¹ The New York Times, October 27, 1971, p. 16 Google ScholarPubMed.

² Lack of aid data on a dyadic basis has required the aggregation of the Soviet Union and other Communist nations into a single actor, referred to some-what unhappily as a “bloc.” For purposes of the 1962–1964 aid data, the Soviet “bloc” comprises the Soviet Union, Bulgaria, Mainland China, Czechoslovakia, East Germany, Hungary, Poland, and Rumania. With the exception of Mainland China, the same states comprise the Soviet “bloc” for purposes of the 1965–1967 aid data.

³ Austria, Belgium, Canada, Denmark, France, Italy, Japan, the Netherlands, Norway, Sweden, and the United Kingdom. Australia, though not a DAC member until 1966, is included in the 1965–1967 analyses.

Although the DAC donors and the Soviet “bloc” nations together contribute the vast bulk of bilateral economic aid flows in the international system, several other states do maintain modestly sized bilateral programs, many of them in the area of technical assistance. Finland and New Zealand are among the developed nations maintaining such programs. Spain, Kuwait, Yugoslavia, the United Arab Republic, Nationalist China, Israel, India, Brazil, Mexico, and Venezuela are among the developing nations that reportedly maintain some such program, often in conjunction with a multilateral effort. See Organization for Economic Cooperation and Development, The Flow of Financial Resources to Less-Developed Countries 1961–1965 (Paris: O.E.C.D., 1967), pp. 59–61 Google Scholar; Organization for Economic Cooperation and Development, Resources for the Developing World (Paris: O.E.C.D., 1970), pp. 289–91Google ScholarPubMed; and Organization for Economic Cooperation and Development, Development Assistance: Efforts and Policies of the Members of the Development Assistance Committee, 1969 Review (Paris: O.E.C.D., 1969), pp. 60–63 Google Scholar. Unfortunately, the data necessary for a systematic analysis of the aid programs of these donor countries comparable to that undertaken for the donors included in the study are not available. In fact, even in the case of the Soviet bloc, as will be noted below, the data are not comparable to those available for the DAC donors, thus leading to the presentation of two separate sets of results. However, the interest usually attached to Soviet aid programs seems to warrant its (their) inclusion.

⁴ See, e.g., Alker, Hayward R. Jr. and Russett, Bruce M., World Politics in the General Assembly (New Haven, Conn.: Yale University Press, 1965), pp. 226–27Google Scholar; Alker, Hayward R. Jr., “Dimensions of Conflict in the General Assembly,” in Approaches to Measurement in International Relations, ed. Mueller, John E. (New York: Appleton-Century-Crofts, 1969), pp. 163–65Google Scholar; and Clark, John F., O'Leary, Michael K., and Wittkopf, Eugene R., “National Attributes Associated with Dimensions of Support for the United Nations,” International Organization, 25 (Winter, 1971), 1–25 CrossRefGoogle Scholar.

⁵ Two examples of studies that use UN votes to examine aid allocations are Kato, Masakatsu, “A Model of U.S. Foreign Aid Allocation: An Application of a Rational Decision-Making Scheme,” in Mueller, , Approaches to Measurement in International Relations, pp. 204–10Google Scholar; and Joshua, Wynfred and Gibert, Stephen P., Arms for the Third World (Baltimore: The Johns Hopkins Press, 1969), pp. 140–43Google Scholar.

⁶ Keohane, Robert O., “Political Influence in the General Assembly,” International Conciliation No. 557 (March, 1966)Google Scholar.

⁷ Keohane, p. 19.

⁸ Keohane, p. 19.

⁹ Keohane, pp. 17–18. See also Kay, David A., “Instruments of Influence in the United Nations Political Process,” in The United Nations Political System, ed. Kay, David A. (New York: John Wiley & Sons, 1967), p. 105 Google Scholar.

Keohane also suggests that “most decisions of the General Assembly … are not important enough to United States interests for it to threaten convincingly to alter the level of aid or support that it is giving, or to exercise other forms of pressure … for the sake of gaining a few more votes in the United Nations” (Keohane, p. 20). And he argues that the changing composition of the United Nations has probably made bilateral pressure less important that it was, say, in the 1950s (pp. 19–20).

Still, this caveat does not obviate the likelihood that foreign aid is a significant factor conditioning states' voting behavior, particularly where aid represents part of a general dependency relationship between a developing nation and an aid donor. In discussing the “appeal for good relations” as a form of bilateral pressure, for example, Keohane observes that the threat implicit in this appeal “… may become a painfully acute form of pressure if it is exercised by greater powers and directed toward states that depend on them for aid or support” (p. 18).

Jack C. Piano and Robert E. Riggs also observe that most UN votes are not sufficiently important to warrant a threat to alter aid levels. See Forging World Order (New York: Macmillan, 1967), p. 154 Google Scholar. In an earlier work, however, Riggs discusses the difficulty of separating specific uses of aid to secure compliance in the UN from the general relationship likely to hold between these two variables; see Riggs, Robert E., Politics in the United Nations: A Study of United States Influence in the General Assembly (Urbana: University of Illinois Press, 1958), pp. 38–39 Google Scholar. The important point, of course, is that whether aid is or is not related to votes remains an empirical question.

¹⁰ Mason, Edward S., Foreign Aid and Foreign Policy (New York: Harper and Row, 1964), p. 33 Google Scholar.

¹¹ Black, Lloyd D., The Strategy of Foreign Aid (Princeton, N.J.: D. Van Nostrand, 1968), p. 19 Google Scholar.

¹² Wilcox, Francis O., “The Nonaligned States and the United Nations,” in Neutralism and Nonalignment, ed. Martin, Laurence W. (New York: Praeger, 1962), pp. 147–48Google Scholar.

¹³ Mason, p. 42.

¹⁴ Wolf, Charles Jr., “Economic Aid Reconsidered,” in The United States and the Developing Economies, ed. Ranis, Gustav (New York: Norton, 1964), p. 52 Google Scholar.

¹⁵ Kaplan, Jacob J., The Challenge of Foreign Aid (New York: Praeger, 1967), p. 248 Google Scholar. See also Kaplan, p. 73.

¹⁶ Westwood, Andrew F., Foreign Aid in a Foreign Policy Framework (Washington, D.C.: The Brookings Institution, 1966), p. 105 Google Scholar.

¹⁷ Westwood, p. 105.

¹⁸ Westwood, p. 105.

¹⁹ Cf. Alker, , “Dimensions of Conflict in General Assembly,” pp. 163–64Google Scholar. Much additional work has been undertaken to determine whether a “critical vote” or “issue-area” approach might be a preferable way to address the question examined in this essay. Generally speaking, the results of these analyses suggest that neither of these alternatives is preferable to the generalized approach that has been chosen. For example, a contingency table analysis of developing nations' voting behavior on the China issue and their U.S. aid receipts during 1961–1967 indicated that aid allocations and voting behavior are quite probably independent of one another. See Wittkopf, Eugene R. and Green, James, “Empirical Perspectives on the China Vote and U.S. Foreign Aid, 1961–1967” (paper, University of Florida, January, 1972)Google Scholar. Similarly, work has been done using only those votes on which the United States and the Soviet Union did not agree. The results obtained differ only marginally from those displayed later, which employ all Assembly roll calls. Finally, an effort was made to replicate Bruce Russett's factor analysis of UN votes in 1963 in which he found five “super issues” in the General Assembly; see Russett, Bruce M., International Regions and the International System (Chicago: Rand McNally, 1967), pp. 63–67 Google Scholar. VRA scores (discussed later) were calculated for those votes that loaded at .5 or greater on each of the three most important dimensions in the 1963 Assembly (cold war, 20 votes; intervention in Africa, 17 votes; supranationalism, 18 votes). The VRA's were then correlated with the 1962–1964 U.S. aid RA's (also discussed later). The results are given below (Table A) . Although some interesting differences across the various dimensions are apparent, the important point is that the conclusions that can be drawn about the association between aid and votes (or lack of association) do not differ dramatically from those that can be drawn on the basis of looking at all roll-call votes. (See Table 3.)

²⁰ West Germany, Portugal, and Switzerland are members of DAC in addition to those nations cited earlier. Aid flows from Germany and Switzerland have been included in the aid matrices analyzed using the transaction-flow model. Portugal is excluded since it extended aid only to its overseas territories.

²¹ These numbers as well as the nations listed in Table 1, include such non-UN members as South Korea, Muscat and Oman, South Vietnam, and Western Samoa. Although these nations are, by definition, excluded from the correlation analyses, they are included here since their aid receipts from the DAC donors have been included in the aid matrices analyzed using the transaction-flow model.

²² Again, there is nothing magic in the number eight; it is simply the minimum number of developing nations included for which data on Soviet bloc assistance are available. But with an N less than eight, the technique of fitting a line to data points is either inappropriate, or the number of observations is so small as to severely restrict the generalizability of the empirical findings.

²³ It is because of the absence of an a priori rationale for assuming which temporal sequence is more likely to hold that correlation rather than regression analysis has been chosen as the analytical technique.

²⁴ See Wittkopf, Eugene R., “The Distribution of Foreign Aid in Comparative Perspective: An Empirical Study of the Flow of Foreign Economic Assistance, 1961–1967” (Ph.D. dissertation, Syracuse University, July, 1971), pp. 38–41 Google Scholar.

²⁵ See, e.g., Russett, , International Regions, pp. 59–93 Google Scholar.

²⁶ Geographical Distribution of Financial Flows to Less Developed Countries 1960–1964 (Paris: O.E.C.D., 1966)Google Scholar; Geographical Distribution of Financial Flows to Less Developed Countries 1965 (Paris: O.E.C.D., 1967)Google Scholar; and Geographical Distribution of Financial Flows to Less Developed Countries 1966–1967 (Paris: O.E.C.D., 1969)Google Scholar.

²⁷ The OECD uses the term “net total official financial flows” to describe these data. The flows are defined as the sum of grants, including grants for technical cooperation; intergovernmental reparation and indemnification payments, a type of disbursement remitted only by Italy and Japan throughout the 1962–1967 time period, and by West Germany until 1965; net official loans; loans repayable in recipients' currencies (net); and transfers of resources through sales for recipients' currencies (principally food aid distributed under the aegis of U.S. Public Law 480). O.E.C.D., Geographical Distribution 1960–1964, p. XII Google Scholar.

It will be noted that the aid data do not distinguish between various budgetary categories and administrative agencies in donor countries. In effect, therefore, it is assumed that the aid data comprise the foreign policy output of an aggregate decisional unit in each donor country. This is admittedly a very powerful assumption (as Allison, Graham T. has clearly demonstrated in his Essence of Decision [Boston: Little, Brown, 1971]Google Scholar) that should perhaps be subjected to additional research. But use of the OECD-DAC data does allow one to move beyond what Baldwin has identified as the tendency to define foreign aid in terms of American institutions. See Baldwin, David A., “Analytical Notes on Foreign Aid and Politics,” Background, 10 (Winter, 1966), 75–76 CrossRefGoogle Scholar.

²⁸ International Flow of Long-term Capital and Official Donations, 1961–1965 (New York: U.N., 1966), pp. 22–23 Google Scholar; and The External Financing of Economic Development (New York: U.N., 1969), pp. 44–45 Google ScholarPubMed.

²⁹ Director of Intelligence and Research, Communist Governments and Developing Nations: Aid and Trade in 1965 (Research Memorandum RSB-50, June 17, 1966), p. 2 Google Scholar; Director of Intelligence and Research, Communist Governments and Developing Nations: Aid and Trade in 1967 (Research Memorandum RSE-120, August 14, 1968), pp. 2–3 Google Scholar; and Director of Intelligence and Research, Communist Governments and Developing Nations: Aid and Trade in 1968 (Research Memorandum RSE-65, September 5, 1969), pp. 2–3 Google Scholar. UN data have been used for the first three years since State Department estimates of Soviet bloc data are not available on an annual basis for all of the years 1962–1964. State Department estimates have been used for the later three-year period rather than UN estimates on the assumption that the former are more accurate than the latter. UN data apparently were collected from unofficial sources such as press releases as well as official Soviet and East European publications. The State Department, on the other hand, presumably received its information directly from recipient nations. It is assumed, therefore, that the margin of error is less in the State Department than in the UN data. See also Vassilev, Vassil, Policy in the Soviet Bloc on Aid to Developing Countries (Paris: Development Centre of the Organization for Economic Cooperation and Development, 1969), pp. 61–66 Google Scholar.

³⁰ Savage, I. Richard and Deutsch, Karl W., “A Statistical Model of the Gross Analysis of Transaction Flows,” Econometrica, 28 (July, 1960), 551–72CrossRefGoogle Scholar.

³¹ Goodman, Leo A., “Statistical Methods for the Preliminary Analysis of Transaction Flows,” Econometrica, 31 (January-April, 1963), 197–208 CrossRefGoogle Scholar; and Goodman, Leo A., “A Short Computer for the Analysis of Transaction Flows,” Behavioral Science, 9 (April, 1964), 176–86CrossRefGoogle Scholar.

³² Brams, Steven J., “A Generalized Computer Program for the Analysis of Transaction Flows,” Behavioral Science, 10 (October, 1965), 487–88Google Scholar.

³³ The more commonly accepted term for transaction-flow RA's is “relative acceptance” indices. “Relative allocation” has been chosen for the aid RA's, however, since this seems a more accurate reflection of the nature of foreign aid monies treated as transaction flows given the absence of a two-way flow of funds.

³⁴ The reason for excluding negative aid flows from the analysis should now be apparent. If both A _ij and E _ij are negative, and the former is greater than the latter, RA _ij can be a positive number, indicating aid flows greater than an indifferent amount, despite the fact that it is based on a negative A _ij . Intuitively, this is, of course, an unreasonable result. More importantly, the RA for a second donor, k, assuming k also allocates aid to recipient j, can be markedly inflated due to the negative flow from i to j, thereby causing RA _kj to have an extreme effect on the correlation co-efficient used to assess the association between donor k's aid allocations and developing nations' voting agreements with k.

³⁵ For a more complete discussion of the transaction-flow model, see Brams, Steven J., “Transaction Flows in the International System,” American Political Science Review, 60 (December, 1966), 882–86CrossRefGoogle Scholar; Brams, Steven J., “Trade in the North Atlantic Area: An Approach to the Analysis of Transformations in a System,” Peace Research Society Papers, VI, Vienna Conference (1966), pp. 144–48Google Scholar; and E. Wittkopf, “The Concentration and Concordance of Foreign Aid Allocations.” My thanks to Professor Brams for the considerable guidance and encouragement he lent to the application of the transaction-flow model and his computer program INDIFF to the data on foreign aid flows.

³⁶ The principal difference between the original Savage-Deutsch model and the Goodman-Brams modification is that the former excludes from the calculation of E _ij 's only a country's transactions with itself, while the latter excludes all pairs not actually linked by known transaction flows. In the case of the foreign aid data, three of the four logical portions of the square foreign aid matrix are, by definition, blank—recipient to recipient, recipient to donor, and donor to donor aid flows. With so many zero entries, it was necessary to assign an arbitrarily small value (equal to $.01) to one cell within each row and column that would otherwise have been entirely blank in order to ensure calculation of E _ij 's for the actual non-blank cells.

³⁷ Todd uses this measure of voting agreements in his study of voting in the Security Council. Todd, James E., “An Analysis of Security Council Voting Behavior,” Western Political Quarterly, 22 (March, 1969), 61–78 CrossRefGoogle Scholar.

³⁸ Brams, Steven J. and O'Leary, Michael K., “An Axiomatic Model of Voting Bodies,” American Political Science Review, 64 (June, 1970), 449–70CrossRefGoogle Scholar.

³⁹ Brams and O'Leary, p. 463.

⁴⁰ Brams and O'Leary, pp. 452–53. The Brams-O'Leary model assumes that the voting behavior of legislative members is independent from one roll call to the next. It can be demonstrated, however, that the value of the expected term in the calculation of VRA is not dependent on the assumption of independence. See Mayer, Lawrence S., “A Note on ‘An Axiomatic Model of Voting Bodies’,” American Political Science Review, 65 (September, 1971), 764–65CrossRefGoogle Scholar, and the response in Brams, Steven J. and O'Leary, Michael K., “Comment on Mayer's ‘A Note on “An Axiomatic Model of Voting Bodies”’,” American Political Science Review, 65 (September, 1971), 766–67CrossRefGoogle Scholar.

⁴¹ Conceptually, the VRA's are virtually identical with the aid RA's. The two indices differ only in the calculation of the expected term; as indicated, the expected values for the aid RA's depend on the row and column marginal totals in the aid matrix, while the expected values for the VRA's depend upon the probabilities of agreement between a randomly-selected pair of Assembly members across a set of roll-call votes.

⁴² Substantively, this means that the patterns of agreements between particular pairs of states are not affected by the overall divisions in the Assembly. See Brams, and O'Leary, , “An Axiomatic Model of Voting Bodies,” p. 463 Google Scholar.

⁴³ Brams and O'Leary, p. 464.

⁴⁴ As a practical matter it would make virtually no difference if subset I were denned to include only the Soviet Union, since the U.S.S.R. and its Eastern European allies exhibit almost identical behavior in their UN voting patterns.

⁴⁵ Developing nations making net repayments of aid (i.e., negative aid recipients) to a particular donor, as well as those that missed more than 40 per cent of the UN roll calls, were subtracted from these numbers for purposes of calculating group VRA's for that donor.

⁴⁶ The appropriateness of significance tests in cases where the sample and the universe are the same, typically the case in cross-national research, remains an unresolved debate among political science and international relations scholars. The use of such tests here is a conservative one and follows the suggestions of Winch and Campbell, who argue that in the case of nonsampled data, tests of significance are appropriate to an analysis of the (null) hypothesis that the magnitude of the relationship between two variables is due to chance. ( Winch, Robert F. and Campbell, Donald T., “Proof? No. Evidence? Yes. The Significance of Tests of Significance,” The American Sociologist, 4 [May, 1969], 143 Google Scholar). Cf. Russett, Bruce M. et al., World Handbook of Political and Social Indicators (New Haven: Yale University Press, 1964), p. 263 Google Scholar; and Singer, J. David and Wallace, Michael, “Intergovernmental Organization and the Preservation of Peace, 1816–1964: Some Bivariate Relationships,” International Organization, 24 (Summer, 1970), 533 CrossRefGoogle Scholar. Significance tests are also particularly useful in situations where there are large variations in the number of observations, as is the case here, which otherwise would make comparisons across foreign policy actors very difficult.

⁴⁷ Because the percentage of pairwise voting agreements and the VRA's are very highly correlated, as noted earlier, the change in the results for the U.S. from those shown in Table 2 can be attributed largely to the transaction-flow model, which, in effect, “transforms” the aid data by adjusting for the size of aid recipients assumed to be reflected in the relative magnitudes of foreign aid transactions between donors and recipients. The magnitude of foreign aid transactions may also indicate the importance of recipients to donors, but this will be reflected in the aid RA's only if the foreign aid transactions for a particular donor are greater than what would be expected on the basis of the proportionate distribution of transactions of all aid donors included in the system analyzed. Essentially, then, the indifference model is an alternative to the more usual ways of “normalizing” foreign aid data, such as by the population size, GNP, or imports of developing nations. Interestingly, in the analysis of U.S. aid and developing nations' voting behavior on the China issue cited earlier (footnote 19), an aid/imports ratio was used as one of the foreign aid variables, but the results were virtually identical to those obtained when only the absolute amounts of aid dollars were used.

⁴⁸ These generally low levels of shared variance are perhaps not surprising given the large number of both political and development considerations presumed to undergrid U.S. aid allocations. For a discussion and empirical analysis of some of these additional considerations, see Wittkopf, E., “The Concentration and Concordance of Foreign Aid Allocations,” and Wittkopf, E., Western Bilateral Aid Allocations: A Comparative Study of Recipient State Attributes and Aid Received (Beverly Hills, Calif.: Sage Publications, 1972)Google Scholar.

⁴⁹ Nations are identified according to their abbreviated names given in Table 1. The reader is cautioned to note that the axes in the two scattergrams are not identical. This is also true for Figure 3 shown on p. 884.

⁵⁰ Estimates of Soviet bloc aid commitments in millions in U.S. dollars to European and Latin American developing nations are as follows:

⁵¹ For a discussion of decolonization issues, with special reference to African states, see Kay, David A., The New Nations in the United Nations, 1960–1967 (New York: Columbia University Press, 1970), pp. 80–85 Google Scholar.

⁵² Kay, p. 84.

⁵³ In a separate analysis of the two basic component categories of foreign aid, grants and gross official loans, it was found that there is a positive (but not significant) relationship between France's VRA's and aid RA's when the latter are calculated only on the basis of grants, and a negative relationship similar to those shown in Tables 4 and 5 when the aid RA's are calculated on the basis of gross loans. The reason lies in the fact that France chose not to allocate grants to its European and Latin American aid recipients, thus excluding those nations most prone to agree with her in the UN as the above results indicate. For details, see Wittkopf, E., “The Distribution of Foreign Aid in Comparative Perspective,” pp. 124–33Google Scholar.

⁵⁴ Mason, p. 33.

⁵⁵ As Benjamin J. Cohen has put it in discussing the rationale for a U.S. development aid program, “In at least one fundamental respect … the promotion of the growth of poor countries is in our national interest: economic concessions on our part serve as a kind of insurance against belligerency on the part of the less developed countries” (“The Third World: The Issues,” in American Foreign Economic Policy, ed. Cohen, Benjamin C., [New York: Harper and Row, 1968], p. 339)Google Scholar.

⁵⁶ This conclusion must, of course, be used advisedly. From the perspective of aid donors, what is being said is if foreign aid donors are motivated to allocate their aid on the basis of how developing nations vote in the UN, then we would expect to find patterns of association different from those found for most donors. The United Kingdom is an illustrative case similar to France. In 1964, Kenya, Tanzania, and Uganda had among the highest aid RA's of any British aid recipient. These states were also among those least prone to agree with Britain in the 1963 General Assembly. The result is an inverse association. If, however, a primary consideration was how these nations voted in the UN vis-à-vis the United Kingdom, then we would expect that Britain would not have allocated as much of its aid proportionately to these nations as it in fact did. Although it is generally the case that developing nations agree with most aid donors (notably DAC donors) relatively infrequently (i.e., virtually all DAC donors' VRA's are negative), clearly some developing nations among those that Britain chose to aid in 1964 agreed relatively more often with the United Kingdom than either Kenya, Tanzania, or Uganda. Accordingly, our hypothesized relationship between aid and votes suggests that these other states should have received a relatively greater amount of British aid than they did.

⁵⁷ Because the Soviet bloc data are commitments rather than disbursements, this is the one donor for which group VRA's could show more than minimal fluctuations from one year to the next if they were calculated for all six years included in the study.

⁵⁸ Because the Soviet bloc and DAC data do differ in kind, it seems reasonable to tentatively regard the absence of comparable results for the United States and the Soviet bloc (a surrogate for the Soviet Union) as at least partially attributable to the absence of comparable data.

⁵⁹ The utility of examining intraorganizational processes is demonstrated in Alger, Chadwick F., “Interaction in a Committee of the United Nations General Assembly,” in Quantitative International Politics, ed. Singer, J. David (New York: The Free Press, 1968), pp. 51–84 Google Scholar; and Weigert, Kathleen Maas and Riggs, Robert E., “Africa and United Nations Elections: An Aggregate Data Analysis,” International Organization, 23 (Winter, 1969), 1–19 CrossRefGoogle Scholar.

⁶⁰ The potential impact of these factors on foreign policy is discussed in Rosenau, James N., “Pre-theories and Theories of Foreign Policy,” in Approaches to Comparative and International Politics, ed. Farrell, R. Barry (Evanston, Ill.: Northwestern University Press, 1966), pp. 27–92 Google Scholar.

⁶¹ Kaplan, p. 248.

⁶² Westwood, p. 105.