphere of competence. Political developmentalists have focused mainly on the domestic political structures of low-income countries and have to a large extent ignored the external setting of these countries. Aid does not appear, for example, in the index of many of the writings of major political developmentalists. Any reference that is made to aid usually takes the form of simple borrowing from economists. Foreign policy analysts have been preoccupied with classifying foreign policy behavior and have tended to ignore the political aspects of economic behavior.

⁹ The concept of commitment has been extensively examined in the social sciences. For a general theoretical discussion, see Roby, Thornton B., “Commitment,” Behavioral Science, v (July 1960)Google Scholar. In international relations, commitment has been most frequently employed in the literature dealing with strategic and alliance policy. For general discussions see, for example, George, Alexander L. and Smoke, Richard, Deterrence in American Foreign Policy (New York: Columbia University Press 1974)Google Scholar; Schelling, Thomas C., Arms and Influence (New Haven: Yale University Press 1966)Google Scholar; Schelling, , The Strategy of Conflict (New York: Oxford University Press 1963)Google Scholar; Weinstein, Franklin B., “The Concept of Commitment in International Relations,” Journal of Conflict Resolution, XIII (March 1969)Google Scholar.

¹⁰ Dependency has become a major concept in the literature dealing with relations between high- and low-income countries. It is frequently examined in the context of such related concepts as domination, imperialism, and neocolonialism. This literature can be confusing and polemical. In this paper we have endeavored to avoid entering into many of the debates associated with dependency, and to provide a simple and unambiguous definition. For examples of some general works dealing with dependency, see Baran, Paul A., The Political Economy of Growth (New York: Monthly Review Press 1968)Google Scholar; Brown, Michael Barratt, After Imperialism (New York: Humanities Press 1970)Google Scholar; Cohen, Benjamin J., The Question of Imperialism (London: MacMillan 1974)Google Scholar; Fann, K. T. and Hodges, Donald C., eds., Readings in U.S. Imperialism (Boston: Porter Sargent 1971)Google Scholar; Frank, Andre Gunder, Capitalism and Underdevelopment in Latin America (New York: Monthly Review Press 1969)Google Scholar; Hayter (fn. 4); Magdoff (fn. 4); Rhodes, Robert I., ed., Imperialism and Underdevelopment: A Reader (New York: Monthly Review Press 1970)Google Scholar; Rosen, Steven J. and Kurth, James R., eds., Testing Theories of Economic Imperialism (Lexington, Mass.: D. C. Heath 1974)Google Scholar.

¹¹ Three qualifications to this analysis must be noted. First, commitment and dependency can be established through a variety of other means such as trade, private investment, and military assistance. Second, the provision of aid does not necessarily guarantee a successful commitment or create dependency. For example, recipients can play one donor off against another, undermining the rationale of commitment; or recipients can refuse aid or repayments, thereby eliminating the potential for dependency. Third, even if commitment and dependency are established, the donor may fail to derive any benefits from the utilities contingent on commitment and dependency. For example, intervention by a hostile party may occur despite the existence of commitments; or in the case of dependency, the benefits accruing to the donor may become so important that it is no longer in a position to terminate the relationship, and at this point the relationship becomes characterized by interdependence rather than dependency.

¹² The publication that has been used to provide gross aid figures is The Geographical Distribution of Financial Flows to Less Developed Countries (Paris: O.E.C.D., D.A.C.). This has been published for the following years: 1960–64, 1965, 1966–67, 1968, 1969, 1970.

¹³ Three countries, A, B, and C, receive different amounts of gross aid (giving them an absolute commitment ranking running from A to C), such that receipts per capita aid are identical. However, once we take per capita income into account, then C, the recipient of the smallest amount of aid, receives the greatest relative commitment.

¹⁴ Asset dependency could be operationalized in a variety of ways. For example, aid could be expressed as a percentage of imports or domestic savings. Although these measures have similar connotations to the index we use, we prefer G.D.P. because it is more comprehensive and stable.

¹⁵ The sources for the data in this study are: United Nations Statistical Yearbook, 1960–1972; United Nations Yearbook of International Trade Statistics, 1961–1972; United Nations Yearbook of National Account Statistics, 1964–1972; International Financial Statistics, 1960–1970 (Washington: International Monetary Fund)Google Scholar; Balance of Payments Yearbook, 1960–1970 (Washington: International Monetary Fund)Google Scholar; Statesman's Yearbook, 1960–1970 (London: MacMillan)Google Scholar; Europa Yearbook, 1958–1970 (London: Europa Publications)Google Scholar; Stfbbins, Richard P. and Amoia, Alma, eds., Political Handbook and Atlas of the World (New York: Simon and Schuster 1970)Google Scholar; Whitaker's Almanac, 1959–1970 (London: Whitaker's Publications);Google ScholarKeesings Contemporary Archives, 1959–1970 (London: Keesings Publications)Google Scholar; The Military Balance, 1960–1970 (London: Institute for Strategic Studies)Google Scholar; Stockholm International Peace Research Institute, Yearbook of World Armaments and Disarmament, 1968–1972 (Stockholm: Almqvist)Google Scholar; Stockholm International Peace Research Institute, Arms Trade with the Third World (Harmondsworth, England: Penguin 1973)Google Scholar; Dupuy, Trevor N. and others, Almanac of World Military Power (Dunn Loring, Virginia: Trevor N. Dupuy and Associates 1970)Google Scholar; Sellers, Richard C., ed., Reference Handbook of Armed Forces of the World (3rd ed.; New York: Praeger 1971).Google Scholar

¹⁶ We are, of course, faced with the problem of a large number of variables. This problem is unavoidable as there are no a priori grounds for anticipating which are the most important. Furthermore, testing the model in separate parts could be seriously misleading; variables, which may be important in simple terms, may well be of trivial importance once other variables are included; alternatively, they may take on a very different importance. It must be emphasized that we are not interested in making a descriptive profile of the characteristics of countries receiving differing amounts of aid, but in identifying the criteria, and the structural arrangement of the criteria, that dictate the pattern of the distribution of aid. This exercise requires a comparative evaluation of potential criteria. We do not, however, wish to retain all the independent variables, as the interpretation would be far too confusing; furthermore, it is conceivable that many may be of little importance. It is also possible, as the regression equation gets longer, to find variables that appear to be of some importance, but only because they are qualifying each other. There are, of course, a number of rules governing inclusion of variables such as size of beta value, level of significance, or magnitude of change in R². There are also certain problems associated with each of these rules; in practice, it is therefore impossible to set a hard and fast rule governing all the regressions. In general, since we are not interested in economy of estimation, we have used low values in order to produce as complex an equation as possible while remaining within the bounds of manageability.

¹⁷ Any multiple regression analysis in the social sciences must confront the problem of multicollinearity. In this context, the use of the standardized partial regression coefficient is very useful. The beta value can be used to partition explained variance into independent and joint effects. (In the three-variable case, R²_1.23 = B²₂ + B²₃ + 2B₂B³R_23.) In the situation where independent variables are uncorrelated, the last term disappears. This is of course unusual in the social sciences. However, this last term is extremely important for a variety of reasons, one of which is that it constantly calls attention to the presence of multicollinearity. In general, we would argue that we do not have any major difficulties. Prior to the analysis, a number of variables with high correlations—as a consequence of one variable being a substitute for another—have been rejected. In other cases, where we wished to retain two highly correlated variables in the analysis, we have standardized and added them (where it was meaningful to do so). Finally, we tested our independent variables by factor analysis. If there was a high degree of interrelation between the independent variables, then a factor analysis would produce a small number of factors with high communality scores. Factor analyses, however, did not produce these results.

¹⁸ As countries become independent during the period examined, the population increases. A number of countries are excluded from each year because data for them are not available. The total number of potential recipients of aid and the actual number included in the analysis (the latter in parenthesis), from i960 to 1970 are: 58 (52), 74 (68), 76 (70), 82 (75), 84 (77), 87 (80), 90 (82), 93 (83). 94 (83), 96 (84), 96 (84).

¹⁹ The authors will supply the results of the regressions of absolute and relative commitment and asset dependency upon receipt of requests accompanied by a self addressed envelope and postage.

²⁰ On the basis of the positive relationships among the development variables, a development syndrome can be identified. Given the existence of a development syndrome, two possible sets of results are compatible with the development hypothesis. First, the hypothesis can be confirmed if a number of development variables appear with positive signs, indicating those aspects of the development syndrome related to relative commitment. Second, the hypothesis can be confirmed if a dominant variable emerges that is capable of explaining all the variance shared by relative commitment and the development syndrome.

²¹ This point is dealt with more extensively in Appendix 3.

²² Although the political interpretation of imperialism has not been as well articulated as the economic, it does have a long pedigree. See, for example, Carr, Edward H., The Twenty Years' Crisis (London: MacMillan 1946)Google Scholar; Morgenthau, Hans J., Politics Among Nations (5th ed.; New York: Alfred A. Knopf 1973)Google Scholar; Schuman, Frederick L., International Politics (New York: McGraw-Hill 1935)Google Scholar; Schwarzenberger, Georg, Power Politics (3rd ed.; London: Stephens 1964)Google Scholar; Waltz, Kenneth N., Man, The State, and War (New York: Columbia University Press 1959)Google Scholar. For a recent analysis of the realist position, see Rothstein, Robert L., “On the Costs of Realism,” Political Science Quarterly, Vol. 87 (September 1972)CrossRefGoogle Scholar. For some recent literature calling attention to the importance of power capabilities, see Cohen (fn. 10); Ferris, Wayne H., The Power Capabilities of Nation-States (Lexington, Mass.: Lexington Books 1973)Google Scholar; Wallace, Michael D., War and Ran Among Nations (Lexington, Mass.; D. C. Heath 1973)Google Scholar.