Skip to main content

Africa and United Nations Elections: An Aggregate Data Analysis


In the United Nations as in other political organizations elective offices are eagerly sought as badges of prestige and levers of influence. Some of the most hotly contested political battles in the United Nations have centred around elections to the prestigious and influential nonpermanent seats on the Security Council. Contests have sometimes run to many ballots before any country emerged with the required two-thirds majority, and deadlock has more than once forced a compromise in which the two-year term was divided between two contenders. Indicative of the feeling sometimes aroused was Indonesia's withdrawal from the United Nations in 1965 which, although prompted by other motives as well, was timed to serve as a protest against the seating of Malaysia on the Security Council. Contests for other positions are generally less spirited, but there are invariably more office seekers than offices.

Hide All

1 Prior to the enlargement of the Security Council from eleven to fifteen members the two-year term was split five times: Yugoslavia and the Philippines, 1956–1957 term; Poland and Turkey, 1960–1961; Liberia and Ireland, 1961–1962; Rumania and the Philippines, 1962–1963; and Czechoslovakia and Malaysia, 1964–1965. An agreement was made to divide the 1965–1966 term between Jordan and Mali if the Security Council was not enlarged. With the enlargement of the Council from eleven to fifteen members, however, each served a full two-year term.

2 E.g., Bailey's Sydney D. chapter on “Elections and Appointments” in The General Assembly of the United Nations: A Study of Procedure and Practice (rev. ed.; New York: Frederick A. Praeger, 1964); Gregg Robert W., “The Economic and Social Council: Politics of Membership,” Western Political Quarterly, 03 1963 (Vol. 16, No. 1), pp. 109132; Padelford Norman J., “Politics and the Future of ECOSOC,” International Organization, Autumn 1961 (Vol. 15, No. 4), pp. 564580; and Padelford Norman J., “Politics and Change in the Security Council,” International Organization, Summer 1960 (Vol. 14, No. 3), pp. 381401. See also Padelford's two dittoed monographs, “Elections in the UN General Assembly” (dealing with Assembly offices) and “Politics in United Nations Elections” (dealing with election of members of the Security Council, Economic and Social Council, and Trusteeship Council), both prepared for the Center for International Studies, Massachusetts Institute of Technology, Cambridge, Massachusetts, 1959.

3 The geographical distribution groups are: 1) Asia and Africa; 2) Eastern Europe; 3) Latin America; and 4) Western Europe and others (the others being Australia, Canada, and New Zealand). A recent article by Catherine Senf Manno suggests that the geographical distribution principle has in some instances been modified or replaced by a practice of “selectively loading” some UN bodies to reflect in a “vaguely defined” way the differences between states in “‘capacities,’ ‘responsibilities,’ ‘interests,’ or ‘willingness to meet Charter obligations.’” Her study, however, relates primarily to sub organs of the General Assembly in which offices are filled by appointment rather than by election. Manno Catherine Senf, “Problems and Trends in the Composition of Non plenary UN Organs,” International Organization, Winter 1965 (Vol. 19, No. 1), p. 38.

4 Singer Marshall R. and Sensenig Barton III, “Elections within the United Nations: An Experimental Study Utilizing Statistical Analysis,” International Organization, Autumn 1963 (Vol. 17, No. 4), pp. 901925. The GNP indicator was assumed to be an index of national power; the voting index was intended to reflect closeness of alignment with the United States in the Cold War; and the aid figure was selected as a measure of economic ties with the United States.

5 Gregg Robert W., “The Latin American Bloc in United Nations Elections,” Southwestern Social Science Quarterly, 09 1965 (Vol. 46, No. 2), pp. 146—154. The reference to the absence of a “more precise statistical measure” is not intended as a criticism of Gregg. He, like Singer and Sensenig, assumes that the economic index—in this case UN contributions—is a suitable measure of power, which itself is not a very precise concept. Since the measurement can scarcely be more precise than the thing being measured, one can persuasively argue that Gregg's verbal description of the relationship between the two rankings represents the underlying concept as accurately and precisely as any statistical correlation coefficient. A correlation coefficient contributes to greater precision, however, when a comparison of the strength of correlations between variables is desired.

6 The literature of African participation in the United Nations has been mainly concerned with African views and activity with respect to substantive questions and with the political behaviour of the African bloc. See, e.g., Karefa-Smart John, “Africa and the United Nations,” International Organization, Summer 1965 (Vol. 19, No. 3), pp. 764773; and Hoskyns Catherine, “The African States and the United Nations 1958—1964,” International Affairs, 07 1964 (Vol. 40, No. 3), pp. 466—480. Even in Hovet's Thomas comprehensive study of Africa in the United Nations (Evanston, III: Northwestern University Press, 1963)discussion of elections is limited largely to comments of a general nature or observations illustrating some broader aspect of African political behaviour. A brief but highly interesting comment on African participation in the United Nations is found in Ellis William W. and Salzberg John, “Africa and the U.N.: A Statistical Note,” American Behavioral Scientist, 04 1965 (Vol. 8, No. 8), pp. 3032. By use of multiple linear regression Ellis and Salzberg relate UN voting behaviour of African states to selected economic, physical, and other national characteristics. The resulting multiple correlation coefficient (.946) indicates “a very strong relationship” between the two. Their data do not extend to UN elections.

7 African states admitted since 1960 include Mauritania, Sierra Leone, and Tanganyika in 1961; Algeria, Burundi, Rwanda, and Uganda in 1962; Kenya and Zanzibar in 1963; Malawi and Zambia in 1964; Gambia in 1965; Botswana and Lesotho in 1966; Equatorial Guinea and Swaziland in 1968. The state of Tanzania was formed from the merger of Tanganyika and Zanzibar in 1964.

8 The Secretary-Generalship and positions on the International Court of Justice (ICJ), filled by concurring vote of the Assembly and the Security Council, are excluded.

9 During the years in question representatives from Asian and African states were expected to hold two or three of the nonpermanent seats on the Security Council and five of eighteen seats on the Economic and Social Council (four in 1961). By General Assembly Resolution 1192 (XII), adopted on December 12, 1957, four of thirteen vice presidencies and two of seven main committee chairmanships were allotted to Africa and Asia. At the end of 1963 diis was revised to give Afro-Asia seven of seventeen vice presidencies and three of the seven chairmanships. See Resolution 1990 (XVIII) of December 17, 1963. By Resolution 1991 (XVIII), also adopted on December 17, the Assembly voted to amend the UN Charter by enlarging the Security Council from eleven to fifteen members and the Economic and Social Council from eighteen to 27. Subject to ratification of the amendments five of the ten nonpermanent Security Council seats were assigned to Africa and Asia as were twelve of the 27 ECOSOC posts. These changes were not effected until 1966, which is beyond the period of this study.

10 Although differences in importance exist, any attempt to reduce the differences to a numerical scale is bound to be somewhat arbitrary. The Gregg system is adopted because it seems as reasonable as any and the use of similar weighting enhances the comparability of results. The weights assigned are as follows: Security Council, five per year; Economic and Social Council, three per year; Trusteeship Council, two per year; president of the General Assembly, ten per session and four per special session; chairman of a main committee, five per session and two per special session; vice chairman, two per session and one per special session; rapporteur, two per session and one per special session. See Gregg , Southwestern Social Science Quarterly, Vol. 46, No. 2, p. 150.

Singer and Sensenig do not assign numerical weights but distinguish between “important” and “other” posts in conducting their analysis. Singer and Sensenig , International Organization, Vol. 17, No. 4, p. 907.

11 Given the Kendall tau (beta) rank order correlation technique, a more accurate measure of association becomes possible as the table for each pair of variables approaches symmetry. Since most of the other variables have fifteen or more ranks, the higher figure for elections would obviously produce greater symmetry. One would not, of course, produce more ranks solely for the purpose of achieving symmetry, but it is an advantage when, as in this case, weighting also gives a closer approximation to reality.

12 Sources of data were as follows: 1) UN mission size: Permanent Missions to the United Nations, 01 1961–December 1965 (Nos. 114–185)(UN Documents ST/SB/Ser.A/i 15–185); 2) speech length, resolutions and amendments, Secretariat personnel: General Assembly Official Records and Annexes (16th–20th sessions), 19611965; 3) armed forces size: Weeks George, “The Armies of Africa,” Africa Report, 01 1964 (Vol. 9, No. 1), pp. 4—21; 4) United States aid: Annual Report to Congress on the Foreign Assistance Program (Washington: Department of State and Agency for International Development, 19611965); 5) gross national product: Proposed Mutual Defense and Assistance Program FY 1964, Summary presentation to the Congress (Washington: Agency for International Development and Department of Defense, 04 1963) ; and Proposed Foreign Aid Program FY 1968, Summary presentation to the Congress (Washington: Agency for International Development, 05 1967); 6) UN budgetary assessment: United Nations Yearbook, 1960–United Nations Yearbook, 1965 (New York: United Nations, 19611965); 7) population: Demographic Yearbook 1964 (New York: United Nations, 1965) ; 8) number of physicians: Statistical Yearbook 1966 (New York: United Nations, 1967); 9) radio receivers and newspaper circulation: World Communications: Press, Radio, Television, Film (4th ed.; New York: United Nations Educational, Scientific and Cultural Organization, 1964); 10) school enrolment: Russett Bruce M. and others, World Handbook, of Political and Social Indicators (New Haven, Conn: Yale University Press, 1964); 11) college enrolment: Statistical Yearbook 1963 (New York: United Nations, 1964).

13 A rank order correlation, as compared with an interval scale such as Pearson's r, appeared to demand less of the data, which is not as reliable as one might hope in the case of some economic and social indices used. An interval scale would also assume a consistent one-to-one relationship between the indicators and the underlying variables which is not necessarily the case. We are not convinced, for example, that one country is twice as developed as another because it has twice as many radios or newspapers per capita. Having opted to treat the data as ordinal rather than interval, we preferred Kendall's tau over Spearman's r because of the large number of ties in the rankings. This problem is discussed in Blalock Hubert M. Jr, Social Statistics (New York: McGraw-Hill Book Company, 1960), p. 317. For fuller discussion see Kendall M. G., Ratify Correlation Methods (New York: Hafner Publishing Company, 1955), especially Chapters 1, 3.

14 Blalock, p. 321. See Blalock, pp. 319–321, for the method by which a tau is computed.

15 Whenever any one of the six differs markedly from the others in its association with a participation or development variable, a common-sense reason can usually be found. For example, population produces a very low tau for each of the per capita (development) variables. This is to be expected because per capita figures by their nature rule out population size as a possible source of variation.

16 Length of UN membership is another variable related to UN participation although it is not a “participation” indicator in the same sense as the four utilized in this study. A tau for length of UN membership and office holding gives the figure of .50 which is slightly higher than any of the participation variables. The relevance for office holding seems clear. Members of the Organization before 1960 had a longer time within which to become familiar with UN processes and issues and to create contacts and obligations that can be cashed in for electoral support. Moreover, chairmanships of Assembly main committees usually go to persons who have qualified themselves through experience at several preceding sessions of the Assembly. Of the nine African states admitted before 1960 eight rank among the top ten states in office holding. On the other hand, length of membership apparently has less to do with the capacities and interests that encourage a state to be active in UN affairs once UN membership is attained. Its correlation with the four participation variables yields coefficients ranging from .28 to .33—all substantially lower than any of the inter correlations of the participation variables.

17 Singer and Sensenig , International Organization, Vol. 17, No. 4, pp. 907910, 914–918. One cannot be sure about the strength of the relationship they found without an examination of their data.

18 Primary and secondary school enrolment is an exception. It has no significant correlation with any participation variable or with office holding. Armed forces is another partial exception. Its tau of .32 for speech length is slightly higher than its .29 for elections.

19 The unexpected correlation coefficients for primary and secondary school enrolment suggest the possibility that available data for these indicators were not very accurate.

20 Gregg , Southwestern Social Science Quarterly, Vol. 46, No. 1, p. 153.

21 We tried to control for length of UN membership by including only African states that were UN Members during the entire five-year time period.

22 Singer and Sensenig , International Organization, Vol. 17, No. 4, p. 924.

23 Singer and Sensenig found a .477 Kendall tau for GNP and “important” offices. The tau for “other” offices was not given, but presumably it was much lower. See ibid., p. 908.

24 Using their data, Singer and Sensenig found:

The correlation coefficient for important offices with economic aid alone in period II [1950–1954] was .472, a highly significant figure. The correlation was not nearly as high for any other period or for other offices.

Ibid., p. 916.

25 Ibid., p. 912.

26 For comment on length of membership as a variable see footnote 16 above.

Studying for the doctorate in the Department of Sociology and Anthropology at the University of Notre Dame, Notre Dame, Indiana

Professor of Political Science at the University of Minnesota, Minneapolis, Minnesota. This study was supported in part by a grant from the Office of International Programs of the University of Minnesota

Recommend this journal

Email your librarian or administrator to recommend adding this journal to your organisation's collection.

International Organization
  • ISSN: 0020-8183
  • EISSN: 1531-5088
  • URL: /core/journals/international-organization
Please enter your name
Please enter a valid email address
Who would you like to send this to? *


Full text views

Total number of HTML views: 0
Total number of PDF views: 2 *
Loading metrics...

Abstract views

Total abstract views: 146 *
Loading metrics...

* Views captured on Cambridge Core between September 2016 - 21st January 2018. This data will be updated every 24 hours.