FOREIGN RELATIONS 0F 34 ATTIIOAN STATES: REGIONAL PATTERNS AND CHANGES AFTER MILITARY coups O’ETAT. ' I954 TIIROUOII 1967 Dissertation for the Degree of Ph. D..— MICHIGAN STATE UNIVERSITY ' CURTIS ELTON RUFF. JR. ' 1924 LIBRARY 7 Michigan Sm University This is to certify that the thesis entitled Foreign Relations of 34 African States: Regional Patterns and Changes After Military Coups D'Etat, 1964 Through 1967 ‘ presented by . @WQE.W , has been Oaccepted towards fulfillment ”9172:2233 Zing Mm MW KW fins/4 1 * Major professor 0-7 639 __ urn-l 'IJ - ‘- count ABSTRACT FOREIGN RELATIONs OF 34 AFRICAN STATES: REGIONAL PATTERNS AND CHANGES AFTER MILITARY cones D'ETAT, 1964 THROUGH 1967 By Curtis Elton Huff, Jr. This dissertation explores empirically the patterns of foreign relations of 34 African states, 1964 through 1967, and it tests the applicability of several images of the military to the foreign behaviors of military governments which took power in ten of these countries by coups d'état. In addition, the dissertation proposes three new parameters for analyzing foreign relations and an eight—fold typology of international actors based on these parameters. Of the 40 independent states in Africa during the four-year period analyzed, the three very newest and smallest states and the very different, white—controlled countries of South Africa and Rhodesia are excluded from the study. Egypt did not figure in the comparison of military and civilian governments‘ policies, since it Was the one country which could not be clearly classified as either military or civilian. Civilian rule was defined in terms of the frequency of presence of military officers in cabinet posts, the size 0f the army, military expenditures as a percentage of gross national PrOduct, and the percentage of the population under arms. —.-ll..-—.. -4— - for I 1967 diplI of f: each occu. NIL inte. patt oft inte indi ten calc YEA: comp Same Sign 8C0: inte Sign Curtis Elton Huff, Jr. Data were collected on 48 variables of foreign interactions for each of the states for each of the four years from 1964 through 1967. The data measure three substantive areas: trade, aid, and diplomacy. Some of the variables were combined into analytic indices of foreign—affairs intensity, extensity, and alignment with respect to each substantive area. Intensity measures the amount or frequency of occurrence of the interaction; extensity measures the proportion of the world's countries and international organizations with which the interaction takes place; and alignment measures the extent to which the patterns of interaction are similar to those of other African states, of the United States, or of the Communist Bloc. The distributions of all 34 countries on four variables of international activity are analyzed for the years 1964 and 1967, to indicate baselines of continental activity and change. Then the trends in foreign relations for the group of military regimes are calculated, measuring the change in each variable and index from one year before the coups d'état to one year after. These trends are compared with the average changes among the non—coup group during the same period. Mann—Whitney U tests are computed to test for the significance of difference between the coup and non-coup groupS' scores. The military group shows a slightly diminished post-coup intensity of trade and aid, relative to pre-coup levels. It also shows significantly less post-coup activity in these areas than the non— coup group. The military regimes do, however, tend to show wider trade relationships among other countries than the non—coup group. The military group receives increasingly intense aid and diplomatic voti comp mili over grea coup inac affa 90w lte Curtis Elton Huff, Jr. interest from the United States, and decreasing aid and diplomatic interest from the Communist Bloc. Aid from multilateral sources to the new military regimes diminishes significantly. In United Nations voting, the military group tends to vote more often with the Soviet Union than with the United States, relative to the non—coup states, although a small subset of each group becomes much more favorable to the United States. The military group shows a significant drop in its voting alignment with the African majority at the United Nations, as compared with the non—coup group, and a substantial majority of the military group votes much more heavily against the African majority than do most non-coup states. On a continental basis, the African states show significant increases in diplomatic activity and alignments over the four years, but the military group shows a significantly greater increase in intensity of multilateral diplomacy than the non- coup group. Overall, there is no evidence of peculiar incompetence, inactivity, or reactionary values which would distinguish the foreign affairs of military from civilian governments in Africa. There is evidence of overriding pragmatism in foreign relations of military governments, suggesting they fit better the perceived "national- interest" rather than the "ideological" style of foreign policy. There is also evidence that these countries suffer in foreign economic relations after military coups, despite the military interest in tangible foreign affairs. Finally, while it does seem true that there is very little exPlicit, shared ideological basis for foreign—policy behavior among Phil regi Curtis Elton Huff, Jr. African military regimes, there are many similarities in orientation to foreign affairs which distinguish the military from civilian regimes on that continent. FOREIGN RELATIONS OF 34 AFRICAN STATES: REGIONAL PATTERNS AND CHANGES AFTER MILITARY COUPS D'ETAT, 1964 THROUGH 1967 BY Curtis Elton Huff, Jr. A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Political Science 1974 @Copyright by CURTIS ELTON HUFF, JR. 1974 oft inte Coll goal Coll Ruth ACKNOWLEDGMENTS Special thanks are due to Dr. Bruce James Smith, who took charge of my supervising committee late in the project and provided the intellectual and moral support for completion; to Dr. John Norman Collins, who helped launch the research, especially by clarifying its goals and methods; and to the staff of the computer center of Williams College, who wrote and explained programs and gave me a key to the computer room. I also wish to thank my wife Jane, my children Lucinda and Anthony, my parents, and parents-in—law for their forebearance, love, and support through the years of this work. ii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . vi LIST OF FIGURES. . . . . . . . . . . . . . . . . ix Chapter I ,I I. INTRODUCTION . . . . . . . . . . . . . . . l 7 The Military and Africa . . . . . . . . . . . 1 Significance of Africa . . . . . . . . . . . 4 Comparative Analysis . . . . . . . . . . . . 8 Elite Analysis . . . . . . . . . . . . . . 16 Theoretical Goals . . . . . . . . . . . . . 22 II. THE AFRICAN STATES: SIMILARITIES AND DIFFERENCES . . . 27 The Sample of 34 Countries . . . . . . . . . . 27 Coups and Coup Countries. . . . . . . . . 37 Differences Between North and Tropical Africa. . . . 55 III. FOREIGN RELATIONS THEORY . . . . . . . . . . . 62 Problems of African Foreign Relations . . . . 62 The Influence of Top Leadership in African Foreign Affairs. . . . . . . . . . . . . 75 Foreign—Relations Theory. . . . . . . . . 83 The Concepts of Foreign Affairs Intensity, Extensity, and Alignment . . . . . . . . . . 94 A Typology of International Actors . . . . . . . 98 IV. THE MILITARY AND FOREIGN RELATIONS THEORY. . . . . . 108 The Armed Forces of Africa . . . . . . . . . . 108 Military and Civilian Regimes . . . . . . . . . 1:: The Military Outlook . . . . . . . . . . . . 136 Military Roles in New states . . . . . . . . 140 Hypotheses on Military Influence on Foreign Relations . CbAptI VI I VII. Chapter V. THE DATA AND THE RESEARCH DESIGN . . . . . . . Selection of Variables and Data Collection . . . Possible Errors and Reliability Problems. . . . Formation of Indices of Intensity, Extensity, and Alignment. . . . . . . . . . . . . Method of Analysis . . . . . . . . . . . VI. ANALYSES OF DATA . . . . . . . . . . . . Baselines of International Activity, 1964 . . . Foreign Behavior on Analytic Parameters, 1964 . . International Activity, 1967. . . . . . . . Patterns of Analytic Parameters, 1967. . . . . Changes in Selected International Activities, 1964— 1967 . . . . . . . Changes in Patterns on Analytic Parameters, 1964-1967 . . . . . Mann-Whitney U Tests of All Trend— Variables. . . Significance Tests of All Trend Indices (Parameters) . . . . . . . . . Composite Change on All Indices. . . . . . . VII. DISCUSSION OF RESULTS. . . . . . . . . . . General Continential Trends, 1964— 67 . . Confirmation or Disproof of Hypotheses of Military Influences . . . . . . . General Implications of Military Influences in African Foreign Relations . . . . . . . . APPENDICES Appendix I. Africa Research Variables . . . . . . . . . II. Sources of Data by Variables . . . . . . . . 111- Size, Social, & Economic Characteristics of African States . . . . . . . . . . . . . . . IV- 1964 Intensity Variables. . . . . . . . . . V. 1967 Intensity Variables. . . . . . . . . . iv Page 155 155 158 162 165 175 175 181 186 190 208 214 224 230 236 236 256 261 263 267 270 Page Chapter Appendix VI. 1964 Extensity Variables . . . . . . . . . . . 274 VII. 1967 Extensity Variables . . . . . . . . . . . 276 VIII. 1964 Alignment Variables . . . . . . . . . . . 278 Ix. 1967 Alignment Variables . . . . . . . . . . . 279 280 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . 5. LIST OF TABLES Thirty-four African Countries, by GNP (1966) . . Comparisons Between Non-Communist Developed and Less- Developed and Less—Developed Areas (1970) . . Political Enculturation in Four Underdeveloped Regions. . . . . . . . Interest-Articulation by Associational Groups, Four Less-Developed Regions. . . Interest-Articulation by Non—Associational Groups, Less-Developed Regions. . . Interest-Articulation by Anomic Groups, Four Less- Developed Regions . . . . Character of Bureaucracy in Four Regions. . . . . . . . Coup and Non—Coup Group Averages Statistics. . . . . . . Coup and Non-Coup Group Averages Statistics. . . . . . . Coup and Non-Coup Group Averages Statistics 0 O O O O O 0 African Coups, 1965-1967. . . Comparison of North and Tropical Comparison of North and Tropical Comparison of North and Tropical vi Less—Developed and Mann-Whitney U and Mann-Whitney and Mann—Whitney Group Averages . Group Averages . Group Averages . Page 29 31 34 35 35 36 36 39 4O 40 45 57 58 58 Decolonization Scores of Thirty African States Ranked by Size of Score and Grouped by Foreign—Policy Actor-Type . . . . . . . . . . . . . Sizes of the Armies of African States (Coup Countries in Parentheses) . . . . . . . . . . . . Changes in Size and Burden of African Armies . . . Partial Indicators of Extent of Military Influence in African Governments, 1964 . . . . . . . . . Distribution of Estimates of Military Influence for 35 States in Pre-Coup Period . . . . . . . . . Predictions of Degrees of shift in Foreign Relations after Coups D'Etat . . . . . . . . . . . International Activities in 1964 (Coup Countries in Parentheses) . . . . . . . . . . . . . 1964 Intercorrelations. . . . . . . . . . . Exports/GNP for Selected Non—African Countries, 1964. International Activities in 1967 (Coup Countries in Parentheses) . . . . . . . . . . . . . 1967 Intercorrelations. . . . . . . . . . . Country Rankings by Amounts of Change Selected International Activities, 1964-67 (Coup Countries Underlined; Tied Scores Bracketed) . . . . . . Country Rankings by Percentage of Change Selected International Activities, 1964-67 (Coup Countries Underlined, low & Change at Top). . . . . . . Country Rankings by Amounts and Direction of Change Selected International Activities, 1964-67 (Coup Countries Underlined, High Positive Change at Top). Ranking of Changes in Correlations. . . . . . . Discrimination Between Coup and Non-Coup Countries: Trend Data Ranked by Probabilities From Mann- Whitney Scores. . . . . . . . . . . . . Page 71 113 126 146 151 176 180 187 189 200 202 207 ale -1l.—-Discrimination Between Coup and Non-Coup Countries of 14. 15. 16. 17. Tropical Africa: Trend-Data Ranked by Probabilities From Mann-Whitney U Scores . . . . . . . . . Rates of Change in UN Voting After Military Coups: Coup Countries Voting Against African Majority, 1964 vs. 1967 . . . . . . . . . . . . . Changes in Discrimination Probabilities After Removing North African Countries . . . . . . . Trend Indices and Probabilities Associated with Mann— Whitney US Scores: North and Tropical Africa. . . Trend Indices: Probabilities Associated With Mann— Whitney U Scores . . . . . . . . . . . . Changes in Discrimination Probabilities After Removing North African Countries . . . . . . . . . . Relations Between Country Economic Measures and Composite Change on All Trend Indices . . . . . Relationship Between (1) Predicted Country Rankings of Degrees of Shift in Foreign Relations, and (2) Actual Changes . . . . . . . . . . . . . 1964 Intercorrelations Among Parameters . . . . . 1967 Intercorrelations Among Parameters . . . . . Intercorrelations of Country Rankings in Among and Direction of Changes on Parameters, 1964-67 . . . viii Page 222 224 227 231 232 234 234 244 245 246 ure LI ST OF FIGURES Cross Tabulation of African Coup Countries on Measures of Population and Gross National Product per Capita . . . . . . . . . . . . . . . . Cross Tabulation of African Coup Countries on Measures of Gross National Product and GNP per Capita . . . Categories of Intervention . . . . . . . . . . Classification of African Groups by Policy— Orientation . . . . . . . . . . . . . . Cross-Tabulation of African States on Measures of Economic Size and Bloc. . . . . . . . . . . Cross—Tabulation of African States by Bloc and Diplomatic Activity (Percentages of Coup States Noted Within Each Group) . . . . . . . . . . Substantive Cumulation of Theory . . . . . . . . Rosenau's Eight Genotypical Actors . . . . . . . Five Sets of Variables Underlying the External Behavior of Societies Ranked According to Their Relative Potencies in Sixteen Types of Societies and Three Types of Issue-Areas . . . . . . . . . . . Categories of Foreign Behavior. . . . . . . . . "Diplomacy" . . . . . . . . . . . . . . . Categories of Foreign Behavior Analysis. . . . . . Typology of Actors in Foreign Affairs . . . . . . ix Page 43 44 49 67 69 73 85 87 88 92 93 94 100 ire Sizes of African Armies vs. National Population Sizes and GNP, 1966 . . . . . . . . . . . . Variables in 3-Dimensiona1 Theory of Civil—Military Relations . . . . . . . . . . . . . . Typology of Military Roles in Politics . . . . . Typology of Military Political Roles and Regimes, ca. 1972 . . . . . . . . . . . . . . . . Relation Burden of Military Expenditures (Coup Countries in Parentheses) . . . . . . . . . Estimates of Military Influence in Pre—Coup Period (Coup Countries in Parentheses) . . . . . . . Predictions of Degrees of Shift in Foreign Relations After Coups D'Etat Based on Social—Economic Development. . . . . . . . . . . . . . Predications of Degrees of Shift in Foreign Relations After Coups D'Etat Based on Social—Economic Size . Predicted Changes in Foreign Relations after Coups D'Etat . . . . . . . . . . . . . Variables Included in Indices of Intensity . . . . Variables Included in Indices of Extensity . . . . Variables Included in Indices of Alignment . . . . 1964 Country Rankings on Parameters . . . . . . Typology of Selected Countries in 1964 . . . . . TYPOlogy of Coup Countries in 1964. . . . . . . 1967 Country Rankings on Parameters (Coup Countries Underlined). . . . . . . . . . . . . . TYP01Ogy of Selected Countries in 1967 . . . . . TYPOlOgy of Coup Countries in 1967. . . . . . . Classification of Countries Showing Highest % of Change Selected International Activ1t1es, 1964—67 . Page 111 119 120 122 143 145 149 150 163 163 182 183 185 191 193 205 Page Country Ranking by Amounts of Change on Parameters, 1964 to 1967 (Coup Countries Underlined, Tied Scores Bracketed) . . . . . . . . . . . . . . 209 Country Ranking by Amount and Direction of Change on Parameters, 1964 to 1967 (Coup Countries Underlined; Tied Scores Bracketed) . . . . . . . . . . 210 Country Rankings Aggregated Across Three Analytic Parameters (Coup Countries Underlined). . . . . 212 Discrimination Between Coup and Non-Coup Countries on Trend Indices Grouped by Types of Foreign Behavior: North and Tropical Africa 35 Countries . 223 Discrimination Between Coup and Non-Coup Countries on Trend Indices Grouped by Types of Foreign Behavior: Tropical Africa 30 Countries . . . . 229 ” Actual Changes in Foreign Relations After Military Coups D'Etat . . . . . . . , . . . . . 253 xi CHAPTER I INTRODUCTION The Military and Africa The 19603 were the first full decade of independence for most ican countries. In an unprecedented way, Virtually a whole continent ame free under various new names and commenced a struggle for recog- fion and development. Accordingly, the decade began with an élan rn of great anticipations and released with the winning of self— ernment. By mid—decade, frustrations attendant on the enormity of ks and the slow pace of progress had set in. Some regimes held ether; some fell. These difficult and turbulent middle years of the ade are analyzed here, with attention focussed on the foreign rela— ns of the African states. During this period, and since, all but one of the regime— nges in Africa have come by military coups d'etat.* There have been electoral changes in government. The central question of this sertation is this: What effect has this intervention of the military, >ne-third of the continent's countries, had on the international wior of these states? For even the experts, this degree of M *President Tubman of Liberia died of natural causes and was Laced by his Vice-President in a smooth and legal succession. tary presence was unforeseen as late as 1966 (Spiro, 1967; Foltz, ). Early in the decade, Lucian Pye (1962) wrote that "it occurred ew students of the underdeveloped regions that the military might me the critical group in shaping the course of nation-building." result was then, and still is, that we are confronted with the ard fact of having no clear theory of the role of the military in political development of the new states. We seem to be caught with two conflicting images of the political character of armies in backward countries: first, there is the early image, derived largely from Latin America and the Balkans, in hich the soldier stands for administrative incompetence and 'naction, and authoritarian, if not reactionary, values; and, econd, there is the newer picture of a dynamic and self—sacrificing ilitary leadership committed to progress and to the task of odernizing transitional societies which have been overpowered by e 'corrupt practices' of politicians. How is it possible to tell 'n the particular case whether army rule will lead to sterile authoritarianism or to vigorous development (Pye, 1962)? :3‘ Since Pye wrote these words, there have been at least 25 ;ions in various African countries when the armed forces have Issfully intervened in politics (Gutteridge, 1970). here has also been an uncertain but substantial number of abortive cups, of which by far the most important were the attempted coup y the Ethiopian Imperial Guard in 1960, the East African mutinies f 1964, and the 1967 attempt to overthrow Colonel Boumédienne in lgeria (Ibid.). Naturally, several studies, large and small, have followed this of military coups. Virtually all of these attempt to deal with lestion of how the military did or could take power (for example, n 1966; Feit, 1968; Gutteridge, 1969; Lee, 1969; and Bienen, For these studies, the dependent variable to be explained is up itself. In the last year or so, scholars have begun to talk of treating “P as an independent variable to explain recent policies and tivities of military regimes. Nonetheless, to my knowledge, there for Africa only one published attempt (Welch, 1970) to describe and plain the effect the military have, or what they actually do, when ey take over, and it is a non-empirical work concerned primarily with mestic change. Eric Nordlinger's more quantitative work (1970) deals th military influence on domenstic development in several less— veloped areas, and it is limited by the Adelman and Morris (1967) ta it uses, since these data pertain mostly to the pre—independence riod, before any coups in Africa. Also, Nordlinger commits the iversal fallacy for Africa: that is, he argues in spite of the fact at the few data he shows indicate that region is categorically dif— rent from all others he considers, that the relationships which hold other areas for which he has better data surely will also hold in rica. But he cannot prove this. One is led to wonder if there is not some general design or aology running through these military interventions occurring so Jidly and widely in Africa, and whether that orientation, if it .sts, fits either of Pye's images or some other. In spite of stide Zolberg's remark (1968) that "it is impossible to specify iables which distinguish §§_a_glg§§ countries where coups have urred from others which have so far been spared," it would seem t the coup is a peculiarly frequent phenomenon through Africa, there 19 coups in well over a third of these states (16 out of 43) between i and 1970. Gutteridge also rejects an inference of common theme throughout e coups, however, saying: One thing is certain, that only in rare cases—-such as that of the Nasser revolution in Egypt-—has there been a firm ideological basis for action; even in Egypt there was only the vestige of a precon- ceived political programme. This being so, it is hard to account for the transformation in the political status of the military which took place during the first half of the 19605 (1970). There are simply no systematic and quantitative studies of the :ies and actions of military regimes. Especially with regard to nilitary orientation in foreign affairs, then, we are left with the Licting images of which Pye spoke, plus other speculations and 32 judgments added over the years. It is the purpose of this irch to explore the variety of writings about the military in :a, to develop some specific hypotheses about the orientations of :ary regimes in foreign policy and international behavior, and to these notions quantitatively across all African military govern— ; in office during the period from 1964 through 1967. Even if a regimes proclaim no common ideology, we may discover common tests and practices and distinguish these from idiosyncratic con— and behavior. Significance of Africa Africa is chosen for study because of its special advantage to -national, comparative analysis, as well as for its interest to uthor. Foreign relations are the behaviors we focus on, because relations are important in themselves, and because they can be ed externally and compared across countries and across time— ds more easily than domestic policies and behaviors. To the t that the African states may be considered a relatively homo— us class of countries, the analysis of changes in foreign ions as a function of changes in national governmental leaderships, ECiallY those changes resulting from military coups, allows the s-national explication of both idiosyncratic and common patterns. he military behave any differently from civilians after taking over can governments, one may be able to measure that difference system- ally. The relations of the development of foreign policies and viors in the new states to other social, economic, and political esses have not been carefully charted. In the case of Africa, this of research does not necessarily reflect laziness or an oppor— ty overlooked, so much as a unique situation: most of the African as have existed as independent entities only a decade or so, and e has not yet been time to chart trends. Twenty years ago there only four independent states in Africa; today there are 43. This asion has come so recently that data have been unavailable, at t data allowing comparisons across several countries. Two major ndia of aggregate indicators of national characteristics published e 19605, The World Hgndbook of Political and Social Indicators ett, £3 31., 1964) and The Dimensionality of Nations Project e1, EE.E£-r 1966) excluded most African nations, either because ata—collection periods were prior to 1960 or because post- endence data were not yet available. The 19705 have brought a few with a more comprehensive View of Africa (Bretton, 1973; Rubin einstein, 1974), and the second edition of The World Handbook of ical and Social Indicators (Taylor and Hudson, 1972) has been ded to include data on most African states. Among currently shed quantitative research, only that of Collins (1971) and of an (1969) deals explicitly with the whole African continent, and of these deals only with a carefully delimited problem in foreign tions. As they indicate, it is only now and in the future that one study with any generality the behaviors of these states. The historical uniqueness of the African states suggests a fold research problem. First, because of this uniqueness, there few traditions, few guidelines, and few indices for measuring the rnal behaviors of these states, except by inference from research on-African states. Even in this research, every author decries the rdevelopment of theory (see, for example, Hoffman, 1960; Rosenau cle in Farrell volume, 1966; Brecher gt 31., 1969). In order to d the possible ethnocentrism of views developed in other areas, , one must begin the empirical study of Africa by throwing a very d data-net. Only by measuring as many different kinds of behavior ossible can one sift out what in Africa is of significance; that is, behavior African states have in common with non—African states, is special about the African states as a subset of the world's es, and what is idiosyncratic about each state within Africa. larly, only with multiple and various measures can one expect to over systematic divergences between military-led and civilian-led es. Accordingly, in this research, data have been assembled on than forty variables of foreign interaction for each of four years for nearly all the African states. Second, the uniqueness of Africa means not only that researchers start from scratch, but also that African governments are faced the special task of creating policy 32.22X2: While options in Lgn relations always pose dilemmas for any government, this is :ially critical for new and developing countries where the goals of )vereign identity, prestige, and self—sufficiency clash with the facts E material, technical, and sometimes even psychological dependence. swan (1968) calls foreign policy "an entirely new dimension of official ancern" for the new African states, their leaders having been oriented p to the 19605 almost solely to the nationalist struggle for indepen— ence. "Basic foreign-policy positions had to be worked out by ersonnel with little or no training or experience and these had to be ransmitted abroad to ambassadors who, perforce, had to learn the dif- icult art of diplomatic negotiation through experience as heads of issions in their first posts" (Ibid). This hurdle of novelty is much amplicated, as Robert C. Good (1962) has pointed out, by the crucial mportance of foreign policy to the process of rapid state-building; he facts of dependence must be squared with the conditions for avelopment. For these reasons, we might expect new departures by the Erican states in the structures of their foreign relations, reflecting pecial individual and continental needs and opportunities. Accord~ Tgly, again, a multivariate, cross-national and comparative study is equired to get a valid picture of African foreign relations. In Africa, we find a cluster of state5--nearly one quarter of m world's countries—-where these problems come true all at once. mh African state is nested in a continent in which countries share dnful sets of problems in foreign affairs and novel conditions for filing with them. These are not several isolated countries, then, m a "set" or a "class" of states, forming a continent, sharing the Periences of recent colonialism, recent independence, and general derdevelopment, as well as a sense of ethnic commonality vis—a—vis e rest of the world. This commonality may be expressed in foreign Affairs by the fact that “all Africans feel in some degree the gravi- :ationa1 pull toward continental solidarity,“ while at the same time feeling the dilemma that “the longing to manage their own affairs and :o be free of European supervision conflicts directly with the urgent med for outside help" (Scipio. 1965). Understanding this "class“ or ’set" of countries identified as "new, poor, African" depends on seeing be common configuration of newness, need, and sense of enigma. Comparative Analysis The extent to which the African configuration of characteristics 5 significantly different from that of other regions of the world hould be systematically described ultimately, and analyzed in measur— ble terms, based on a theory of comparison. The notion of Africa as n important "class" of countries implies comparison with other sets, uch as Southeast Asia or Latin America or East Europe. Data are resented in Part One of Chapter Two to substantiate the claim that frica is such a distinct class of states, but no general theory of egions as classes is expounded. Thus, while I argue that Africa is an mportant and distinct region, there is no explicit theory here for aneralizing the findings in Africa to other areas of the world. The urpose here is a more narrow one for which analysis across several alatively similar states——Africa—-is appropriate, namely to identify nether and in what ways military governments have a distinct orienta— ion in foreign affairs, by analyzing the range, variation, and changes Eempirical patterns of foreign interactions. The notion that Africa, as a "class“ of countries, is more Ian a geographic region is quite important to the goal and design of is research, therefore. Either military governments have some mmon orientations to foreign affairs because they are military, or ey have not. This goal of saying something empirically verifiable out the orientations of the military toward African states' foreign lations requires comparison of measures of these phenomena across veral countries, in order to distinguish common from idiosyncratic fects. Comparison is useful, though, only to the extent to which the veral countries differ in known ways which affect the dependent enomenon. Selection of a set of countries which are relatively nilar reduces the likelihood that the test for common military lentations may be confounded by unknown or uncontrolled variables. is argued that Africa is optimal in this respect, with countries Eficiently alike to be called a "class" or socio-political region. African countries are no random selection from some larger \ .verse, then. Hence, research results from this region may not be eralizable to other regions. Africa is explicitly a skewed sample, fact, but that is its justification, a justification based on my ire to control for national variations which might influence the nomena under study but are beyond our present theoretical purview. e there grand theory available specifying all national character— ics relevant to the phenomena under study, and were all these iables accurately measured, available multivariate techniques could ceivably specify relations among these dimensions. In such a aretical and data paradise, research could account for differences fig regions or "classes" of countries. Of course, such grand theory Ld obviate the need for the more narrow research here. The present :arch may be an important first step toward formulating such theory. a wad“... .. .. w [as wrote}. 10 If the geographic region of Africa can be shown to be a socio- political "class" of states, and if the characteristic configuration of that class can be shown to account appreciably for the phenomena under study, then we have a relatively powerful opportunity for analyzing those phenomena. This is "configurative analysis" in Hayward Alker's terms (1964), in that it describes a structural context important to the research phenomena. The assumption is that the meanings of given events or processes may vary from context to context-—that the relation- ships among given variables (here, military rule and foreign policy) may be different in different geographic or cultural contexts. Con— figurative analysis helps elucidate the phenomena under study by isolating and specifying the contexts in which they may or do occur and thereby helps "correct for the universal fallacy of inferring anything about particular regions or stages of development from universalistic relationships" (lbid.). Indeed, we not only have a "class" of states identified by a >articular configuration of national characteristics (new, poor, wakly organized, African), but an extraordinary frequency of coups Pétat here, perhaps itself a class characteristic. Africa seems mlike any other region in the ease and frequency of military inter- ention during the 19605. Neither Latin America nor Asia had such a li9h rate of coups during the decade. Thus, Africa provides both a Emarkable context for studying the military and, also, a remarkable Mantity of military behavior that may have had a bearing on foreign olicy. Something may be lost, of course, in cross—national comparative nalysis which could be gained by case studies of individual states. ll conceptions held by specific elities, their personal histories, the liar influences of local ethnic complexities, the specific triggers casualties of coups, the exact sequences of events, and so on, are sed over by the comparative design of this research because of ts of theory, time, and data. But those losses are not crucial to objective of this research, which is to ascertain whether and in ways military governments of Africa show common orientations in ign relations distinct from those of civilian regimes. Without blishing first that there is a correlation between military govern— vi a] w s and certain policy-orientations, there is no adequate theoretical e for detailed case research. Without the background of a compara- , cross-national study, a case researcher could not be sure whether ad identified an idiosyncracy in his country of study or whether he i evidence of a pattern of behavior typical of, say, former British :ies or of trpical agricultural economies generally, or of the Lnent of Africa. In other words, case research pg£_§g cannot alish theory; it must be guided by theory or by observed general :ionships, if it is to have theoretical payoff. And the question ‘9 us is not "Why do military governments always do x, Y, and Z in .gn affairs?", but "Given alternative speculations about their :y-orientations, do military governments in fact show charateristic ltations in foreign affairs, and which of the speculations, if any, to be confirmed by the findings?" The impact of the military on the African states' foreign- ?y behavior depends on two sets of variables--the nature of the ary and the nature of the states which they govern-—and the way two sets engage one another. All the African states have certain 12 ternal attributes whose configuration sets them apart as a class and y have a bearing on their foreign-affairs behavior. The relationship attributes to that behavior may be summarized in the following uation: General foreign-affairs f(similar colonial experiences + behaviors of African states = + recency of independence + + economic underdevelopment + + weak, political and diplomatic institutions, etc.) e view of the military might summarize their behavioral style in this uation: Military style = f (professionalism + low ideological interest + + nationalistic orientation + economic development orientation + technical skills) >ther view of the military might summarize their behavioral style in 5 equation: Military style = f (bully mentality + conservative values + + self-serving orientation + isolation from non—military knowledge + limited bargaining skills) The task of this research is to ascertain which, if either, of views of military style seems to be borne out in the behaviors rican military governments. One could then symbolize the effects itary coups on African states' foreign-affairs behaviors by ing the equations above as follows: Foreign-affairs behaviors f[(style of the military) + (gen— under military regimes = eral foreign-policy orientations of African states) + (points of interaction)] ending on the style of the military that we postulate as accurate, nay predict different foreign—affairs behaviors under military ' es. Also, of course, the African states are not identical and the Lens military regimes which have taken power in several countries not identical, so a more accurate equation would be: Foreign-affairs behaviors f[(military commonality) + (military under military regimes = idiosyncracy) + (state common- ality) + (state idiosyncracy) + (interactions of common and idiosyncratic factors)] In this research, data and argument are presented to support view that the African states, in contradistinction from other ns of the world, show a greater commonality than individual syncracy in attributes important to general foreign-affairs dors and military styles; that is the basis for calling them a . Comparative analysis of two subsets of states-—one having had ary coups and one having continuous civilian governments——will :he limits to the commonality of the general foreign—affairs ors of African states and the foreign-affairs behaviors under :y regimes. We have, then, a quasi—experimental study approxi- the "nonequivalent-control-group" design suggested by Campbell iley (1963), modified by the facts that the analysis is §x_22§E ,nd we can assume, though not experimentally find, appreciable nce between the groups in the sample of states. Iome difficulties are unavoidable when we use an §x_pg§3 facto, erimental design. In particular, looking back on completed was not allow for a full test of proposed explanations of l4 lose actions; a more critical test would be possible if we could pre— .ct a future coup d'état and the military reorientations of foreign dicy which would derive from it, and sit back and watch the events fold. Though theoretically a more powerful test, this seems an even ss likely method for building theory than the less—controlled sign employed here. Whatever its difficulties, there is a distinct payoff in llowing through with the design here. As Campbell and Stanley (1963) d others (e.g., Stinchcombe, 1968) indicate, that payoff comes in ing able to eliminate certain rival hypotheses about the phenomena of terest. Two sets of rival hypotheses are important to the following First, there are alternative notions about the configura— apters. ans, or class characteristics, of the African countries. These :ions imply different probabilities or different kinds of change in eign relations after military coups d'état. Three such alternatives ht be the following: I. One might argue that the options in foreign affairs are so narrowly drawn for the African countries, given their general underdevelopment, that the patterns of foreign-affairs behavior will not change significantly after any change in regime, whether civilian or military. One might argue that foreign relations are essentially an elite 1969, for phenomenon in Africa (see Morrison and Stevenson, such an argument) because bureaucracies and pressure groups are poorly developed and masses unmobilized and, thus, the decision- latitude of elites is great, resulting in clear changes in foreign behavior after coups to the extent that the military are systematically or idiosyncratically different from the civilians they replace. One might argue that the commonality of African states is not all that great or relevant, or that options are not severely constrained in foreign affairs (the paradox of "leverage through weakness," C. T. Thorne, Jr., in African Diplomacy, 1966), or that civilian and military elites are not systemat— ically different in their approaches to foreign relations 15 (for example, both may hold the goal of rapid development as primary). In these cases, one would expect a variety of changes, or no changes at all, in the patterns of foreign relations of countries experiencing military coups. These three alternative notions describe different ideas about 1e military regimes and the "classness" of African states. They are Lso mutually exclusive for the set of African states; the empirical aasurement of foreign—relations patterns for this set of states can )nfirm only one. In part, the significance of this research derives mm our inability to be sure which notion is most nearly correct, .though the research is begun with the hypothesis that the second ttion is the most probable. A second set of rival hypotheses deals with alternative images the military. Even if one can properly assume the second of the ternative propositions described above, change in policy after litary coups is not predictable unless we have a correct image of the litary. Given the dichotomy of which Pye speaks (1962), we may pothesize at least two different patterns of outcomes after coups, aending upon which image is chosen. One such image suggests that 2 military in the new African states are by comparison in the African ilian elites, incompetent, poorly organized and undisciplined, ecure in status, tribalistic and reactionary. Another image suggests t the military may, in fact, be the most developed and most profes- nalized elite group at this stage of the development of the new ltries; that it is, relatively, technically competent, impatient [the bungling and machinations of politicians, and both willing and to get on with the job of development. Other models of military es are described in the research literature, each of which would 16 imply another hypothesis about military behavior. (See, for example, Victor Alba's trichotomy (1962) of "the barracks groups," "the school I! officers," and "the laboratory men, as well as Morris Janowitz's aristocratic, democratic, totalitarian, and garrison-state models (1964).) These images of the military are discussed in Chapter Three, and several hypotheses are drawn for testing interpretations of mili- tary influences on foreign relations. Elite Analysis* The military holding political power are one kind of political elite, and so this research adds an increment to the long history of elite analysis. Two assumptions underlie this research. First, some individuals and grOups will have more influence in the political decisionmaking process than others, and those who have most influence are called the political elite. Second, the relationships among those who compose the political elite, between these persons and other persons in society, and their orientations to political problems will form patterns. These patterns can be conceptualized in terms of the structure of the elite and the behavioral norms of the elite. Patterns ill vary somewhat from situation to situation and from political tructure to political structure, but general types of elites, such as e military, can be identified by their distinctive structures and orms, which appear in a similar way in all political situations. *In this section, I am indebted to the summary of purposes and esigns in elite analysis provided by Carl Beck and James Malloy in olitical Elites: A Mode of Anal sis, University of Pittsburgh, 1971. '17 Elites are sets of persons who have relatively, a great deal of r and exercise relatively, a great deal of control. I measure tical power according to David Easton's model (1957) of politics he process by which values are authoritatively allocated in a ectivity. "Any individual's or group's power is therefore a tion of the number of alternative lines of action (or entry), or he qualitative character of lines of action (or entry), that that vidual or group has in relation to the authoritative value- eating process" (Beck and Malloy, 1971). Individuals or groups who . not exploited successfully their lines of entry are termed ntial elites. The actual elites are "those who have to some ure turned their potential into some degree of control over an ct or some aspects of the authoritative value-allocating process" 3.). The range of questions one can ask about elites is quite broad. Elite analysis runs the gamut from relatively specific studies of in elite in small-town U.S.A. (Hunter, 1953) to the highly {eneralized analysis of the changing composition of national elites >ver broad time-spans (Pareto, 1963). In some studies various :ocial aggregates such as the military (Huntington, 1957), members if legislatures (Matthews, 1960), bureaucrats (Bendix, 1949), and usinessmen (Mills, 1945), are singled out for specific analysis. ome studies attempt to discover the real elite (Mills, 1956, and unter, 1959). . . . Some studies emphasize the social character- stics of a segment of the elite with the assumption that these re determiners of political behavior (Matthews, 1954). Other tudies center on the character of social stratification within a >ciety, using elite-constituency relationships as an index of 1e character of the elite (Janowitz, 1956, and Kornhauser, 1959) Beck and Malloy, Op. Cit., page 2). Like Huntington's work mentioned above (1957), the present selects one elite social aggregate, the military, for analysis. lge of questions I ask, however, is quite small. I am not :ing to discover potential elites; I assume that holders of __,...—- —_;_ r___— ’— e, especially those who have taken office by force, are elites ding to my definition. My question is simply this: Do elites ified in one category, the military, show patterns of policy in 'gn affairs distinct from elites classified in another category, lians? There are two complementary approaches to describing elites and iining their behavior: (1) in terms of socio—economic background tcteristics, and (2) in terms of functional or skill characteris- While socio—economic characteristics have often been used as the ( L of description and explanation, a number of problems suggest that is the less usable approach to adopt in studying the military. , elite study based on socio—economic characteristics assumes that characteristics are a principal key to behavior, yet observation shows that persons of the same socio-economic background adopt ally different political postures (Matthews, 1954). Persons of ame socio—economic background may not be equally conscious of or of that background, or at least they may not see themselves as an st group. Second, naive explanation of behavior in terms of economic background may tend to overlook the influence of current stances surrounding behavior. It may be that elites face a t crisis of such intensity that their traditional conceptions of interests become ambivalent or multivalent or confused; or they ce an issue of such overriding importance to them that potential cts among them based on socio-economic differences are too much uxury to be entertained. Third, explanation in terms of socio-economic background may cause situations demanding behavior may engage different 19 ects of backgrounds of persons who fall within the same socio— nomic category. Typically, the categories developed to sort out sons according to socio-economic background are crude conceptions, en with culture—bound or value—laden implications, such as middle iSS, bourgeoisie, and proletariat. The persons presumed to fit the :egories simply may not fit with the implications that the category fries. For example, is there a proletariat in the United States as x predicted? And the categories are so gross and encompassing that y do not specify the causes of behavior. For example, the energy sis which leads to short supplies of gasoline and higher prices may y differently affect two middle-class Americans living side-by-side good neighbors, if one of them is a travelling salesman who depends driving a great deal in his job and the other is an accountant for arge oil company whose office is an easy bus—ride away. Fourth, explanation in terms of socio-economic background faces Eiculty when these conditions change. In modern societies, people a about and change jobs a great deal, and so categories based on raphy or occupation are not very stable. Sometimes such mobility s to changes in class status, and there may be ambiguity as to the nt to which previous or current situations, or both, will determine vior. Lastly, education may modify or confound behavioral tenden— based on socio-economic characteristics. Harold Lasswell has suggested an alternative way to identify explain the behavior of elites that supplements the traditional -economic classification. He advocates the use of functional or categories (1952), such as managers of violence, technocrats, mic managers, agitators, leaders of the masses, etc. These 20 gories are applicable across a number of polities. Such categories also be related to various levels of socio—economic and political lopment, providing thereby a basis for generating hypotheses of a ctural—functional and genetic nature. The functional needs of any polity will be influenced by four r factors: (1) the level of socio-economic development, (2) the of polity, (3) the political position of the polity,* and (4) the 1iques and mechanisms by which relationships around the decision— mg process are controlled. These factors, then, define the elite -s which will be most important to that polity. I propose to adopt the approach of functional and skill cate— es in the analysis of African military elites, for several reasons. , the military are, by definition, a skill—group, organized to rm similar functions in different polities. While socio-economic round may influence recruitment and promotion within an army, the aimed aim of many military organizations is to erase background rences and substitute professional attitudes and skills and ance to hierarchical authority. Second, in this comparative analysis among countries of similar mic and political development, we are defining a category of elites is not bound to specific cultures or other idiosyncracies of states. While it is doubtful that any person, including military I is ever completely divorced from the political influence of economic background, the military often make an attempt to change personnel in exactly this way. To the extent that the military *Revolutionary, post—revolutionary, stabilizing, conservatiz- tc. 21 successful, we should expect to find similar policy-orientations lecision-making processes in several countries. If we do not find similar orientations, it may be evidence that different socio- Dmic influences are breaking through the otherwise common military Jres; it might also be evidence of other differences in influences, as differences in decision-making situations faced by the various nes or differences in their non—military capabilities for dealing them. But our theory suggests that we will find common orienta— ; among military regimes; and if other influences produce idiosyn— .c orientations, this research will not be invalidated. Rather, .11 have specific knowledge about the limits of common military (tations of African regimes. Third, the military are a relatively precisely defined elite citly performing a set of decision—making functions in those nments where they have taken power. In all cases of coups in a, the military who have taken over have representedhmilitary asts and used military techniques, organization, and power. To my edge, in no cases have military leaders posed as representatives :ticular ethnic or socio-economic groups, although they may show endencies in these directions as days go on. The point is that, e of their words and actions, our first approximation to an ation of their orientations may be made in terms of their shared ry roles and skills. Chapter Four discusses these roles and and their probably influence on military men as decision-makers, lg hypothesized patterns of foreign policy for military regimes. As many students of politics have pointed out, the study of in new and developing societies, or "non-crystallized" ieties as Beck and Malloy term them (1971), is especially important. E. Black has argued that the "consolidation of modernizing leader- 'p" (1966) is the first stage on the road to modernization. He sees transfer of power from traditional to modern elites as involving litical struggles of the first magnitude," one source of the new dership espousing modernization being the new armies. This view ms to fit the African case well, for the "modernizing" military e rebelled and taken over governments in many states, and the uggle to consolidate this leadership continues. I discuss in later pters the special character of military leadership and the special blems of development in African states; the point here is that the solidation of modernizing elites is the crucial stage faced by most ican states in the 19505 and 19605, and the military may be a cial category of elites for establishing that consolidation. Theoretical Goals There are two general theoretical goals in this research. at, I hope to say something precise about the common orientations xfrican military regimes in foreign affairs. In particular, I thesize that, by contrast to civilian regimes, military governments tend to be less intensely but more parochially involved in diplo— 0 activities, to be more intensely and more parochially involved rade and receipt of aid, and to show less concern with the rhetoric ractice of alignment and nonalignment. Second, although this research and its conclusions are limited kica, I try to develop some general methods and concepts for the Sis of the foreign—affairs behavior of states and of types of tors in international affairs. My concern is to raise the discussion a level which will allow subsequent comparison with the foreign- airs behavior of military and civilian regimes at other times and other geographic areas. I accept the challenge of Przeworski and ne's "postulate of substitutability" (1970), which says that "the 'dge between historical observations and general theory is the sub- 'tution of variables for proper names of social systems in the course comparative research." In this research, then, I speak very little the foreign policies of particular countries--e.g., Nigeria, Ghana, on, etc.——but I organize the discussion in terms of categories of Intries (e.g., military and civilian regimes) and in terms of leral parameters of foreign-affairs behavior. I have defined three 2h general parameters--intensity, extensity, and alignment-—and these 'be used to measure the foreign-affairs behvior of any state, African other. In addition, I have combined these three parameters to ive an eight-fold typology of international actors which can be used help classify and interpret the behavior of any state. When I utinize the foreign-affairs behavior of a particular African state, 3 so by comparing that state with the ideal types in my typology, trends are discussed in terms of movement from one actor type to ther or across a "property space" which figuratively displays the ationships among actor types. 'Because Africa has been slighted in empirical work to date and muse we are deficient in rigorous theories of foreign-policy-making ew states, this research must be first descriptive and, second, Ipretative. My development of three parameters of foreign-affairs vior and eight actor—types should contribute to making this ussion more general. Each of the 34 countries in the study is ribed in terms of more than forty variables, which are sorted into ories of trade, aid, and diplomacy. Each country is described by ehavior on each variable for each of four years, 1964 through , and trends on each variable are calculated for the military and ian groups of states. These two groups are also described by intensity, extensity, and alignment patterns on the variable ories of trade, aid, and diplomacy. Selected countries, including the military governments, are analyzed in terms of the typology of r-types. Hypotheses are tested by comparing trends of the military :ivilian groups on parameter patterns and actor-types. Because this research provides an international overview, n one region, of foreign behavior, its conclusions come as ments of correlation rather than causation. As J. David Singer rgued (1961), the international level of analysis enables us to a comprehensive overview of the behavior of states and to study :ns of interaction which would be overlooked in single-country studies, although this level of analysis does suffer from the ;ary dearth of detail. "And though this may be an inadequate Ltion upon which to base any causal statements, it offers a [ably adequate basis for correlative statements. More specific— it permits us to observe and measure correlations between certain or stimuli which seem to impinge upon the nation and the behavior ns which are the apparent consequence of these stimuli."* *Przeworski and Teune come to similar conclusions in their dis— n of what they call "most similar systems designs," and they agree are times when research questions require such designs despite limitation to producing correlative, not causal, statements. use the problem under study here requires first establishing that e are associations or correlations between military regimes and ial patterns of foreign-affairs behavior, the international level alysis is the level at which we should begin. My goal in this research, then, is middle-range theory in s of the scope of the conclusions and the type of statements which be made. In scope, this research is limited to African states and comparison of military and civilian regimes' orientations in ign affairs. Statements of association (correlation) among omena can be firmly made, but arguments about causality can only entative, awaiting more detailed information at lower levels of ysis. I believe this understanding of what middle-range theory can do are precise and, indeed, more valid than the understanding of ain earlier writers, including those who introduced the concept. example, Harry Eckstein says middle-range theories are '. . . theories that go beyond mere description and common—sense leneralizations, that are based upon some explicit theoretical Frame of reference, that permit some rigor in formulating and :esting hypotheses, and that yet do not present iron—clad laws tr total interpretations of the meaning of social life" (Eckstein & pter, 1964:28). It seems to me that there is no such thing as "mere descrip- " The ingredients of any description are always chosen with an of their importance and interrelations. Indeed, as we refine our about any particular phenomenon, as we may discover underlying tures, we are likely to revise our definition and description of henomenon. It also seems to me that middle~level statements can on—clad" and even lawlike. Statements of correlation, for 26 nple, about the voting behavior of African military regimes at United Nations, may be quite precise and unerring, but still be I middle-range because they may fail to indicate linkage to other :ements'about the behavior of African governments or to statements 1ilitary orientations in other areas of the world. Thus, I accept )avid Singer's trichotomy of scientific knowledge into existential, ‘elational, and causal statements (1972), and I would place my : largely in his second category. This understanding is not :imilar from David Easton's much earlier concept of "synthetic or ow—gauge theory," contrasted with lower-level "singular generali— ons," and higher—level "broad—gauge or systematic theory." Easton narrow-gauge theory consists of a set of interrelated propositions that are designed to synthesize the data contained in an unorganized body of singular generalizations. But in the process of synthesis, the theory that is developed goes beyond the actual data included in the original cluster of generalizations. It becomes possible to understand not Dnly the phenomena to which these generalizations originally related, but also other phenomena which had hitherto been shrouded in doubt (Easton, 1953:56). CHAPTER II THE AFRICAN STATES: SIMILARITIES AND DIFFERENCES The Sample of 34 Countries There are at least 43 independent states in Africa today, ting only gg_jg£e governments, though that number can be challenged. :al territories under Portuguese rule are apparently in transition idependence; independence has recently been granted to Guinea- 1u, and the new Portuguese government promises independence for >ther colonies. Elsewhere, the legitimacy and even legal status {facto governments are in doubt, most clearly in Rhodesia. .ly, there is the question of whether to classify neighboring d states as part of the African group. The Malagasy Republic is ntionally included as an African state, but such convention is s clear for Mauritius. Thirty-three states on the African continent and the Malagasy lic are studied here for the period 1964 through 1967. The very nd small states, including Botswana, Buinea-Bissau, Gambia, 10, Mauritius, and Swaziland, for which data are not adequately able, and the very different, white-controlled Republic of South 1 and Rhodesia are excluded from the analysis. In the early ; of the study, the United Arab Republic was included in the 3, but was removed when it appeared on closer inspection not to 27 28 fit the definition employed here of either military or civilian regime. These 34 countries are listed in Table 2-1, divided into groups of ten countries with military coups d'état and twenty—four countries with continuous civilian rule through the four—year period under study. In this table, the countries are ranked according to their gross national products in 1966, the mid-point of the study, and their dates of independence are given in parentheses. Clearly, from this table, the African countries vary a good deal on the dimension of GNP; in fact, the highest GNP is 35 times greater than the lowest. By what logic may one call such a diverse group a distinct class? Several notions are borrowed from Russett (1967) and Alker (in Russett gt gl., 1964). First, a "class" of states must have relative homogeneity on certain internal attributes, mt there need not be integration among these states. That is, the :tates, however similar on attributes, need not be functionally inter- ependent or systematically interlinked to be called a class. Indeed, he notion of class signifies a set of separate units rather than a ystem of parts. Second, to define a class of states, an attribute r set of attributes must appear in one set of states but not in ther sets, or it must appear in a different amount in each set and Lth less variation in amount within each set than between sets. fird, the term "class" may refer to a particular configuration of tributes in a group of states which is differentiable from con- gurations in other regions. 29 mle 2-l.—-Thirty—four African Countries, by GNP (1966) Coup GNPl Non-Coup .geria 4603 geria 3040 2545 Morocco ana 2492 ngo Kinshasa2 1730 1561 Libya 1483 Ethiopia 1400 Sudan 1114 Kenya 1014 Zambia 1005 Ivory Coast 940 Tunisia 830 Tanzania 811 Senegal 734 Cameroon 709 Uganda 707 Malagasy Republic arra Leone 361 358 Mali 295 Niger 289 Guinea 251 Chad er Volta 245 229 Liberia 0 209 203 Malawi omey 193 tral African Republic 187 184 Gabon 162 COngo Brazzaville 1ndi 154 143 Mauritania 129 Somalia 128 Rwanda N=24 lGNP is in millions of U.S. dollars, measured in 1966. The have been culled and cross—referenced from several sources, the ary sources being the United Nations Statistical Yearbooks and the Agency for International Development's Selected Economic Data for Less Developed Countries. 2The name Congo Kinshasa is used throughout this research for iepublic of Zaire since that was the name for the country during rears under study. 30 For our set of African states, all these aspects of the definition of class apply. First, the states are separate, sovereign entities only loosely linked economically or socially. Second, on a great many measures of social, economic, and political development mese countries show relative homogeneity as a group and difference Tom other regions. By and large, on these measures of development, frica in the 1960's appeared to be the least-developed of the so- alled less—developed areas of the world. This leads to the third lass characteristic: it is not just one or two or even three variables aat show Africa to be less developed than other regions, but nearly 1e whole gamut of measures that are commonly reported in the litera- lre. The unique configuration in Africa is the consistency in lowest :vels of development, in a wide variety of measures, as compared th any or all other geographic areas. The data presented in Table 2—2 show that, in regional average atistics, Africa is clearly different from the other developing eas of the world on measures of non-political development. These Jional comparisons, assembled and published periodically by the Lted States Agency for International Development, regularly show :ica to be the least developed of the world's geographic areas on hpolitical variables. The data also show Africa to be a huge tinent, sparsely populated, but with a steadily growing population. has the greatest availability of agricultural land and the lowest e of food production. Only in GNP per capita, among all the per- mance variables, does Africa show better performance than another ion, and even here Africa's performance may become the worst, since 3 1 , Table 2-2.--Comparisons between xv“ " ' t ‘ r ‘ and Less ‘ i ’ Areas (1970) Developed Areasa Less ‘ ’ Non-” ' L Are as East Latin Near East 1 United C Tota States Total Africa Asia America & S. Asra Population Total in millions 660 205 1,800 290 310 260 860 Annual growth, in percent 1.0 1.1 2.6 2.6 2.7 2.8 2.5 People per square mi e 53 57 70 27 180 34 190 Urbanization, in percent 71 74 29 2O 26 56 23 Land Total area, in 1000 square miles 12,300 3,615 25,500 10,800 1,700 7,700 4,600 Agricultural land, in a of area 39 47 30 33 16 28 36 Agricultural acres per capita 4.6 5.2 2.7 7.8 0.6 5.3 1.2 Gross National Product Tot GNP, in billions of dollars 1,870 931 373 42 53 120 117 Current Growth rate, in percent 3.8 1.2 6.4 5.4 8.2 6.3 6.0 GNP per capita in dollars 2,850 4,584 210 145 175 470 140 Food Production Production index in 1970d d 119 113 124 113 125 127 125 Per capita production index, 1970 111 104 103 94 104 104 105 Electric Power per capita, KWH/year 4,850 7,640 210 100 150 510 130 Health Life expectancy, in years 71 71 51 45 51 59 50 People per physician 730 650 3,400 17,000 4,380 1,500 4,260 Education - Literacy, in percent 96 98 38 19 57 59 28 Students as a of 5-19 age group 76 85 38 24 48 52 34 aGenerally the industrial countries of Western Europe, United States, Canada, Australia, New Zealand, Japan, and South Africa. bAfrica excludes South Africa; East Asia excludes Japan; Latin America includes 19 republics. cIncludes countries not in regional totals. dIndex based on 1961-1965 = 100. - Source: Selected Economic Data for the Less Developed Countries, U.S. Agency for International Development, Office of Statistics and Reports, June 1971. 32 the African population is growing fast and its regional GNP is the slowest-growing of all the regions. While the ranges of and variations from these averages within each region are not given in Table 2-2, the averages are fairly com- parable across the regions because only a small fractiOn of African states show scores closer to the averages of other areas than to those of Africa. Also, in none of the variables in this table do African countries show as great a range and variation as on GNP and on popu- lation size. Thus, while African countries are not very homogeneous in GNP and population size, they are quite homogeneous and distinct from other regions in economic, health, and other social variables. The foregoing table presents indicators of size and economic and social development and modernization; it does not directly tell umch about political development. Of course, economic and social indicators are properly seen as important conditional or contributory variables for political development, although it could be argued that a poor, rural, illiterate society could be considered politically developed if political functions were carried out satisfactorily and efficiently, and there existed a sense of political identity and loyalty. Such a society could meet the kinds of indicators of political development that Potholm gives--(1) effective linkage, (2) wide political participation, (3) differentiation of the struc- tural and functional aspects of political institutions, (4) flexi— bility, and (5) rationality-~since these are not dependent on indus— trialization, urban living patterns, literacy, or other essentially "modern" measures. Such a view of political development is closer 33 :o Potholm (1970), Apter (1965), Almond and Verba (1963), and Geertz I1963)--authors who emphasize political culture and ideology and heir congruence as between masses and e1ites—-than to analyses based m institutional complexity (Fred Riggs, 1964), economic size and omplexity (Rostow, 1960; Organski, 1965), or types of elites Shils, 1962). Of course, in the modern international context, such View of domestic political development may be too narrow or even nadequate, since a small, poor, rural, illiterate society, however ell-developed politically within, may not be able to defend itself Jainst a hostile and powerful environment. The issue of how to define political development need not : resolved here. It does seem important, however, to distinguish m sets of indicators of political development which go beyond the cial and economic measures given above. On the one hand, develop- nt is indicated by measures of the modernity of the political lture, capabilities, and activities of the masses. On the other nd, development may be indicated by the modernity of the goals, yles, legitimacy, power, and skills of political elites. The rst speaks of measures of civic competence, political participation, litical identity, and means of interest—articulation and interest- Jregation. The second speaks of what ruling regimes seek, allow, encourage, and how they do that. The first set of measures is mrtant to the hypothesis that Africa is a distinguishable class states politically, as the data in Table 2—2 have suggested from ial and economic background characteristics. The second set aks directly to the hypothesis under research——that military 34 regimes differ from civilian in patterns of foreign relations——not to the hypothesis that Africa is a class of states. This second aspect of political development points to the heart of this research and is treated primarily in Chapters IV, VI, and VII, where military and civilian regimes are contrasted. Being able to specify certain measures of mass political ievelopment--the first set of indicators distinguished above——only Leads one to the frustration that there are not, so far as I know, good regional or national statistics available for comparison of the Erican states, among themselves or with other regions. The best mproximation is probably to be found in the relatively "soft" data I Banks and Textor (1963), including variables of political encultu- ation, interest-articulation, and the development of non—political ureaucracies. The following five tables of these variables show a onsistency of traditionalism in African politics unrivaled by any ther region. able 2—3a.--Politica1 Enculturation in Four Underdeveloped Regions Region Highly Integrated Moderate Low N “RICA 0% (0)b 55% (16) 45% (13) (29) :ia 14% (2) 29% (4) 57% (8) (14) min America 7% (1) 27% (4) 66% (10) (15) ddle East 9% (1) 18% (2) 73% (8) (11) Average 6% (4) 38% (26) 56% (39) (69) aData are from The Cross-Polity Survey, by Arthur S. Banks d Robert B. Textor (Cambridge, The M.I.T. Press, 1963). bThe numbers in parentheses are the sample—sizes. 35 Table 2-4a.--Interest-Articu1ation by Associational Groups, Four Less-Developed Regions Region Significant Moderate Limited Negligible N AFRICA 0% (0)b 0% (0) 13% (4) 87% (28) (32) Asia 12% (2) 0% (0) 29% (5) 59% (10) (17) Latin America 9% (2) 18% (4) 59% (13) 14% (3) (22) Middle East 0% (0) 20% (2) 30% (3) 50% (5) (10) Average 5% (4) 7% (6) 31% (25) 57% (46) (81) aData are from The Cross—Polity Survey, by Arthur S. Banks and Robert B. Textor (Cambridge, The M.I.T. Press, 1963). bThe numbers in parentheses are the sample-sizes. Table 2~5a.--Interest—Articulation by Non-Associational Groups, Less- Developed Regions Region Significant Moderate Limited Negligible N 82% (27)b 18% (6) 0% (0) 0% (0) (33) AFRICA Asia 94% (17) 6% (l) 0% (O) 0% (0) (18) Latin America 0% (O) 53% (9) 47% (8) 0% (0) (17) Middle East 82% (9) 18% (2) 0% (0) 0% (0) (11) Average 67% (53) 23% (18) 10% (8) 0% (0) (79) aData are from The Cross-Polity Survey, by Arthur S. Banks and Robert B. Textor (Cambridge, The M.I.T. Press, 1963). b . . The numbers in parentheses are the sample-Sizes. 36 Table 2—6a.--Interest-Articulation by Anomic Groups, Four Less- Developed Regions Region Frequent Occasional Infrequent Very Infreq. N AFRICA 12% (4)b 66% (21) 22% (7) 0% (O) (32) Asia 33% (5) 60% (9) 7% (l) 0% (0) (15) Latin America 0% (O) 71% (10) 29% (4) 0% (0) (l4) niddle East 44% (4) 44% (4) 12% (1) 0% (o) (9) Average 19% (13) 63% (44) 18% (13) 0% (O) (70) aData are from The Cross—Polity Survey, by Arthur 5. Banks and mbert B. Textor (Cambridge, The M.I.T. Press, 1963). b The numbers in parentheses are the sample-sizes. able 2-7a.--Character of Bureaucracy in Four Less-Developed Regions Region Modern Semi-Modern Post—Colonial Traditional N FRICA 0% (0)b 22% (7) 75% (24) 3% (1) (32) sia 0% (O) 67% (10) 6% (l) 27% (4) (15) atin America 0% (0) 100% (22) 0% (0) 0% (0) (22) iddle East 0% (0) 50% (4) 0% (0) 50% (4) (8) Average 0% (O) 56% (43) 32% (25) 12% (9) (77) aData are from The Cross—Polity Survey, by Arthur S. Banks and mert B. Textor (Cambridge, The M.I.T. Press, 1963). bThe number in parentheses are the sample-sizes. bles 2-4 and 2-7 were originally presented in this form in John Collins' unpublished Ph.D. dissertation, Foreign—Conflict Behavior d Domestic Disorder in Africa, Northwestern University, 1967. 37 There is another side to the argument that Africa is a special class of states. On the one hand, I argue that Africa is a unique and advantageous class of states for studying the possible special foreign policies of military regimes, and that argument is grounded in the discussion and data presented above to show the homogeneity and speciality of this class. On the other hand, this distinctiveness of Africa sets it apart from other areas, so that the findings from a study of military policies in Africa may not be generalizeable to )ther areas. Coups and Coup Countries The frequency of military coups d'état in Africa raises two mestions about that class of states. First, are those states which ave experienced coups d'état importantly different from the states hich have had continuous civilian rule, thereby undercutting the rgument that Africa is relatively homogeneous class of states? That s, are there really two distinct classes of states in Africa, one Asceptible to military coups d‘état and one not so susceptible? An iswer to this question may be approximated by studying the social, :onomic, and political situations of the two groups——coup countries 0 non-coup countries--and comparing them as groups. If the answer (that the groups are systematically different, that they are sub— asses distinguished by more than that one group is made up of litary regimes and the other of civilian regimes, then it will be at difficult to isolate a set of policies peculiar to military Jimes, as distinct from civilian regimes. 38 Even if we find no systematic difference between the states which have military regimes and those which do not, another question remains. Are there peculiarities within the group of coup countries whiCh would make it impossible to isolate military influences? In particular, are the leaders of coups so different in personal or corporate backgrounds and support, or so different in political intentions, that the resulting military regimes are not comparable for policy analysis? An answer to this question may be approximated by noting the backgrounds of coup leaders in the ten countries under question, the extent of corporate support in their take-overs, and their announced intentions as well as their behavior in office. If the answer to this question is that the African coups and their leaders are widely divergent in motive and situation, then it will be quite difficult to ascertain any special military influence on policy. Indeed, by definition of any great differences in personal interests, corporate situations, or announced motives, the discovery of a special military phenomenon becomes impossible. Because infor- mation is simply not available to portray definitively the personal attributes of these military leaders, their coups, or their corporate situations, and because the causal link between being military and making particular policies has never been systematically explored, this study can continue. If a set of policies or behaviors peculiar to military regimes is discovered, then it may be appropriate to dig more vigorously for information with which to compare the military regimes with one another. But if no such policies or behaviors are discovered, then the whole notion of special military politics will 39 e disproven, and the difficult task of comparing military regimes ith respect to personnel or corporate structure and function or hatever may be superfluous. Tables 2-8, 2-9, and 2-10 present group averages for twelve ocial and economic variables for the coup and non-coup sets of ountries. Also, statistics are given for Mann-Whitney tests of xtent to which each variable differentiates the two groups. U scores ere computed from rankings of all 34 countries on each of the vari— bles, then these U scores were transformed into z scores according o the formula given by Sigel (1956, page 121). The probabilities iven are those associated with these standardized z scores, and they ndicate the probabilities that the distribution of coup and non—coup roups would appear at random on these variables. 1e 2-8a.-—Coup and Non—Coup Group Averages and Mann-Whitney U Statistics Annual Rate Group Population of Population Population Urbaniza- x 1000 Growth, % Density tion, % >up 10,520 2.27 33.70 7.17 >n-Coup 5,429 2.50 17.41 7.95 score 128 99 165 127.5 Score .3024 -.7939 1.686 .2835 pbability .3812 .2136 .0459 .3883 aGroup averages are computed from data presented by L Gray wan in his book The Dilemmas of African Independence, 1968, revised ition, New York, Walker and Company. The full set of data is esented in Appendix III. 40 able 2-9a.--Coup and Non-Coup Group Averages and Mann-Whitney Statistics Land Agricultural GNP Group Area Land (acres GNP per (sq. mi.) per capita) capita cup 272.80 4.70 1321.4 109.0 on-Coup 270.29 9.78 717.6 118.3 score 94 63.5 131 100 score —.9829 -2.018 .4158 -.7561 :obability .1628 .0218 .3387 .2247 aGroup averages are computed from data presented by L Gray >wan in his book The Dilemmas of African Independence, 1968, revised lition, New York, Walker and Company. The full set of data is resented in Appendix III. ble 2-10a.——Coup and Non—Coup Group Averages and Mann—Whitney Statistics Electric Inhabitants Students as Group Power per Literacy % of 5-19 per capita Physician Age~Group up 77.3 33,768 17.2 22.1 n-Coup 113.5 24,444 13.9 25.4 score 100 124 119 111.5 score -.7561 .8546 .4448 -.3213 Dbability .2247 .1953 .3282 .3740 aGroup averages are computed from data presented by L Gray an in his book The Dilemmas of African Independence, 1968, revised .tion, New York, Walker and Company. The full set of data is asented in Appendix III. To a substantial extent, Aristide Zolberg's judgment in 1968 (page 71 in Bienen volume, 1968) seems to be borne out, that "it is mmossible to specify variables which distinguish as a class countries mere coups have occurred from others which have so far been spared." mough neither Zolberg nor any other analyst I have read presents tatistical analysis of the background characteristics of coup and on-coup states, ten of the twelve variables analyzed here show no ignificant discrimination between the two groups. Two variables do discriminate significantly: population density nd agricultural land per capita. No doubt these measures would be egatively correlated with one another, since countries with high >pu1ation density would logically have fewer acres per person than >untries with low density, and therefore would probably also have awer acres of agricultural land per person. These two variables :e more measures of population-pressure than of social performance ' even development, and so would not seem to indicate a crucial fference between the coup and non-coup groups of states. This view y be corroborated by the fact that the next most significant vari- la in these tables is a measure strictly of size—-that is, of land ea (p = .1628)-—and all of the performance and development variables e less significant. It should be noted how unreliable it would be to infer dif- rences between the two groups on the basis of group averages alone. 3 measures of population and GNP show quite different averages for 3 groups, with the coup-group averages almost double those of the I-coup group, yet in both cases the probabilities are better than 42 a in three (p = .3812 and p = .3387) that the distributions could re occurred by chance. The explanation for this phenomenon is 1nd in the unnormal distribution of states on these variables and a fact that coup countries appear at both extremes of these dis— .butions. On the other hand, among the other insignificant vari- .es, the coup countries tend to be grouped in the middle of the :tributions; this is the case with the percent of students in the 9 age group, population-growth rates, and levels of urbanization. these variables, there is a tendency for the coup countries to be istinct group, perhaps a set of states in agonies of either nsition or limbo between extremes. This hypothesis——that the coup up may be special in terms of theSe three variables--should not pushed too far, however. Two of these measures are not very useful ause so many states have tied scores, and on all three measures Eind states close to the extremes have experienced military coups :at since 1968. Thus, countries throughout these distributions vulnerable to coups. Figures 2—1 and 2-2 present cross-tabulations of the coup tries on population and gross national product versus GNP per ta, respectively. These figures show how the coup countries spread over a broad range of sizes. There is a tendency for coup group to be made up of two sub-groups, the first consisting ur larger states having higher per-capita products, and the d of six smaller states having smaller per—capita products. The a1 pattern of development is not different in these two 43 Population, in thousands High Medium Low (over (4700 (less 10,000) to than 8000) 3000) High (over Algeria Ghana u $200) 0 o t O u m m H fl Medium Sierra 2 % ($85 Nigeria Leone .3 U to T090 13’ 3‘) $120) CAR 2 94 m U) 8 Low Upper Burundi 0 ($70 Congo K. Volta Dahomey and under) \ Fig. 2-l.--Cross Tabulation of African Coup Countries on Measures of Population and Gross National Product per Capita 44 Gross National Product in millions U.S. dollars High Medium Low (over ($400 (less g $1700) to than 3‘ $1700) $400) ? High Algeria (over Ghana u $200) 0 c c o - u m m . H g Medium Sierra 2 % ($85 Nigeria Leone _3 U to Togo 1;; 3 $120) CAR 2 94 m U) a Low Burundi 0 ($70 Congo K. Dahomey and Upper under) Volta ig. 2-2.--Cross TabulatiOn of African Coup Countries on Measures of Gross National Product and GNP per Capita 45 sub-groups, however, nor does there seem to be a consistent dis- tinction in types of coups or stability of military governments. Most of the variables used in Table 2-2 to compare regions have been used in Tables 2—8 through 2—10 to compare the coup and non-coup groups. Those variables from Table 2-2 not used in the later tables were dropped from the latter analysis because there were not sufficient data for each of the countries. Similarly, so few countries of Africa are included in the Banks and Textor study, from vhich regional statistics have been drawn for Tables 2-3 through 2—7, :hat no adequate comparison can be made between the coup and non-coup groups within Africa on such political variables. I know of no >etter source for such political data. Table 2-11 lists the African countries which underwent dlitary coups d'état during the period 1965 through 1967, the coup able 2-1l.-—African Coups, 1965—1967 Country Date New Head of Government lgeria June 19, 1965 Houari Boumedienne ongo Kinshasa Nov. 25, 1965 Joseph Mobutu ahomey Dec. 22, 1965 Christopher Soglo entral African Republic Jan. 1, 1966 Jean-Bedel Bokassa pper Volta Jan. 3, 1966 Sangoule Lamizana igeria Jan. 15, 1966 J. T. U. Aguiyi-Ironsi July 29, 1966 Yakubu Gowon hana . Feb. 24, 1966 Joseph Ankrah urundi Nov. 28, 1966 Michel Micombero 390 Jan. 13, 1967 Etienne Eyadema ierra Leone Mar. 22, 1967 David Lansana, then Andrew Juxon—Smith 46 dates and the new heads of government who emerged. All these new government heads were military men, though not always the primary leaders of the coups which brought them to power.* Of the officers who took over governments in 1965 through 1967, most were still in power as early 1974. Seven of the ten men in this group who were heads of state at the end of 1967 were still in charge at the beginning of 1974. All those who fell from power were displaced by other military officers (Dahomey, Sierra Leone, Nigeria) or voluntarily removed themselves (Ghana). The most unstable government in this group of states, perhaps in all of Africa, has seen in Dahomey, where General Soglo led the first coup in December L963. Almost immediately, Soglo turned over the government to other :ivilians, without any structural changes. In December of 1965, he Ias either less forgiving of others or more ambitious himself, for .n this coup he took and attempted to keep full control of the "overnment for himself. Almost exactly two years later, he was verthrown by a coalition of younger officers who complained of ectional discrimination and incompetence on Soglo's part. At least *In Ghana, General Ankrah seems to have been recruited to ad the government after others, especially Lt. Col. Kotoka and lice Commissioner Harlley, completed the take-over. In Nigeria, jor-General Aguiyi~Ironsi actually resisted the January coup, and asserting control over the army in the south reached a position ere both the former cabinet and the coup leaders were willing to t him form a government. In the July coup, Lt. Col. Gowon was used a negotiator between the government and the rebellious junior ficers, and he took control of the government only after the more nior officers had either been killed or fled the country. In part, won agreed to take control in order to forestall the secession of e Northern Region; as a northerner from a minority tribe, he took adership with least prejudice from north or south. 47 o more coups have occurred subsequently in Dahomey, again at two- ar intervals. For the time period under investigation, however, eral Soglo was the controlling officer. In Sierra Leone, Brigadier David Lansana held control of government for only a couple of days, being then deposed by a up of younger officers headed by Col. Andrew Juxon—Smith. In eria, there were two coups d'état in the first half of 1966. In second coup, Major—General Ironsi, who had headed the government ce the first coup, was assassinated. Lt. Col. Gowon, who then k control, survived, leading the country through a very difficult i1 war and later into reconciliation and prosperity. He announced would not relinquish power before 1976. In Ghana, General Ankrah was forced from power after two rs as head of the National Liberation Council, after it was dis— sed that he had accepted money improperly. His military colleagues :he NLC maintained control of the government, but they made plans a new civilian regime, and, in 1969, under a new constitution ten by the NLC, Dr. Kofi Busia was elected the new Prime Minister. a's government was overthrown in December of 1971 by a group of officers headed by Colonel I. K. Acheampong. This group is rently unrelated to the NLC which had held power from 1966 19h 1969; indeed, the current NRC has detained some members of former NLC. With these exceptions, the military governments had all ved as of early 1974; at least, the heads of states had main— d themselves. Also, since 1967 eight more states of the 24 in 48 ‘the non-coup group (Table 2—1) had fallen to military coups d'état. In temporal order, they are Congo Brazzaville, Mali, Somalia, Sudan, ribya, Uganda, Rwanda, and the Malagasy Republic. While it appears that there is no particular type of African state especially susceptible to or protected from coups, one wonders whether there may be distinctive and different types of coups. Can we distinguish a variety of motives and circumstances in the coups which vould differently condition the policies which emerge from the new overnments? In particular, do the coups and their leaders differ mmortantly in ways that affect foreign policies? ‘ We can separate three schemes of classification: those of :ountries, of armies, and of coups. I have discussed a classifi- ation of states in the first section of this chapter, and have argued hat the African states can be seen as a relatively homogeneous lass. A classification of African armies has been attempted by ierre van den Berghe (in Welch volume, 1970), and it is not important 3 review here except as such distinctions are expressed through Lfferent coups. Three analysts have suggested typologies of military inter- mtion in internal affairs. Zolberg (in Bienen volume, 1968) dif— rentiates military strikes, referee actions, and take-overs. lch (1970) differentiates three types of military influence also, mely passivity, mutiny, and coups d'état. Lee suggests a four-way assification of the types of military intervention in Africa, based the composition of the army and what he calls the problems of the me. I have constructed Figure 2—3 as a representation of Lee's eme. Unfortunately, Lee does not label his conceptual categories 49 Extent of Corporate Action Broad Army Support Group or Sectional Limited Interests Access Competition for office prevents effective government MAJOR PROBLEMS Fig. 2-3.--Categories of Intervention nor discuss them in a clearly delineated way. Indeed, none of these analysts provides lists of countries or governments which are sup— posed to fit their classifications, nor operational measures for classifying countries. The spinning out of speculation can continue almost indefi— nitely, but so far it seems that classifications of coups follow from a few simple distinctions. On the one hand, we might classify coups either according to motive or according to the group involved. Robin Luckham (1971) distinguishes the two coups in Nigeria in 1966 according to group, labelling the first the majors‘ coup and the second the junior officers' and NCOs' coup. Classifications according to motive boil down to three motives: (1) In some cases, military officers take over governments because they want personal benefits for themselves or for the ethnic or sectional groups with which they identify inside or outside the army. Some say the first coup in Nigeria was a device to put Ibos in power. Some say that Col. Acheampong overthrew the Ghana government in 1971 because he was angry at the loss of certain 50 fringe benefits, such as a provided limousine. In Dahomey, it has been said that Major Kouandete overthrew General Soglo in 1967 because Kouandete, a northerner, believed that Soglo, a southerner, was discriminating against northern officers in the army. (2) In some cases, military officers may take power primarily to feed the corporate interests of the army as a whole. It has been argued, for example, that a primary reason for the overthrow of Ben Bella in Algeria was the desire of many officers to maintain the i power and integrity of the army separate from the politicians and their parties. In Ghana, the coup of 1966 may have occurred in part because the army was tired of being pushed into civilian work projects, and felt demoralized and weakened by the creation of Nkrumah's separate palace guard. (3) In some cases, military officers may take power primarily to meet broad, compelling national interests which they feel are endangered by the government in power. In Ghana in 1966, the army seems to have felt that the Nkrumah regime had spent the country into literal bankruptcy, undercutting development at home and destroying financial credibility abroad. Coup leaders in Nigeria and Dahomey also argued that politicians had been so corrupt and consumed by bickering among themselves that effective and efficient government was nonexistent. Depending on the motives and situations surrounding a coup d'état, a military regime may have problems in insuring its own Stability and the implementation of its programs. In the first case abOVe: the army is likely to be divided over the coup, and coup Ea";— 51 leaders are vulnerable to counter-coups by other parts of the armed forces. Programs coming out of coups triggered by personal or sectional interests are likely to be divisive and therefore resisted by some. In the second case above, the army is likely to be more unified in its support of the coup, and so the military government may be more stable than in the first case. Here we assume that the army has a monopoly of the use of force and, if united, is unchal- lengeable. However, if the coup does not result in the anticipated benefits, or if different parts of the army profit quite differently, then unity may break down. In the third case above, where personal or corporate interests of the armed forces are not so directly involved, and where the military men attempt to takeover the roles of national politicians, the support of the public and the pragmatic competence of the new administration may most determine the character and success of the coup. Whatever the triggers to the coups in the Congo or in Nigeria, and whatever the support of the army and the public at that time, it seems clear today that Presidents Mobutu and Gowon gained firm control 0f their governments because those governments led their countries through trying times and into significant prosperity. After all this discussion of types of coups, however, it may be that the classification of coups is not important to this study. At least, I would argue that work on conceptualizing the nature of Comps should come after, and be directed by, the findings of this StUdY- If, in the realm of foreign affairs, all the military regimes of Africa show similar policies, or similar modifications of old 52 patterns, then differences in the circumstances of their coming to power will be irrelevant. While it is always true that a researcher needs theory to guide his work--the random accumulation of facts does not create meaning--too much theory can prejudice research.* This research commences from the notion that the military, because of similarities in training, organization, function, experience, and so forth, heading similar states in Africa, will develop similar foreign policies. We must expect that there will be some differences, great or small, in their foreign policies, but this is not the place to attempt a definitive typology of coups and coup leaders. That would be to put the cart before the horse. If we turn back to Zolberg's article in which he argues that it is impossible to distinguish coup countries as a special class of states, we also find this judgment: Although the behavior of the military rulers varies according to their own backgrounds, the situations at the time of the take-over, which contributed to these events, and the problems they are facing, there are striking similarities in the way in which they are approaching the problems of government and in the instrumentalities they are attempting to create to implement their goals. (Zolberg, in Bienen volume, 1968:87). Although he does not quantitatively or even systematically defend this judgment, Zolberg is aware that the bulk of policies and lThis is the theme of Graham Allison when he argues that analysts are often captured by their conceptual frameworks, especially a model of rational action, so much so that they raise questions and come to conclusions more because of the way their concethal back- grounds directed them than because the situation under study held the meaning they found (Allison, 1971). Robert Jervis plays on the same theme when he speaks of theory as being both "necessary and dangerous" (Jervis, 1968). 53 practices emerging from military regimes may follow similar lines, despite background differences. He finds it "more difficult to discover common patterns in the field of foreign policy, where variations among countries were initially greatest" (p. 92), but, nevertheless, finds seven similar trends in the foreign relations of these military rulers. Several other caveats show the inadvisability of beginning with the classification of coups rather than of policy—patterns. First, there simply are not enough data available to classify coups with confidence. By their very nature, coups are clandestinely— planned and rapidly—executed events. This makes their unraveling difficult, especially from a distance, and makes detailed comparison across several states next to impossible. Also, by the nature of assuming governmental control, the new military leaders can all be expected to speak of the most exalted and impersonal motives and to hide or repress or overlook embarassing or difficult problems. Second, coups have multiple motives. It would be naive to believe otherwise. Indeed, it may be that a coalition powerful enough to topple a government can be organized only after a mul— tiplicity of grievances has accumulated, touching enough people to incite them to such dramatic action. In Ghana, for example, when Nkrumah was toppled in 1966, the new military council spoke of legal motives (change the constitution to restore the independence 0f the legislative and judicial branches and protect against develop- ment of an all-powerful executive), economic motives (cut out wasteful Spending and rebuild from national bankruptcy), and social motives 54 (root out corrupt politicians and the ethic which allowed politicians to see governmental work as an opportunity for personal profit). There may have been other, unspoken motives, as well. In any case, it would be most difficult to classify this regime until its behavior had confirmed these motives and displayed priorities among them. Third, the motives for coups may change over time. In Nigeria, Lt. Col. Gowon may have taken power initially in 1966 because there simply was no other officer available and acceptable to a majority of the army. Once in power, he had to give immediate priority to curbing the rioting and mass killing throughout the country. Later he moved to create twelve new states within the Federation in order to alleviate constitutionally the sectional and tribal tensions which had been so important to the coup. Then he had to recruit and lead a large army to meet the challenge of secession. Later, he turned attention to economic concerns, with oil reserve to exploit and an economy to stimulate. In other words, whatever the motives involved in a coup, situations change and prob— lems force themselves on governments, and the meaning of a particular regime emerges only as we can trace a pattern of decision and behavior. Fourth, even if motives and coup situations have differed in Africa, nearly all the military regimes which have come to power have attempted to stay in pOWer. Apparently they have come to believe that they can do a better job than any alternative elite in the country, at least for that time. Thus, they all seem to believe in their special competence or destiny, and, if they remain in power OVer some years, they find themselves faced with some general and 55 common problems, such as foreign relations. The longer they are in power, the better able we are to assess comparable patterns of policy and behavior, therefore, and to rise above idiosyncracies in coup situations which may hide similarities between regimes in the short run. The real judgment of what a coup means or what a military government represents must be made in terms of what the military regime does. Any government, military or civilian, may say one thing and do another. Or it may not even speak to all its own interests in the first days or weeks it holds power; and it cannot therefore be fully evaluated by its first utterances. Similarly, it is risky to judge a regime in terms of the backgrounds of its leaders. Should we classify a military government by the primary schooling, religions, ethnicities, economic classes, and so on of its leaders? The thesis that there is a peculiar military approach to politics assumes that the earlier background influences which led people to undertake military careers are accentuated, erased or, at least, largely offset by later military training, organization, and experi- ence. The potency and nature of these earlier or later influences-— educational, occupational, sectional, military, economic, situational, or whatever—-can be assessed only be observing and comparing behavior. Differences Between North and Tropical Africa While Africa tends to be made up of a relatively homogeneous set of states, some states do not fit the mode. I have already acknowledged the difference of the white-controlled states in the southern portion of the continent by not including them in the sample 56 under study here. A second area which may not fit the mode very well is North Africa. While the states of North Africa are certainly all young and developing, there are reasons to believe that they may be sig- nificantly distinct from the states of Tropical Africa. The states of North Africa have traditionally been separated from Tropical states by the great Sahara Desert. North Africa is ethnically more homogeneous and more unified by a single, evangelical religion, Islam. The French administration of the Mahgreb states was certainly dif— ferent from French administration of tropical colonies. In the North, the French looked on their colonies much more as integral parts of France-~Algeria being the most dramatic example——and the French cultural and economic influence was much more intense, more pervasive, and long-term than in the tropical colonies. Of the more than thirty colonies in Africa which achieved independence during the 19505 and 19605, only Algeria had to fight a war for its freedom. I have already eliminated the United Arab Republic from the sample under study because its government is a combination of civilian and military persons which will not allow the classification necessary for this study. Because of its other characteristics, such as its politico-cultural leadership in the Arab world, orientation Ito Middle East concerns, long—term independence, and relatively greater development, the UAR would also not fit very well into a single class of African states. 57 The states which provide a question mark for this study, then, are, west to east, Morocco, Algeria, Tunisia, Libya, and the Sudan. Of these, the first four are most alike among themselves and probably most different from their tropical neighbors. The Sudan is really a divided state, partaking of both Northern and Tropical African characteristics. The Arab northerners have ruled in this country, and during the 19605 there was a protracted and debilitating war between these elites and the black population in the southern portion. Because the black population is in the majority in this country, and because the Sudan has never been inte— grated into North African politics as thoroughly as the other states, the Sudan will not be considered a North African state. Tables 2-12, 2—13, and 2-14 present group averages for twelve social and economic variables for the North and Tropical African sets of countries. These are the same variables, and the averages are based on the same individual country data, as were used in comparing the coup and non-coup groups in Tables 2-8 through 2-10. Mann-Whitney U tests were not performed in the comparison of Table 2-12.--Comparison of North and Tropical Group Averages Gro Population Population Population Urbaniza- “P x 1000 annual rate Density tion, as of growth,% North 7514.5 2.35 18.00 18.70 Tropical 6848.1 3.06 . 22.76 6.25 58 Table 2-13.—-Comparison of North and Tropical Group Averages Land Agricultural GNP Group area, land, acres GNP per sq. mi. per capita capita North 458.25 8.25 2021.5 197.5 Tropical 245.90 8.24 931.3 104.6 Table 2—14.-—Comparison of North and Tropical Group Averages Electric Inhabitants Students as Group power per per Literacy % of 5-19 capita Physician age group North 151.25 8,152 24.7 35.0 Tropical 96.40 32,107 13.4 22.3 North and Tropical groups, however, because the two groups are so dif— ferent in sample size and the Northern group is made up of only four states. When interpreting these tables, it must be remembered that the simple comparison of group averages can give a distorted View Of the relationship of the groups, because of the unnormal distri— bution of countries. Nonetheless, some interesting and tentative observations can be made. First, those variables which best dis— criminated between the coup and non-coup groups seem much less dis- tinguishing between the North and Tropical groups. While North African countries seem much larger in land area than tropical states, Population density and the availability of agricultural land are much 59 more similar in the Tropical than the Northern group. Similarly, the average population size is about the same in the states of each group. On the other hand, in measures of productivity and development, it appears that the North African states are c0nsistently far more advanced than Tropical Africa, and in several variables the group averages are so disparate that it seems reasonable to infer signifi- cant differences between these groups, even without the aid of statistical tests. The North African group shows the following advan— tages: three times the urbanization rate of Tropical Africa, better than twice the average GNP, almost twice the GNP per capita, about 55 percent higher output of electric power per capita, nearly twice the literacy rate, more than 50 percent higher percentage of young people in school, and a four—fold improvement in the ratio of inhabi— tants to physicians. In sum, then, North Africa must be considered different from Tropical Africa. Comparing these averages with the data on Table 2-2, however, we may infer that North Africa is probably more like Tropical Africa than is any other region of the less-developed world. We must remember that the data in Table 2-2 are not fully comparable to the group averages presented in the later tables, because the data in Table 2-2 were taken at a later date, approximately five years later. Nonetheless, on urbanization, inhabitants Per physician, literacy, and the percentage of students, North Africa is clearly closer to Tropical Africa in averages than to other less-developed areas. In economic measures, such as the per capita production of electric 60 wer and GNP per capita, North Africa does show averages similar those of the next least developed area. Unfortunately, because the insufficiency of data, the North and Tropical groups cannot compared on political variables, just as the coup and non-coup oups could not. Because North Africa must be seen as significantly different 3m Tropical Africa in some ways, especially in economic development, 3 tests in Chapter VI for differences in foreign policies between 1p and non-coup countries will be run in two groups. Military and rilian regimes will first be compared while leaving the North :ican states in the sample, and then compared after the four :thern states (Morocco, Algeria, Tunisia, and Libya) have been loved from the sample. Discussion of the effect of removing the 'th African states from the sample will be presented in Chapters and VII. This two-stage analysis should illuminate the findings the development of foreign relations, and for that reason alone re is justification for keeping the Northern states in the analysis. addition, as William Zartman noted some years ago (1966), the :h and Tropical African states are comparable in their under- :lopment (perhaps especially in the particular aspect of politics .involves foreign policy); they are oriented to each other as raphical neighbors must be; they share allegiance to a surpris- y strong regional organization (The Organization of African Unity); Datterns of relations among these states have appeared which are me extent self—contained. Thus, to drop North Africa entirely 61 om the sample of African states would be to overlook an important pect of the politics under study. CHAPTER III FOREIGN RELATIONS THEORY Problems of African Foreign Relations With the achievement of independence, a new state is faced ith a host of apparent options in foreign relations, and, as for .11 states, foreign—policy decisions are linked to internal policies nd external conditions. Among the external conditions that African ountries face are the virtual control of world events by the big owers, in which African states may be pawns; tension and change in he structure of international relations; and distance from and poor ommunication with sources of power.* Karl Deutsch argues (l966:8, 1 Farrell volume) that the impact of external events and conditions :ould be said to decline with the stability and autonomy of the Iternal decision-making system and with the looseness of the coupling mween the outside environment and the internal decision-system." om either standpoint, the African states have been highly vulnerable the international environment. On the one hand, most African states We not yet achieved stability and autonomy; and, on the other hand, Lle these states are loosely coupled to the outside environment that they make little impact on it, they are quite dependent on side powers. In Deutsch's terms, then, African states cannot, at *Quincy Wright discusses these and other factors in detail, 62 ast in the near future, master or control the international environ- nt; most feel they dare not be isolated from it; and very few seem have adjusted their domestic systems so as to balance their depen- ncies abroad with adequate controls.* The general underdevelopment portrayed statistically in Part 1 Chapter II indicates how much African states are likely to be endent for a long time on external aid to make the economic, social, political transformations which virtually all espouse. Their lestic conditions and goals make foreign relations necessary because estic policy cannot be implemented by resources within the boun— ies of the African states.** Foreign policy is also a response the ideology which focuses and interprets domestic and inter— ional situations. This complex interplay among real conditions, icy-implementation, and ideology may vary from country to country. .s the argument of this research that African conditions are more >geneous than African regime ideologies. In particular, it is :ly that military regimes differ consistently from civilian es in their approach to foreign relations. Whether or not this hypothesis can be confirmed empirically, government must decide its priorities in foreign affairs. There ractical questions of which other countries to recognize, how and what types of diplomatic delegations to send and receive, *Potholm (1970) seems to suggest that Tanzania has gone est along lines which Deutsch believes would minimize external ts; still, he recognizes that Tanzania has developed a vulnera— y to Chinese influence. **This is the central argument of Robert Good's article, 1962, e Martin volume. ich international organizations to join, what bargains to make for 1de and aid, and so on. These decisions must be weighed in the [ht of such limiting factors as cost, the availability of trained rsonnel, the political, economic, and cultural implications of .ablishing or not establishing relations with certain states, and nature of internal political, economic, and cultural wants and s.* In this regard, most new states are confronted with the fact relations may be easiest and most productive with the former opole with which they have just broken and from which they are ampting to disengage. The question of "normalizing" relatiOns with the former ropole involves obvious risks and poses foreign—policy dilemmas. .e external assistance is necessary for domestic development, this ndence dramatizes the unequal, sometimes almost artificial acter of the new state. The African countries are thus prime ples of "penetrated systems," to use Rosenau's term (1966 article rrell volume), although certainly not always happy about this ng of decision-making. Whether the example be the United ns' forces in the Congo or the United states' advisers on ia's national planning board,** it is clear that non—members of ew states participate directly and authoritatively in the ation of values and the mobilization of support on behalf of *A good discussion of these options is developed by Patrick n in his unpublished paper read at the African Studies 'ation meetings in New York City, 1967. **See Wolfgang Stolper, Planning Without Facts. 65 als in these countries.* Indeed, national development goals are \ arently often inspired by external models; both communist and non- unist. At the same time, the African leaders may see these foreign ivities and influences as temporary, and as instrumental to their purposes, and for reasons of strategy or ideology they often gain for advantages from competing major powers. In addition, eign penetration--even where providing development resources—~may hologically and politically oppose or retard the process of ting new nation-states. The senses of separate nationality and elf-sufficiency, and the anti-colonial nationalisms of the Os and 19603, are the antithesis of penetration, and often are t promoted by aggressive policies which contrast the new state 1 other states. In this situation, foreign policy may become a mix or reflec- of two basic needs: First, the need for domestic solidarity— iing so that the state will be viewed at home and abroad as an 11" in the international arena; and, second, the need to solve specific, tangible problems that require the use of external rces. William Zartman has conceptualized such a dichotomy of :ives as "ideological" foreign policy and "national-interest" n policy.** *For further discussion, see articles by James Rosenau and IHanrieder in the APSR, December, 1967. **See Zartman's article in McKay volume, 1966, and his book, 1966. 66 Ideology is idealistic, activist, combative, revisionist, visionary, purist, maladaptive, and deductive; national interest is realistic, modest, constructive, conservative, evaluative, compromising, adaptive, and inductive. Foreign policy as a continuation of domestic, solidarity- ilding ideology might take the form of confrontation with real or gined external enemies-—as in boundary disputes, the "struggle inst nee-colonialism," and propagandizing. Or it might take some er form which works to identify and to establish the symbolic lity of the new state vis-a-vis other states, both for inter— ional and domestic audiences, such as assuming the role of mediator een power blocs in international issues. In contrast with this marily symbolic foreign policy is tangible, problem—solving policy—— tman's "national-interest" foreign policy-—which would better be sured in terms of levels of negotiation for trade and aid, technical istance, defense agreements, and the like. Both sorts of foreign policy may be carried out simultaneously :he same country, sometimes apparently opposing each other and times reinforcing. Thus, while Kwame Nkrumah denounced the ad States for neo—colonialism and intervention in Africa, be con— 2d for some years to request and maintain Peace Corps projects in . An example of somewhat better coordination of ideological ealistic policies may be Julius Nyerere's request for British to put down a Tanganyikan Army mutiny, even though they were replaced by African troops. Although Nyerere had professed 1 self—help, non—alignment, and independence as much as any list leader, he asked for the aid of the former metropole. 67 Thus foreign relations may be seen both as projections of arnal problems and as resources for solving internal problems. mix of motivations and actual policies of the African countries not been charted in these terms, however, although we do know a of the fairly consistent divisions among these states up to mid-19605. For one, Zartman has applied his distinction ween ideological and national-interest policies to the series of :a-continental alliances among West African states, from which 1re 3-1 below is derived (Zartman, 1966: Chapter I, International ltions in the New Africa). \ brily National-Interest Policy Primarily Ideological Policy. m fiNational independence Federalism, Pan-Africanism J,',,"Government politics Party politics 'flEconomic development Political development 3 Council of the Entente (1959) Mali Federation :y Coast, Niger, Upper Volta) (Senegal, Mali, Mauritania) Entente (1960) UAS mey, Ivory Coast, Niger, (Ghana, Guinea, Mali) Volta) Monrovia Bloc (1961) Casablanca Bloc ria, Sierra Leone, Nigeria, (Morocco, Algeria, Ghana, Ivory Coast, Dahomey, Niger, Guinea, Mali, Libya, UAR) Volta, Senegal, Mauritania) . 3-l.--Classification of African Groups by Policy-Orientation Dorothy Dodge (1967) has analyzed African voting in the Nations for the period 1960-65, and has confirmed the existence 2 major groupings of these states, two of which continue from s dichotomy above. Dodge's clusters are based on a 68 e-Beyle pair—agreement analysis of 220 United Nations General embly roll-call votes. Although these clusters do not have .ctly the same composition as the original Brazzaville, Casablanca, l Monrovia blocs, a substantial majority is the same in each case. lge says that Guinea, Kenya, and Libya do not show high cohesion :h any other African states or clusters in this period, while ritania, Central African Republic, and Senegal are closest to ster I, the Sudan closest to Cluster II, and Burundi, Malawi, and ia are unaccounted for. I have added some unclassified states odge's clusters, based on later voting—tendencies and other ign affairs behavior, and cross—tabulated these clusters, or s, with a trichotomy of African gross national products, pro— Lng Figure 3-2 on the next page. It should be noted that Figure produced a distribution, at least for the coup countries, almost (tical to that in Figure 2—1. Thus, it would appear that, for coup countries in the early 19605, there is a high correlation een bloc association and GNP per capita. Countries of highest yer capita aligned with the most "radical" bloc (Casablanca); :ries of moderate GNP per capita and also generally English— ing aligned in the more moderate Monrovia Bloc; and the poorest ties, which were also nearly all French—speaking, formed the [ville Bloc. The exact nature of the differences among the clusters and , in terms of issue-areas, extent of foreign commitments, and Ind aid, has yet to be determined in detail. In addition, both and Zartman's analyses deal with Africa before the post—1965 69 Gross National Product in millions U.S. dollars High Medium LOW (over $1700) ($400-$1700) (less than $400 5 0 3 (Algeria) Casablanca Ethiopia Guinea Bloc (Ghana) Mali (II) Morocco Somalia Sudan 2 4 3 Monrovia Libya Kenya Liberia Bloc (Nigeria) Tanzania (Sierra Leone) (III) Tunisia (Togo) Uganda 1 4 9 (CAR), Chad Cameroon Congo Brazza Brazzaville (Congo Ivory Coast (Dahomey) Bloc Kinshasa) Malagasy Rep. Gabon (I) Senegal Mauritania Niger, Rwanda (Upper Volta) . 3-2.--Cross—Tabulation of African States on Measures of Economic Size and Blocs )countries appear in parentheses. number I, II, III are empirically derived groups from research nited Nations roll-call voting, done by Dorothy Dodge and reported Erica Report, October, 1967. These groups overlap very closely the older Casablanca, Monrovia, and Brazzaville groups. di, Malawi, and Zambia are not classified because of lack of nation on their fit into blocs. 70 :h of military coups, and I have seen no comparable study of the .ntenance of or changes in the earlier alliances. If these pre-coup lsters did show generally different patterns of foreign relations, 2y would provide different baselines for the measurement of posts 1p changes in patterns, calling for different amounts of "correction" the military men who took power in each cluster, assuming the .itary interests to be the same in all states, which is the hypoth- .s under study here. Another indication of clear variation in African foreign- ‘airs activities is given by Patrick Mc-Gowan, using the Sudan and the Ivory Coast as illustrations. In 1963—64, the Sudan had embassies in 27 countries and belonged to 112 non-governmental international organi- zations; it did 13% of its foreign trade with communist states in the same years and in 1963 to 1965 voted 60% of the time with the USSR in the U.N. General Assembly; in addition, as of 1963 the Sudan was not an associate member of the European Economic Community and did only 24% of its trade with its main trading partner, the United Kingdom. Comparable data for the Ivory Coast reveal that the government maintained 12 embassies abroad, belonged to 83 INGO's, did only 1% of its foreign trade with communist states, voted only 30% of the time with the USSR, was an associate member of the SEC, and did 57% of its trade with its main partner, ’rance (1967: 1-2). McGowan has also developed a typology of African states which close comparison to Dodge's clusters; his sets and clusters lite similar, though based on completely different data and ic techniques. McGowan's clusters are presented in Table 3-1. tor—analyzing the extent of these countries' interactions with mmnist bloc and their general diplomatic interactions, and by the countries on scales of these types of international 71 de 3-l.--Decolonization Scores of Thirty African States Ranked by Size of Score and Grouped by Foreign—Policy Actor-Type tive-Independent Transitional Inactive-Dependent k State Score Rank State Score Rank State Score Sudan 9.9 1 Senegal —14.4 1 Mauritania - 8.4 U.A.R. - 8.3 2 Uganda 9.9 2 C.A.R. — 8.0 Morocco 6.5 3 Guinea 7.9 3 Chad - 7.5 Ghana - 3.1 4.5 Congo K. 7.6 4 Malagasy — 6.2 Algeria 0.8 4.5 Somalia 7.6 5 Sierra Leone 4.7 6 Ethiopia 6.7 6 Niger — 4.6 7 Tanzania 5.8 7.5 Congo Brazza - 4.5 8 Nigeria — 5.4 7.5 Dahomey — 4.5 9 Mali 4.4 9 Ivory Coast - 4.2 10 Libya 3.2 10 Liberia — 3.0 11 Tunisia 0.8 11 Togo 2.9 ’ 12 Gabon 2 .6 13 Upper Volta 2.1 14 Cameroon — 0.3 vity, McGowan found three consistent groupings of states which ames in terms of the intensity and independence* of their foreign tions. The extent to which the countries had "decolonized" their rnational activities—-that is, the extent to which they carried greater proportion of their relations with states other than former metropole--was generally correlated with intensity and )endence of activities. The decolonization scores given, then, neasures of deviance from predicted values, given the overall Elation between extent of decolonization and diplomatic activity .ndependence. Positive scores show greater decolonization than .cted, and negative scores show less. It is important to notice limilarity of trends in each group and to remember that the ____________________ *McGowan's concept of "independence" is similar to what I have d "Parochialism" and "extensity" in this study. 72 :cores are deviances from individually predicted values. Thus, the udan's score and Uganda's score are both given as 9.9, but this does ot indicate similar extents of decolonization, only similar deviances rom their different predicted scores. There are some internal similarities between the analyses resented above, indicating some consistency in African foreign—policy atterns in the period up to 1965, whether measured by ideological Locs (Zartman), by roll—call voting in the United Nations (Dodge), : by scope and independence of international diplomatic activities chowan). Furthermore, within each of the groupings that each :searcher has produced, we can record one or more successful .litary coups d'état, apparently indicating no strong relationship, 'any, betWeen policy-groups and the occurrence of coups. This formation is laid out in Figure 3-3 on the next page. Figure 3—3 shows a fairly high overlap between McGowan's pology of countries by levels and styles of diplomatic activity, the one hand, and alliance-blocs and voting clusters, on the her. The Casablanca Bloc-Cluster II states all fall within the hive-Independent and Transitional groups; the Monrovia Bloc— lster III states fall within the Transitional and Inactive— >endent groups, with a slightly higher tendency to appear in the idle, Transitional group; and the Brazzaville Bloc-Cluster I ltes also fall within the Transitional and Inactive-Dependent >ups, but with a very high tendency to appear in the latter. mgh the numbers within cells are too small to make elaborate parisons, it appears that McGowan's typology somewhat better age-4"- . Diplomatic Activity* Active- Transitional Inactive- Independent Dependent (50% coup) (18% coup) (36% coup) 4 4 Casablanca (Algeria) Ethiopia Bloc (Ghana) Guinea (II) Morocco Mali (25% coup) Sudan Somalia (50% coup) (0% coup) 5 3 m Monrovia Libya Liberia 0 Bloc (Nigeria) (Sierra Leone) 0 (III) Tanzania (Togo) H (37% coup) Tunisia m Uganda (20% coup) (67% coup) 2 11 (CAR), Chad, _ (Congo Cameroon, Congo BrazzaVille Kinshasa) Brazza, Gabon, Bloc Senegal (Dahomey), Ivory (I) Coast, Malagasy (27% coup) (50% coup) Rep., Mauritania, Niger, (Upper olta)(27% coup) L 3—3.--Cross-Tabulation of African States by Bloc and Diplomatic Activity (Percentages of Coup States Noted within Each Group) m Countries appear in parentheses. .assification by diplomatic activity is taken directly from McGowan's 'k (1967), "Factors and Correlations in African Foreign Policy." UHdi, Kenya, Malawi, Rwanda, Zambia are unclassified because of k of data. 74 lates the tendency for coups to occur than does the typology of cs and clusters. The tendency which McGowan's typology seems to gest is a greater likelihood for coups to appear at the extremes diplomatic activity, i.e., to occur in either active—independent inactive—dependent states rather than in transitionals. We should ember that these extremes of diplomatic activity also correspond Lh extremes of size, measured by land—area, population, and gross :ional product. Thus, the larger states tend to be the active- N: : {Lid a... a--. - . . f _ lependent states and have the highest rate of coups d'état (50% Figure 3-3), while the smallest states tend to be inactive- »endent states with a lower but substantial rate of coups (36% in 'ure 3—3). Middle-sized states tend to be transitional in diplo- ic activity and have a distinctly lower rate of coups than the er categories (18% in Figure 3-3). The relationship between categories of diplomatic activity the occurrence of coups is moderate at best. It suggests a stion not central to this research, the possibility that certain les of diplomatic activity may be very dissatisfying to the itary, so much so that this particular dissatisfaction becomes I important in the equation which accounts for military coups tat. This research is not concerned with accounting for the Irrence of coups, however. Rather, I seek to describe the foreign .cies which emerge after establishment of military regimes and L whether these policies are significantly different from the cies of civilian regimes. Confirmation of a hypothesis that the tary in every country or many countries show a basic similarity 75 influence on foreign policy would require showing similar trends foreign policy changes after coups in all or many of these Intries. The point of Figure 3-1 through 3—3 and Table 3—1 is it different groups of African countries are separable in terms of Eferent aspects of foreign relations. The Influence of Top Leadership in African Foreign Affairs A relatively small number of political leaders in the new ates may be expected to wield an essentially unrestricted influence foreign policy because of their role as nationalist, anti- Lonial symbols, because institutionalized processes of decision- cing which might constrain them are not yet developed, and because ase leaders and their followers appear to have a preference for :sonalization of relations. Zartman has noted that "foreign .ations are carried on primarily among party leaders rather than ng governments acting in the name of states representing people" 66 book: 144). Here we are speaking primarily of political tes--the executive officials and party leaders who appear as focal sonalities in the new states--rather than of the administrative, 'eaucratic, interest, or communications elites which might be ected to share in the process of policy-initiation and formation a more developed system.* These latter elites are not well eloped in Africa, where, in Cyril Black's developmental typology, *Gabriel Almond distinguishes these different elites as arable influences on foreign policy—making, in his book The rican People and Foreign Policy (1950). 76 1e states are in the throes of "consolidating modern leadership" L966). There are four parts to Zartman's argument which, though eveloped to account for intra-continental relations of North and ast Africa, seem generalizable to the relations of all African :ates. After noting the primacy of party leaders' activities in rreign relations, Zartman says that the second key explanatory Incept is power. For African states, that means the lack of power. least, compared to developed states, African states tend to have ttle power beyond the charisma and expertise of their leaders. As ese leaders approach international questions, they find they have w institutional capabilities to bring to bear and few restraints their personal behavior. Sanctions and force are either disallowed or impractical; domestic public opinion and pressure groups have little weight in decision-making processes and foreign affairs leaders frequently have little access to them; the possibilities for (giving) aid are insignificant; pur- chase, reason, and support, in a bargaining process, are rarely used because there are few clear goals, flexible policies, and national—interest criteria to which they can be applied. Finally, the informational input on which capability must be based is small (1966 book: 145). This personalism and relative institutional weakness in sign affairs lead to Zartman's third characteristic of developing ions' foreign policies, the primacy of ideology. Having no power but only trampled rights, having fallen under the domination of stronger states, and having overthrown the colonial system largely by manipulation of slogans and by default,. . . African states attempt to reject international relations based on power and Speak of a new era, one no longer characterized by the domination of the stronger over the weak (1966 book: 145). 77 Ideology is pursued and created by elites, not by the masses. But the high role of ideology leads us to a fourth characteristic of African foreign policy: its questing nature. That is, there is little systematic, clear, unified, or accepted ideology in African foreign relations. Definitions are still very much being sought, as are national interests and goals, let alone general acceptance of them or unity around them. Thus, much of African foreign relations is a search, through discrete actions, alliances, conferences, and rhetoric, for mental clarity and ideology. A corollary of this search is the predominance of ag_hgg_policy-making at the present stage of development, again encouraging and allowing primary influence and action by top leadership. A very few leaders--perhaps as few as one or two or three per country, such as Touré, Nkrumah, Nyerere, Houphouet-Boigny, and enyatta-—stood out in the 19603 as symbols of the new states for hom there seemed no alternatives. These were the leaders of the nti—colonial movements, and often the organizers of the intellectuals, he group whom Shils (1962) and Kautsky (1962), among other analysts, ave described as the motivating and successful force in creating the ew states. They have been characterized as intensely politicized, ationalistic, socialist, populist, uprooted, oppositional, and ncivil (Shils, 1962), and so, additionally, charismatic, intransigent, nd detached from both modern and traditional roles. Neither Shils r Kautsky was sanguine, in his writings about the capacity of such eaders to administer the new states, yet for lack of others these en symbolically are the states in some cases. Some of these 78 >parently very powerful, §Ei_generis leaders have been rather easily aposed to live in exile (e.g., Nkrumah, Azikiwe, Keita, Obote), It in all cases except Ghana the lack of any countervailing group is allowed the entrenchment of the military.* In the absence of an experienced diplomatic and legislative >rps, and because of the restriction of slim national budgets, there ; little challenge from seasoned elites or even developing ones, 1d the challenges that do appear are often treated as threatening ither than constructive. Zolberg argues that in terms of West frican one-party states, given the overwhelming importance of perSOnal leadership in all five countries, there may be no alternative to an understanding of the personalities and backgrounds of the men formulating the ideology, a goal which we cannot achieve given the present state of our knowledge of these men (1966: 61). .SO, acting in accordance with an ideological map which defines opposition as illegitimate, the ruling group attempts not only to neutralize and control visible political opponents, but furthermore to anticipate possible manifestations of opposition by establishing new electoral rules and coercive measures. Deprived of most legal channels for expressing discontent, various groups in the society engage in illegal action; the more dangerous this becomes, the more desperate their attempts to change the situation. ‘ This involves not only changing a team of rulers, but because the rulers are so closely identified with the political order, the entire regime (1966: 91). *Seven of the ten military heads of state in 1967 were still 0 charge at the beginning of 1974. In Nigeria, Gowon had said he Duld stay in power at least until 1976. Also, consider the descrip— ion of General Mobutu's control of the Congo in The New York Times rticles of 9/21/69 and 10/26/69. 79 Furthermore, there exist only a minimum of social and political structures and interest-groups which could bring influence to bear on leaders. The populations are so poorly mobilized—-even in the so~ called mass, one-party states, as Zolberg has shown (1966, especially Chapter One)-—that the politicians may despair more of social-political inactivity than of over-activity. This picture of personalistic and individualistic diplomacy is accentuated by the fact that the leaders of the African states are inveterate travellers, with an emphasis, especially within the continent, on relations by personal communica— tion.* The one group which must be considered an exception to these observations is the military, who have monopolistic, even if sometimes minor, physical power and an identity and interest which consolidates them. Yet the conclusion that there are few alternatives to the personal influence of a handful of leaders on foreign-policy forma- :ion in the new states is probably too simplistic. To the extent :hat foreign policy is a tangible and realistic problem-solving )rocess rather than an ideological cathartic--the alternatives dis— :ussed above-—it must be a policy with roots in conditions beyond he individual idiosyncracies of political leaders. While any olicy will have symbolic effects and may be subject to the whims f political elites, foreign relations may still be motivated by or esult in more tangible values. Membership in international organi— ations may provide a new state with nebulous "world status," but 150 with measurable technical assistance in agriculture, business, *L. Gray Cowan's observation in McKay volume (1966: 121-2). 80 :ommunications, and so on. Trade and aid with developed countries ray provide the motivational stimulus and material substance for hange. Also, one result of, if net stimulus to, foreign relations ill be the growth of new expectations of social needs and changes, ncluding the development of new interests and influences in foreign- olicy formation. Particularly among governments committed to hange, therefore, foreign relations provide not only a two-edged, deological and practical, instrument extending domestic politics, but 150 a source of interests and influences which may compete with the alitical elite. Thus, at least in time, foreign affairs may become so complex : have so much domestic impact that top leadership will not have a tee hand in its development or conduct. If events, situations, or roblems control leaders more than the other way around, then a vecial approach to foreign affairs which would distinguish military 'om civilian regimes will be increasingly difficult to isolate. This could eSpecially be a problem for research if national terests are real, objective, external things, as Hans Morgenthau ieves,* and these interests have been identified so clearly that ry alert regime knows its limits and what it must do at all times. this "rational model"** of foreign-policy—making, there would be tle leeway for idiosyncratic elite behavior or even for policy lecting special corporate interests such as the military. * See the first chapter of any edition of Morgenthau's itics Amon Nations. **I use this concept as developed by Graham Allison in ence of Decision (1971). 81 If, on the other hand, national interests are not objective things to be discovered, only goals to be wisely formulated (see Roger Fisher, 1969: 10), then personal or corporate interests of elites which may condition that wisdom become important. In other words, if there is much uncertainty and genuine disagreement about African foreign affairs, there cannot be full rationality in policy— making. Policymaking will inevitably involve other processes and calculi, including organizational processes and politicking among iecisionmakers, as Graham Allison has shown (1971). Allison has proposed an "organizational—process" model of foreign-policy—making that sees foreign policy and behavior more as the output of certain Lnstitutionalized routines or patterns than as choice based on the :areful and informed calculation of gains and losses clearly related :0 an explicit and consistent value-system. Seeing governmental ehavior and policy as output rather than choice reflects a View of "overnment as a conglomerate of semi-autonomous, loosely allied rganizations, each with a substantial independence and life of its wn. Each of these existing organizations has a fixed set of tandard operating procedures, programs, and interests which condition he way gOVernment perceives problems, defines alternative solutions nd consequences, and grinds out responses. Policy often comes to be afined by the largest or most pervasive organization within or ahind the government, or by the values and routines of the most lfluential group or groups plodding through a problem, or by the EPartment which first perceives or is assigned a problem and incor— mates it into its ongoing routine and world View. Allison has also proposed a "governmental-politics" model that s somewhat at variance with the assumption of rational decision- aking based on objective "rational interest." In this model, policy 5 seen as outcome or output of various overlapping bargaining games ong relatively rational players arranged hierarchically in national overnment. Here, policy is a function of the perceptions, motiva- ions, positions, power, and maneuvers of individuals who are volved in a well-calculated competitive game for power, income, estige and similar values. Policy is not chosen, nor even pro- ssed, by institutional routines intended to define and advance a ational interest," but rather emerges as the result of compromises, )alitions, competitions, and confusions among individual rational fficials who have personal perceptions, priorities, and stakes in Le issue. While we cannot resolve here the issue between Hans Morgenthau d Roger Fisher of whether national interests are ultimately objective ings, external from individual whims or interests, or the creations wise institutions or leaders playing the art of the possible, this estion is not crucial to this study. In Africa, probably more than other region, national interests have not been fully or unequi— ally defined. African states, with their generally recent entry 0 formal diplomatic relations, their underdeveloped foreign-relations titutions, and other impediments to rational policy making, are ely to formulate policy and act in international affairs more in s of Allison's two models. Although these models obviously do not aust all possible views of the policy—making process, they are the most fully developed in the literature and seem particularly sug- gestive here. To wit, if African foreign-policy-making better fits the governmental-politics model, the primary influence and actiOn of top leadership is to be expected. If policymaking better fits the organizational-process model, that would occur where substantial corporate bodies can be identified; the most likely and most visible is the army, although a legislature or civilian interest—groups may complicate the picture. Thus, where the military take over the government, any special and systematic approach to foreign affairs should be evident, whether policy processes fit an organizational— process or a governmental-politics model. Foreign-Relations Theory There are two sides to foreign-relations theory, quite asymmetrically treated in the literature. On the one hand, we may speak of the sources or causes and conditions of foreign relations, and, on the other hand, of the outcomes or events or results of foreign—policy behaviors. Innumerable books and articles have been written on the subject of sources of foreign policy and behavior, but :omparatively little on the subject of outcomes. Thus, in spite of he fact that virtually all authors attempting to explain foreign elations complain about the paucity of general theory, there is in act a plethora of approaches to such theory, while at the same time We need more here is little discussion of what is to be explained. esearch into the styles and patterns, even details, of both foreign- °1iCY inputs and foreign—policy outcomes. 84 For the development of quantitative theory, this is a serious :oblem justifying this current exploration of patterns of outcomes 5 African states. Not only must concepts operationalizable as uriables be developed, but the variables of any comparative theory )st occur in a sufficient number of regularly recurring instances to bilitate statistically significant accounting. If, e.g., a very i rtant outcome in foreign relations is wa£,* and that outcome is dependent variable to be explained, the empirical researcher may e some difficulty counting enough comparable wars to make much dway in quantitative theory. If the outcomes to be explained are Lisions** made by the foreign office, State Department, or other éncy charged with foreign policy, then, no doubt, there are a great y occurrences to be counted. But, of course, the question arises: ch decisions? Some substantive classification of decisions is Jired to grasp conceptually the outcomes and link them to under- ldable antecedents. It is the purpose of this section to present ’erent approaches to conceptualizing the inputs or sources of ign policy and relate these to various types of outputs. Several typologies or theories of foreign—policy inputs are .able in recent writings. It is useful to see these typologies rms of the levels of theory important to this research. I have iy indicated that the hypotheses and conclusions to be developed *This is the orientation of Kenneth Waltz's book, Man, The and War, 1959. **This is the orientation of Joseph Frankel's book, The Making eign Policy (1963) and the work of Snyder, Bruck, and Sapin 85 (ere are limited to Africa, but ultimately one could imagine this heory being incorporated, first, into a more general theory of the oreign relations of developing areas, and, second, into a completely eneral theory applicable to any state or area. Figure 3-4 displays his substantive cumulation of theory which is a scientist's goal. General Foreign Relations Developing Africa Fig. 3—4. Substantive Cumulation of Theory One of the difficulties of some typologies is that they are differentiated by their relative applicability to developed as ,nst developing areas, let alone to Africa. Some typologies are d on varieties of issues in foreign relations, others emphasize ings of values, others the levels of a variety of inputs to :ion-making. At least three dimensions are necessary to classify ogies so as to sort out their relevance to Africa: differentiation ) system—level, (2) issues, and (3) state or area configuration. .rst of these dimensions classifies system—levels, the second only nominal distinctions, and the third may be used either 11y or ordinally. 86 Among system-level typologies, one would include those of Jacobson and Zimmerman (1969) and Rosenau (1966, in Farrell volume), which distinguish systemic, environmental, societal, governmental, and idiosyncratic or psychological approaches to explaining foreign policy. One could also include the trichotomy of Brecher, Steinberg, and Stein (1969), which distinguishes power-theorists, decision- making theorists, and input—output theorists. They point out that most power—theorists deal primarily with the worldhsystem level or with the level of the national state; the main stream of decision- making theory tends to deal with the small-group or individual idio- syncratic level; and input-output theory often deals with the inter- national or national levels. Issue—approaches to analyzing foreign policy include Rosenau's four-part set of territorial, status, human-resources, and nonhuman- 'esources issue-areas (1966, in Farrell volume), and also Brecher, iteinberg, and Stein's military-security, political—diplomatic, conomic—developmental, and cultural-status issue-areas (1969). Kurt ondon implies an issue typology when he deals with the following actors in foreign-policy decision—making: military capabilities, Dmestic opinions, economic structure, national security policy, and 1ministrative influences (1965). Kenneth Waltz selects an outcome Fprime interest——namely, war-—and deals with it in terms of three evels of influences: human behaviors, internal structures of states, d the traditional international anarchy. Rosenau is one of the few theorists who propose state or area nfiguration typologies, deriving eight genotypical actor—states in 87 rms of three parameters: population size, degrees of economic velopment, and degrees of political accountability. Dichotomizing ese parameters, he derives the eight genotypical actors listed in gure 3-5. Parameter Combinations State Labels large, open, developed democratic superpowers large, closed, developed authoritarian superpowers large, open, underdeveloped democratic great powers large, closed, underdeveloped authoritarian great powers small, open, developed democratic middle pOWers small, closed, developed authoritarian middle powers small, open, underdeveloped democratic small powers small, closed, underdeveloped authoritarian small powers Fig. 3-5.--Rosenau's Eight Genotypical Actors Rosenau argues that each of these types of state is faced :h its own structure of problems in foreign policy. There is, in :t, a logical kind of cross-tabulation among actor-types, 0n the hand, and the salience of different issue-areas and system-levels, the other. To a large extent, the kind of state an actor is tates the issue-areas which will be of most concern to that actor the system-levels which will wield the most influence. Within 5 rather deterministic context a state must "adapt," to use enau's notion (1970); that is, cope with and try to benefit from international environment. Adding one more parameter--"pene- ion" by other states——Rosenau comes up with the Figure on the owing page of sixteen kinds of actors, each with its ranking respect to potencies of system-level variables in various issue— 88 . .I _ 1 (1.11114!) .Hrom .mm . $me .395 muwmuofiaa :Mwuwosfiuoz . £002:va :ouuum suumm .m an cause» .muauaaom . . can on . as = suaaom amauuom we moauoore can moauowaeuoua= .smonmom .oz magma "mounom ammoumawsmmH mo mocha. ounce can woauoaoom uo mags noouxam 5." mwaunwuom 033.30% .3..an on onwpuogg possum mwwuwwoom mo ”Simeon Hmcuflxw on» 9336an moaofluugy uo muom minimum .mum ill mOmv ouomu o 0mm m 0mm omomm emowm a w .m H a ..n omomxm enema m o m u owm anomam omonhm a a a M a ._.. o m on m m on 0mm H 0mm u m m on m m on m on m m 0mm >m>mom m m cm hmomhm 0mm hm omhmm >m>m>m hmhmam o m m u n u u u u u u on new. on H ..n w a a A. com >m and u m m m xmwmaw Oman» Summon m m on m m m m com 3033 .ammmhm a 4.. a hmhmMm a .m a anamhm u u w xmhmom u u on u H u u u u u M on u u u u" H H ..n w ..n cum H on?“ M as." hmhmam “an... mmxmkfi wn..» zmhmxm nun mmhmam saw. «as, Ha.“ Haw uHu nu... “new uuom 1 Ii] 0.. on. .u. 0.. ow“ on“ mm“ mm“ was was ,ow. cu. . s on. on. . mom mmn Mom mmm mm: WW2 “We Hue Que Hum mwm mwfl mwm “on mom mwfi aw. an...» em: a .4 a .4 e .4 n m mum 9H1 94.4 a .4 e .4 9W3 9W3 am. We 1mm .mm 1mm 1mm .mm “mm 2%. .ms “mm 1%“ “mm 1mm 1mm 1m“ 1mm a m E nu. “an nun was mun sum a.” a.“ sun mum as“ “an a“. n“. a“. “m. 3:” “as was 919 one era are 91% era v.19 are axe min ”I... a z a n s.. ... .s. 5s. 3s. as. 5s. .s .s. .s. 5s. 5s. .s. s.” “as was s s o o o o o o o s s o o o o n n n n n n o o o o s m m m m . . . s . . m m m m m m a . a a a a a a a a a . a o . e a e a e a a e a m m a. v. e e a. n n n n n u n w 1 z m n n m m e e a a a a v. n. u. w w. s e a a 1 E E a. E E 9 e 9 w w. canny» euuuuu uaunuu uwumuu uoumnu caucus vouuuu filuulurlln m a locum mouauu locum poumuu locum uwumuu mama couch» locum uuumwu locum couch locum coumuu v we“ .. 1 Inc locum 1.62 locum ncoz nwnom 1.62 :05 I: I n uoumuu coz once 2 o oz 05m 182 05m :62 locum iii one suaaoa ammo uumoao suaaom :uao unmoao sud o and o .8 .H m o oomoHU 3.30m some aeonoom woo o>wo if! saonoom oumoaopoeuoean u a seasons vwooau>wuuovan saoaoom eaaoaoaoa r suucsoo Hanan asou we 89 By Rosenau's definitions, all the African countries would fall within the economically underdeveloped categories. In six of the eight types of actors which are economically underdeveloped, idiosyncratic elite variables at the national level are expected to be most important in determining foreign policies in all issue-areas. In the other two, international—system variables play the topmost role. One would infer, then, that all African countries would have foreign policies which reflect the anxieties, aspirations, percep— tions, talents, and personal experiences of their political elites more than they reflect any other kinds of variables, with the exception of certain small states which are heavily penetrated and thereby strongly influenced by internation systemic variables. A few African countries might be considered large in geo- graphic and physical resources (for example, Algeria, Libya, igeria, and Congo Zaire Kinshasa), and for them international-system ariables would be most important after national elite idiosyncratic ariables. Most African countries are small, however, so these ystemic variables are likely to be important along with elite diosyncratic ones. Even for these small countries, though, nternation systemic variables are expected to be next most important. at should be noted is that virtually no African country would be xpected to have foreign policy much influenced by institutional r societal variables. Rosenau's View corroborates that of Zartman sported in Part Two of this chapter, then, portraying the elites as ery frequently and deeply insulated from or unable to read public ?inion, and as presiding over amorphous and weak governments. 90 In sum, the top leaderships of African governments can be (pected to define, and even largely carry out, their countries' )reign policies, mainly by following their instincts, interests, 1d their perceptions of the roles in which they find themselves. -the extent, then, that the political elites of these countries n be said to share goals, skills, and perceptual screens, one would pect similar foreign-relations patterns. Military regimes may ve this commonality because of the inculcation of particular per— nal values, the development of certain skills, the orientation the armed forces' routines, and the military's definition of vernmental roles. It is the purpose of this research to test for a monality of military policies as distinct from civilian. There are actually two levels of outcomes in foreign— icy-making at which we might look. First, there are decisions en, agreements and bargains made to seek certain goals. For mple, the Nigerian government decides to try to enter the Common ket, the Tanzania government decides to seek international funding a transnational railroad, the Ivory Coast government decides to i the invited "dialogue" with South Africa, and so on. These lsions are about agreements or bargains, or about preferences-- it the ordering of values that have external or international )nsions. The decisions, themselves, however, are not automatically ized in practice. Actions must be taken to achieve the goals ed upon. The second level of outcome, then, is the action taken plement policy. While it would be useful for the development mparative theory to observe simultaneously the deliberations of 91 African governments with regard to all foreign—policy questions, is, of course, impossible. One can observe and count only the ond level of outcome, the actions which result from earlier isions. Of course, any government takes myriad sorts of actions that e foreign—policy implications. Tariffs are raised or dropped, sons are built and deployed, votes are cast in the United Nations, Lomatic recognition is extended or ambassadors are recalled, eign investment is discouraged or invited, foreign nationals are Lted or deported, hostile or friendly remarks are made, and so on. : of us believe that even much more informal events also indicate eign—policy decisions. For example, embassy cocktail parties are an supposed to reveal sensitivities, preferences, and strategies, .heir invitation lists, their timing, their conversations. And, rly, the internal events of nearly any country can be seen to :international repercussions. The conflict in Vietnam came to oil the United States far beyond its domestic political capac- “; the largely symbolic changes within Czechoslovakia in early [brought the military intervention of the Soviet Union. ) Out of such a variety of actions, one needs to isolate major iers or dimensions for analysis. Rosenau has suggested the means ctivities of foreign relations, can be dichotomized into ble and intangible aspects (1966: 85-86, in Farrell volume). %ility speaks of specific concrete, material rewards and costs. Fibility, then, must mean symbolic, diffuse, relational, con- 1 T T l--in a word, psychopolitical-—costs and benefits. We can 92 orate this dichotomy slightly int 0 more traditional dimensions: e, aid, and diplomacy. Though there are obviously rough edges ach of these categories, the following figure shows the approximate eptual relations. Tangible Intangible Trade I Aid Diplomacy Fig. 3—7.—-Categories of Foreign Behavior These latter three categories--trade, aid, and diplomacy—- be used as the conceptual boxes for foreign—relations activities 3 observed in this study. Trade is measured by monetary values :ports and imports. Aid is measured by governmental loans, grants, Inity payments, and the like. Diplomacy is measured by many Iators which do not directly involve the transfer of material 7 examples are votes in the United Nations, memberships in national organizations, numbers of diplomats sent abroad and ved, attendance and participation in international conferences, visits, exchanges of letters, propaganda, and official press- ses and speeches. These three traditional dimensions of foreign—affairs behavior 1tuitive1y and historically derived. It would be preferable, in ltitative dissertation, to derive dimensions empirically from exhaustive list of all conceivable external activities, but that [dous—-and, perhaps, conceptually impossible-—job is beyond this Suffice it to say that the published work which best flea» ‘ 93 (pproximates such a job shows dimensions of external behavior which .argely overlap the parameters and measures given above, except that lid is not as important on world-wide consideration. The following Figure shows the major patterns of external behavior which Rummel [erived from factor analyses of behavior—indicators on random and :elected samples of dyadic relations from a population of nearly all he world's countries. Patterns 1, 2, 4, 5, and 10 are not neces— ‘arily official government activities or even derived from such, lthough they may be. All the other patterns may be encompassed by y general dimension of "diplomacy." Major Patterns Indicators Measured l. Salience Tourists, A to B 2. Emigration and Communications Emigrants, A to B/A‘s population 3. United Nations Voting Weighed UN voting-distances 4. Foreign Students Students, A to B/A's students to all nations i. Exports Exports, A to B/A's GNP E International Organizations B's/A's IGO membership ). Official Conflict Behavior Military violence score, A to B ) Diplomatic Representation Embassy or legation, A to B/A's [ total embassies or legations ) Self-determination Voting UN voting-distance on "Self i Determination" issue patterns - Anti-foreign Demonstrations Anti—foreign behavior score, 1 AtoB Fig. 3—8. "Diplomacy“ lrce: R. J. Rummel, "Indicators of Cross—National and International Patterns," American Political Science Review, March, 1969, Page 131. 94 The Concepts of Foreign Affairs Intensity, Extensity, and Alignment Even though we decide to measure, to count particular indi- >rs like those above, there remain the questions of where to look how to arrange the counts. Logically, just three parameters are assary to describe the patterns of occurrence of any foreign-affairs Lvity at a particular time: First, the frequency or scale of 1rrence; second, the area of the world social system with respect which it occurs; and, third, the extent to which one country's set of countries' activities coincide with those of another country set of countries. I will label these parameters “intensity," :ensity,“ and "alignment." We have, then, the following cross-tabulation of parameters IhiCh to measure foreign relations, representing substantive :nsions and distributive or analytic dimensions. Substantive Areas Trade Aid Diplomacy Intensity . . . . . . if; a o Extensity ; - - - . . "‘5 m a g Alignment ' ' ° ° ° ' Fig. 3-9.—-Categories of Foreign Behavior Analysis Intensity refers to the amount or frequency of a phenomenon. ntry with 2000 foreign—service officers stationed abroad is intensely involved in international relations than a country 95 which sends out 1000. A country which votes 100% of the time in United Nations roll—call votes is more intensely involved than one which votes only 50% of the time. A country which annually exports goods worth 75% of its GNP is more intense than one which exports goods worth 10% of its GNP and so on. As elementary as measures of intensity may seem, they can be ambiguous. In particular, intensity of involvement in foreign affairs may not be a result of similar choices in all countries, even when we are speaking of strictly gOVernmental activities. Intensity may vary grossly from one country to another simply because the countries differ in size. One would expect Nigeria to have more foreign service officers in the field than Gabon just because of their different population-sizes; with 50—60 million people, Nigeria will have more people available, more needs, and more resources to support international efforts than Gabon, whose population is barely half a million. Similarly, although Nigerian citizens enjoy a GNP/ capita only about one-third that of Gabonese citizens, one can expect igeria to export and import much more by value, than does Gabon. though Gabon's GNP per capita is three times that of Nigeria, igeria's GNP is 25 times that of Gabon. Level of development may ave an effect on intensity in addition to, or contributing to, the ffect of size. For example, Zambia and Malawi have about the same opulation—sizes, but Zambia is much more developed industrially and hows a GNP per capita about four times greater than Malawi's. We an expect the more developed Zambia to be generally more active in nternational affairs. 96 Thus, both aspects of intensity seem important to measure and :0 consider. Even if the differences in GNP make it appear "unfair" :0 compare the values of the exports and imports of Nigeria and :abon, we should not fail to appreciate that Nigeria i§_more visible, 1ctive, and probably more influential in the world social system than :abon. To a certain extent, the intensity of Nigeria's activity is in artifact of its size, undoubtedly; but that activity is still a :eal thing which will have ramifications domestically and inter— 1ationally, and which is subject to variation over time. Thus, I propose to measure and to compare the intensities of international >ehavior both in gross terms (e.g., the number of foreign-service >fficers abroad and the amount of exports) and in standardized terms (e.g., the number of foreign-service officers divided by the size ff the population, and the amount of exports divided by the GNP). 1 en I aggregate several variables to produce an index of intensity, 0th gross and standardized measured will be included. "Extensity" refers to the breadth or parochialism of a state's ternational affairs. While the number of foreign-service officers road may measure intensity, the number of countries and inter— tional organizations to which these officers are sent is a measure extensity. In this regard, countries will vary in the proportions the world social system which they touch with their activities, and the variety of states or international organizations they deal th. Two countries which have the same number of foreign—service ficers abroad may actually show quite different patterns of behavior. e of the countries may concentrate all its officers in embassies 97 a few rich states from whom aid or other rewards may be expected; 2 other country may spread its officers thinly around the globe ung many states and organizations for symbolic or other reasons. :t as a government must decide how many diplomats to send and reive, so it must also decide to and from how many states it will ead these diplomats. “Alignment" refers to the extent to which the pattern of one ntry's foreign behavior is similar to that of another state. The e similar the patterns, the more the countries are said to be gned. While alignment of any two countries might be assessed on tually any type of foreign—affairs behavior, two major alignment- terns are of particular interest here: alignment with one or ther of the big powers on Cold War issues, and alignment with er African states. These two general areas of alignment seem : important to the African states. The big powers, especially United States and the Soviet Union, quite disproportionately .uence international affairs, and therefore affect the African [tries whether the latter like it or not. The big powers have .ncial, technical, and military power which African countries can 1y ignore, and, since the big powers tend to be mutually antagon- c, the African states are always under pressure to choose between . The direction and tightness of such alignment is measurable in a, aid, and diplomacy. Similarly, the African countries speak very often as though would like to present a united front to the world. In union, : is some power or appearance of pOWer, for these generally 98 powerless states, and, in addition, the leaders of these states tend to feel a bond in the commonality of their situations, their geo— graphic location, and even their race. Since 1963, the pressure for unity has been institutionalized in the Organization for African Unity. That organization is certainly not all-powerful, but it has weathered many storms and lasted longer than many non-Africans predicted. It has organized substantial material and symbolic support for the liberation movements in southern Africa, and it has effectively mediated Several intra-African disputes. The extent, then, to which any particular state follows the will of a majority of the continent is quite important to all states on the continent. The breakdown of foreign relations into these nine analytic dimensions--intensity, extensity and alignment in trade, aid and iiplomacy--sharpens our understanding of this complex behavior and improves the precision of our description. As logical constructs, they provide ingredients for the development of cohesive theory. 3eing logically derived from general historical experience rather man emerging from peculiarities of the African situation, they can e used in the study of politics in other regions and can be vehicles Or the development of more general theory of foreign-relations atterns and their changes. A Typology_pf International Actors Not only can the concepts of intensity, extensity, and Lignment be used as parameters of foreign affairs; they can also 3 Combined to develop a descriptive typology of international :tors. Each of these three traits which define a state's 99 international activity can be conceived as a dimension with scale positions for every possible quantity of each trait. When taken together at right angles, these three scales form a three—dimensional space in which all actors can be located according to their degrees of intensity, extensity, and alignment. In Figure 3-10, this space has been closed to form a cube, the eight corners of which represent all possible combinations of the extreme values of the three dimen- sions. Thus, the corners of the cube represent extreme ideal types of actors with respect to intensity, extensity, and alignment. Few, if any, actual actors can be considered to correspond to these extreme types, but as the location in the cube of a specific actor approaches one of the corners, that actor can be treated as displaying activity in a manner similar to the ideal type.* For the optimum development of quantitative theory, it would be necessary to define numerically what "high" and "low" mean on each dimension and to show equal-interval, if not ratio, levels of measurement on each. Chapters V and VI speak to these problems, arguing that ratio-scales can be developed for these dimensions. Here our purpose is to use the figure as a heuristic device to facilitate visualizing the types of actors. To illustrate the location of an actor on the cube, consider the foreign-affairs behaviors of the following non-African countries. Inspection of United Nations voting records and data in the 1972 *The idea for displaying types of actors in a three- dimensional space was suggested by a reading Charles Hermann's "International Crisis as a Situational Variable,“ in Rosenau (editor) (1969) . 100 Consumer Vulnerable F G Low I I Alignmen I I I I Proté é gHi I B I CVassal Low I I I I E ________________ Inde ndent a P6 Fe 1e g /’ m ./ § / x / W / 2’ / /' /’ High / A D . High Low Calculating AggreSSive Intensity A. Aggressive B. Protégé 2. Bassal Calculating E. Figure 3-10.—-Typology of Actors in Foreign Affairs Independent Consumer Vulnerable Feeble 5",: , . , 101 edition of the World Handbook of Political and Social Indicators (Taylor and Hudson, 1972) shows that France, the United Kingdom, and Vest Germany all send out very high numbers of diplomats, spread :heir exports among a very large number of countries, and tend to lots with the United States significantly more than with the Soviet Jnion. Thus, these three West European countries tend to be aggressive .ctors in international affairs. In contrast, Iran and Pakistan show only a moderate intensity, eing ranked mid—way in the distribution of countries in terms of the umbers of diplomats they send abroad (ranks of 42.5 and 32, espectively) and in terms of the number of memberships in inter- ational organizations (77 and 30.5, respectively). They are among 1e top three ranks of countries in extensity, however, as measured 7 the number of countries to which exports are sent. They differ Jmewhat in their alignment patterns, although both are more aligned .th the United States than with the Soviet Union. Both receive mge amounts of aid from the United States, although Pakistan gets ch more and Iran gets nearly as much from the USSR. Both are mbers of CENTO, and Pakistan is a member of SEATO in addition. us, we might classify Pakistan as more calculating than aggressive d Iran as somewhere between independent and calculating. Canada is an interesting and fairly easy state to classify. shows high intensity in terms of diplomats sent abroad and its nberships in international organizations, and it is quite closely Lgned with the United States in ideology and security arrangements. shows surprisingly low extensity, hOWever, being one of the ten 102 countries with the highest concentrations of exports. For Canada, this means a very high proportion of its exports go to the United States, reinforcing a diplomatic alignment. Canada, then, looks something like a protege of the United States. Our typology of international actors can provide more than descriptive pigeon holes, also. Differences in actors' styles imply differences in values placed on foreign affairs by the government, differences in influence with other governments, differences in the flexibility and capacity for change in policy and action. On a world—wide perspective, we may find that African countries simply cannot attain levels of intensity and extensity that are possible for other states. Also, each actor-type may require certain national resources or characteristics to support it; Gabon may not be able to support 100 embassies abroad, and Ghana may not have been a big enough power for Kwame Nkrumah to use his position to try to mediate the Vietnam War in 1966. Some actors may be playing world—system roles which do not fit them, or for which there is tension between the role and the country's most important characteristics or needs. The degree of fit or tension between actor-type and country—type may tell us something about the impact and success of foreign policy and of the regime. In sum, while elites are expected to have great latitude in making foreign policy for African states, and the actor—type which emerges will largely reflect the governing elite's idiosyncracies, a country can become over- or under-extended or -involved or -aligned. Each actor-type in our typology implies certain risks or capacities behind it. 103 "Aggressive" actors are characterized by or tend toward high involvement and forwardness in international affairs. They are relatively vigorous, assertive, and sure of themselves, tending to be ideological bullies. They are more likely to take initiative, to assert leadership, and to use force than other actors. They are apt to be pugnacious, enterprising, and pushy. To carry out such a role in international affairs, a government must be relatively stable and strong and have a strong economy behind it. The aggressive role is a committed, visible, and potentially influential one; it can also imply inflexibility and risk if there is not power to back it up. A "protégé" state is highly involved in international rela- tions, but largely limited to dealing with one of the great powers with which it is aligned. The intensity of its involvement implies domestic power; its high alignment and concentration of interactions with the alignment-leader implies special concerns which do not allow it to become more independent or take leadership itself. There may be some important limit to its power, or it may be a state in flux, or there may be some special historical, ideological, or other relationship with the alliance leader. Because of the protégé‘s real or potential power, however, the alliance leader is likely to be much interested in it, as well as vice versa. In sum, the protégé is a state under the patronage, protection, or other care of a great power interested in its destiny or welfare; the protege is a state with tensions, committed internationally for the time but perhaps having the potential to go it alone someday. It may hope to reap the 104 rewards of alignment in the short run, but grow to independence or leadership later. A "vassal" state is aligned, like the protege, but is much weaker. It is little involved in foreign affairs, and that little activity is likely to be concentrated upon a big power. Probably, it is aligned in order to get payoffs to offset its weakness; thus, it tends to be trapped in its situation, inflexible, uninfluential, and unadaptable. In return for the crucial resources provided, the vassal gives homage, fealty, or other service to the big power; the vassal is a subject, subordinate, follower, or retainer. The "calculating" actor is generally weak, like the vassal, as evidenced by the low intensity of its foreign relations, but it is likely to be less resigned to its position. It is highly aligned with an aggressive, big power-—probably for the same reasons as the vassal state-~but it does not limit its relationships to the alliance- leader. Its alignment may indicate that it is an ideological actor, but it will not let principles destroy it. The high extensity of its actions may indicate that it is shrewd, scheming, or prudent, unwilling to be caught in total dependence on the alliance-leader. While the vassal encounters risks of being overlooked or squashed because of its physical or symbolic weakness, the calculating actor runs the risks of spreading itself too thin or alienating the alliance- leader because of its behavior far afield. The calculating actor is likely to be less predictable and more flexible and influential and adaptable than the vassal, but, because of its weakness, less so than the protege. 105 The "independent" actor is not much influenced or controlled by others. It is highly involved in foreign affairs and spreads its activities among many other actors, but it is not oriented to any bloc. This position implies strength of its national regime and economy, but not assumption of a leadership role in the world system. It can be flexible and adaptable, because of its strength, but its influence is irregular and uncertain, because of its distance or independence of action. Like aggressive states, it must have real strength to back up its intensive and extensive activities, but, unlike aggressive states, it is relatively uncommitted ideologically and therefore probably more flexible in foreign affairs. That flexi— bility may not imply leadership, however, since it is apt to be self-centered and unconcerned with fitting in with others or orga- nizing others. What influence it does have is likely to be very largely related to the apparent domestic strength which underwrites its intensity and extensity. "Consumer" states show high intensity, but low extensity and low alignment. These states are interested in much foreign activity, but not in ideology or alliance. Probably their orientation is essentially materialistic; their involvement abroad is limited to states where they assume that the financial or material rewards will be greatest. If their high intensity reflects real domestic strength, they may have a viable posture in international affairs, but if their high activity reflects substantial domestic needs, they they may be forced into an alignment. As consumers, they are probably uncommitted and probably uninfluential, more interested in amassing wealth for 106 themselves than in ideological questions or others' problems. A Absorbed in their own narrow concerns, consumers need real physical strength to protect themselves and to adapt to changes in the environ- ment. "Vulnerable" states have the least desirable foreign affairs posture of all. Their low intensity shows weakness, and their low extensity shows dependence on a few others. They have all their few eggs in small baskets, and get little or no payoff from alignment with a big power. They are easily susceptible of being wounded or hurt, open to attack, and difficult to defend. Their weakness and nonalignment suggest a vacuum which invites foreign intervention, also. They are uncommitted, inflexible, and uninfluential, perhaps not viable for long as sovereign states. "Feeble" states are physically weak, enervated, or infirm. Like vulnerable states, they get no payoff from alignment with a big power, but unlike vulnerable states, they make efforts at relations with a wide variety of states. In this latter regard, they are like the calculating actors, attempting to develop options and visibility, although they are unwilling to make alliance commitments. They are likely to be spread too thin, and thus be ineffective. Depending on the type of actor a state is, it may find it relatively easy or difficult to change its policies and activities in foreign affairs. New regimes which emerge in it——as in military coups d'état--may find international postures more or less easy to change and more or less to their liking. Military regimes, in Particular, may feel uncomfortable with certain types of 107 international activity and tend toward certain positions on the cube of actors developed here. These military orientations are the sub- ject of the next chapter. CHAPTER IV THE MILITARY AND FOREIGN RELATIONS THEORY The Armed Forces of Africa Military requirements and military techniques have always influenced foreign policy. Indeed, national military establishments are often created and justified primarily with foreign relations in mind. They are supposed to be guarantors of security from foreign threats. Perceived military weakness can decisively limit the flex- ibility and credibility of other foreign-policy instruments, themselves often underdeveloped in the new states. Because of the military's missions and capability, the military can be expected to be a potent influence on foreign—policy formation. The military in new states are likely to have strength of organization and virtual monopoly of force, as well as a stake in the relations with other states, both for security and supply reasons. Despite their importance, the armies of African states are comparatively small. As the table on the following page shows, very few exceeded 10,000 men by 1968, and, in 1964, nearly a third did not exceed 1000 men. The larger armies are found in states which have clearly identifiable military interests at home or abroad: in the Arab countries, no doubt in large part because of the general Middle Eastern tension with Israel; in the Sudan and Nigeria, which were involved in protracted civil wars; and in Cengo Kinshasa, where 108 109 military problems began in the first week of independence. Ethiopia may be an exception, although it had border clashes with both Somalia and Kenya in this period. J. M. Lee explains that Ethiopia has "a long tradition of military retainers" which would inflate the size of the military (1969: 6). Army size seems to have no relationship with the occurrence of coups d'état. Looking at the figures for 1966, one finds that the ten countries where the military took over governments during the period 1965-67 include the country with the smallest army--Central African Republic--and the country with the second largest army—— Algeria. There may be a slight curvilinear relationship between army size and the occurrence of coups d'état, since four of the coup countries appear in the lowest nine ranks of army size and six of these countries appear in the thirteen highest ranks, with no coup states in the middle twelve ranks. A clearer association with army size is demonstrated by Figure 4-1, which cross-tabulates army size with population size, at the top of the page, and with gross national product, at the bottom. Each of the three variables has been trichotomized into high, medium, and low groups of roughly equal numbers of states. The top Part of the table shows a substantial relationship between army Size and population size, with more populous countries having larger armies. Such an association is hardly remarkable; the table is more useful for picking out the few states which do not fit the pattern. Figure 4—1 shows even more clearly that army size is positively related to size of gross national product. The cross-tabulation of 110 Table 4-l.-—Sizes of the Armies of African States (Coup Countries in Parentheses) Country 1964* 1966** l967*** l968*** (Algeria) 48,000 48,000 58,000 57 000 (Burundi) 800 950 2 , cameroon ,000 1,600 . 2,700 3,500 3,800 4 350 (Central African Republic) 500 600 1 000 1,100 Chad I I 400 900 1,500 2,650 Congo Brazzaville 700 1 800 1 900 2 200 . I I (Congo Kinshasa) 30,000 32,000 35,000 38:250 éDghomey) 1,000 1,800 1,750 2,250 G:b::pla 30,000 35,000 40,000 45,400 600 750 450 1,050 éGhana) 8,000 17,000 14,000 15,900 Iuinea 4,800 5,000 5,000 5,400 Kvory Coast 4,000 4,000 3,500 4,500 L::::. 2,500 4,775 5,000 5,400 la 3,580 3,200 4,000 4,150 Libya . 5,000 7,000 8,000 15,000 Malagasy Republic 2,600 4,000 3,750 4,500 Ealaw1 n.a. 850 1,086 1,150 “Sui t _ 3,100 3,500 3,000 3,650 r1 ania 500 1,000 1,400 1,530 Morocco 34,848 44,800 n.a. 50,000 N19er . 1,200 1,200 2,000 2,100 (Nigerla) 8,000 11,500 12,000 163,500 :wanda 900 1,500 1,500 2,750 enegal 2,500 5,500 5,000 5,850 (Sierra Leone) 1,850 1,360 1,200 1,600 :omali Republic 4,600 9,500 11,000 12,000 udan . 11,000 18,500 26,500 27,450 Tanzania 2,000 1,800 5,000 7,900 (Togo) 250 1,450 700 1,250 Tunisia 20,000 17,000 23,000 21,050 Uganda 2,000 5,960 7,000 5'700 United Arab Republic 120,000 180,000 n-a- 213'000 (Upper Volta) 1,000 1,500 1,700 1,800 Zambia 2,900 3,000 3,900 4,400 **Source: Africa Report, January, 1964. 1*Source: Adelphi Paper 27. *::::ource= J. 14. Lee (1969) WEE—121995. . ource: Claude E. Welch, Jro, Regimes: A Typology and Analysis of Post-C Africa"; calculated from original figures 1 Forces of African States (London: \——————— "Radical and Conservative Military oup Governments in Tropical n Richard Booth, The Armed Adelphi Paper #67, 1970). Population in thousands dollars Gross National Product in millions U.S. Fig. 111 SizDS of African 'es High Medium Low 48000-9000 8000-3000 2000—600 7 1 1 (Algeria), UAR High (Congo K.) 60000-9000 (Nigeria) Kenya Tanzania Ethiopia Morocco Sudan 2 8 5 CamerOOn, Mali Chad Medium (Ghana) Guinea Malawi 8000-3000 Tunisia Ivory Coast Niger Malagasy Rep. Rwanda Senegal (Upper Volta) Uganda, Zambia 1 2 8 Low (Burundi),(CAR) 2000—600 Somalia Libya (Dahomey),(Togo) Liberia Congo 3., Gabon (Sierra Leone) Mauritania 8 0 0 (Algeria), High UAR, (Congo $5000—1400 K.), (Ghana) (Nigeria) Ethiopia Morocco. Sudan 1 8 1 Cameroon, Kenya Medium Tunisia Ivory Coast Tanzania 31300—400 Walagasy Rep. Senegal, Libya Uganda, Zambia 1 3 l3 (Burundi) (CAR), Dahomey Guinea Chad, Congo B. Low Somalia Liberia Gabon, Niger 5400-120 Mali Malawi, Rwanda 4-l.--Sizes of African Armies vs. GNP, 1966 (Sierra L)I(Togo (Upper Volta) Maurifania National Population Sizes and these two variables in the lower half of the figure shows that, out of 35 countries charted, only five states deviate from the pattern. That pattern is visible in both parts of the figure as a concentration of states in a diagonal from upper left to lower right. TWelve states fall outside the diagonal in the upper part of the figure showing the relationship between army and population sizes; five states fall out- side the typical relationship between army size and gross national product. Looking at both parts of the figure, we see only four states consistently fall outside the typical relationship. Of the four deviant cases elicited by Figure 4—1, three have armies larger than would be predicted (Tunisia, Somalia, and Liberia). Of these, Somalia is the most extreme case; it is low in both popu- lation and gross national product, but has one of the largest armies of Africa. One state, Tanzania, shows a smaller army than would be predicted, probably reflecting the low priority given the military function p§£_§g_by the Tanzanian political elite (see Bienen, 1968 and Magrui, 1969). No coup states appear among the deviant cases. Looking at Table 4-2, We discover rather large diversity in the changes in size of armies and in my estimate of the burden of these changes. In the first column, a simple percentage rate of change in size, from 1964 to 1966, has been calculated. Since this statistic seemed too gross by itself to indicate the importance of changing army size, these figures were standardized, in a sense, by multiplying each by the respective army size in 1964. Thus, the second column gives the number of persons added to the army during the two-year period, a different view of change. A large percentage 113 Table 4-2.-—Changes in Size and Burden of African Armies* %Change Number %of Estimated Country in Size, Soldiers Population Increased 1964466 Added, 1966 Burden 1964—66 1964—66** (Algeria) 0 0 .395 O (Burundi) 18.75 150 .029 .435 Cameroon 29.62 800 .065 5.200 (Central African Rep.) 20.00 100 .042 .420 Chad 125.00 500 .027 1.350 Congo Brazzaville 157.14 1100 .212 23.320 (Congo Kinshasa) 6.67 2000 .200 40.020 (Dahomey) 80.00 800 .074 5.920 Ethiopia 16.67 5000 .152 76.015 Gabon 25.00 150 .160 2.400 (Ghana) 112.50 9000 .214 192.600 Guinea 4.16 200 .138 2.760 Ivory Coast 0 0 .102 0 Kenya 91.00 2275 .049 11.147 Liberia -10.61 -380 .293 —11.134 Libya 40.00 2000 .417 83.400 Malagasy Republic 53.84 1400 .064 8.960 Malawi 0 O .021 0 Mali 12.90 400 .075 3.000 Mauritania 100.00 500 .093 4.650 Morocco 28.56 9952 .326 324.435 Niger 0 0 .035 0 (Nigeria) 43.75 3500 .019 6.650 Rwanda 66.67 600 .047 2.820 Senegal 120.00 3000 .154 46.200 (Sierra Leone) -26.48 -49O .056 -2.744 Somali Republic 106.52 4900 .368 180.320 Sudan 68.18 7500 .133 99.750 Tanzania -l0.00 -200 .015 -.300 (Togo) 480.00 1200 .086 10.320 Tunisia -15.00 -3000 .381 -ll4.300 Uganda 198.00 3960 .077 236.412 United Arab Republic 50.00 60000 .597 3582.000 (Upper Volta) 50.00 500 .030 1.500 Zambia 3.45 100 .078 .780 *Calculated from data presented in Table 4—1. Coup countries are in parentheses. **Increased burden is defined as the product of % Change in Size 1964-66 x Size in 1964 x % of Population. 114 increase for a small army may be quite easy to achieve since it involves relatively few men, and it could be easier than a similar percentage increase in size for a much larger army, though other factors would obviously be involved. A 100% increase in size for Mauritania is probably easily accomplished since it means adding only 500 men to the army. A little over 100% increase for Somalia is comparatively more onerous, since it means adding almost 5000 men, ten times as many as Mauritania. Inspection of the first columns of Tables 4—1 and 4-2 suggests that there is little relation between initial army size and rate of change in size. There may be somewhat more association between initial size and number of men added. Somewhat more interesting figures are contained in the last two columns of Table 4-2. The third column shows the percentage of the country's total population which is in the army in 1966, and we see that in all cases far less than one percent are under arms. These percentages show a great range, however, from a high, for the United Arab Republic, of 0.597% to a low, for Tanzania, of 0.015%. This is a ratio of nearly 40 to 1. The last column of Table 4-2 shows an even greater range on a crude estimate of the "burden" of changing the size of the army. These figures are computed by multiplying the states' scores in column two (number of soldiers added) by those in column three (% of Population), in order to show rates of change in army size based not Only on comparison of initial sizes, but also on percentages of the Population involved. Presumably it is more burdensome for a country 115.. with a large number of men under arms to add an additional large number of men to the army, especially if that total comes to a comparatively large percentage of the whole population. This is the case of the United Arab Republic. At the opposite end of the spectrum here is Tunisia, in the sense that it had a fairly large army in 1964 but cut back the army significantly by 1966, although the army still represented a fairly high percentage of the total population. One should also note that some countries not only show no change in army size, but have no important portion of their population included in the military. Niger and Malawi are examples. Tanzania is a more extreme case of their pattern, in that it shows the lowest percentage of population included in its army, plus an actual decline in army size. The great variety in the statistics of changes in size of army and burden suggests the hypothesis that, whatever similarity in outlook the military may share in these countries, they play different roles in different African states. These roles, outlooks, and the influence of the military on civilian politicians are explored in the next three sections of the chapter. Military and Civilian Regimes Since this thesis depends on making distinctions between military and civilian regimes, measurement of the influence of the military in politics must be explored in some detail. At first glance, it seems clear that military regimes are those governments which are headed by men who are professional soldiers and who do not relinquish their military careers when directing the government. 116 But clearly the military may have substantial, even controlling, influence on governments even when civilians head the states. Military officers may staff numerous important positions in govern- ments, as bureaucrats and as decision-makers. The military may be employed in internal development projects or may be necessary for the maintenance of domestic tranquility, so that they must be consulted often and about important matters. Conversely, in regimes headed by the military, the influence of civilians may be high. For example, military governments, especially those which take power in the crisis of a coup d'état, may have to rely on established bureaucracies and their ongoing routines in order to achieve effective government, or they may have to recruit some civilian politicians to their support in order to achieve a crucial link to an important number of citizens. The question is this: How can we measure the extent to which the military have influence on political decision-making? A. R. Luckham has neatly stated for domestic affairs (1971 article), as I attempt to show in this dissertation for foreign affairs, that there is in the literature on civil-military relations an enormous proliferation of ad hoc generalizations on the subject, particularly where the military in new nations is concerned. Many of these contradict each other. Often the theories are not adequate to the facts. And frequently they depend on so many other things being equal that by the time one has made all the necessary qualifi— cations there is not much explanatory force left in the original hypothesis" (Luckham, 1971 article, page 5). Luckham further points out that the main weakness of the existing theories of civil- military relations which take on this task is that they still either concentrate on the characteristics of civilian politics and their influence on military 117 intervention and civil-military relations, to the exclu— sion of the organizational and professional qualities of the military itself; or they give emphasis to the latter, to the exclusion of the social and political environment.* The answer, according to Luckham, is to develop a theory which takes into account both sets of characteristics and the boundaries between them, namely: (a) the strength or weakness of civilian institutions; (b) the strength or weakness of the military, the coercive, political and organizational resources at its disposal; and, the nature of the boundaries between the military establish- ment and its socio—political environment. v (c I shall depend on Luckham's approach to determining the influence of the military, because he integrates the insights of earlier analysts without getting caught in contradictions or con- ceptual culs-de-sac, and because he derives comparative measures for each of his dimensions. To my knowledge, only one other analyst (Claude E. Welch, Jr., 1973 APSA paper) has come to grips with this sort of multi-dimensional theory of civil-military relations, and this later work depends very heavily on Luckham's earlier article. I have brought together the discussions of these two analysts in the next three tables, which show variables in the theory and typologies of roles and regimes which are derived from the theory. Both Welch and Luckham argue that military influence in politics is a function of the relative strengths of the military and of civilian groups and of the clarity and effectiveness of boundaries *"The most important exponent of the former position is S. E. Finer, The Man on Horseback, and of the latter, Janowitz, The Military in the Political Development of New Nations, and Huntington, The Soldier and the State." See Luckham's footnote 20, page 9. 118 between them. As Figure 4-2 indicates, the strength of civilian institutions is a function of public mobilization and support and of those institutions' internal effectiveness. The strength of the military is more a function of size and physical power and its organization, although also of public support and internal regulation. Boundaries are both conceptual and physical, being defined by dif- ferences in goals, legitimation, expertise, responsibilities, command- structures, and routines of action. Implicit in Luckham's argument is the idea that the relative influences of civilians and military men cannot be known from information on who heads a state. Whether a head of state is a military officer or a civilian may indicate the influence of the military in politics, but not necessarily so. For example, even in a situation of determinative military influence on decision-making, we may find a civilian "puppet" as head of state, because the civilian helps legitimate the government, or he is a useful front-man for the military, or he has been installed according to some pattern which does not challenge the military. In any case, Luckham does not care to dichotomize between civilian and military regimes, believing those categories to be too gross. Rather, he derives a nine-fold typology based on the interplay of civilian power, military power, and boundary type. Luckham's typology, which is reproduced in Figure 4-3, does carry a hint of the judgment that civilian control of the military is the normal arrangement. That juflgment seems to be based on two observations: First, by definition, to the extent that there is a distinct military, that group is 119 I. Strength of Civilian Subsystem of National Social System A. Public Support 1. Amount and scope of allocative activity 2. Degree of political mobilization (e.g., voting—turnout) 3. Ascribed legitimacy B. Effectiveness of Political Structures 1. Internal coherence 2. Internal self—regulation and discipline II. Strength of Military Subsystem A. Coercive and Strategic 1. Size (absolute and relative to population) 2. Share of budget and percent of GNP 3. Firepower utilizeable 4. Organization (see below) B. Organizational 1. Number of officers (absolute & as % of total force) 2. Quality of officers (experience, education, personality) C. Political 1. Levels of coercive and organizational strength (above) 2. Convertibility of above to political resources 3. Integration with civilian decision—making (e.g., parti- cipation in cabinets, friendships, kinships, ethnicity, etc.) 4. Ascribed legitimacy III. Boundaries Between Military and Civilian Subsystems A. Integral l. Narrowly defined focus; expertise in the management of violence, especially for external defense 2. Corporate autonomy and institutional coherence 3. Command structure is primary channel through which contacts are made with other institutions B. Fragmented 1. Diffuse focus: mixing of internal security and international-defense responsibilities 2. Direct contacts between members of the armed forces and civilians, by-passing the command structure 3. Conflicting loyalties and legitimations, based on ascriptive and institutional bases C. Permeated Such complete fusion of goals and of organization between the armed forces and civilian groups that they are substantially indistinguishable Figure 4-2.--Variables in 3-Dimensional Theory of Civil—Military Relations 120 .NN ovum .Hhma .uoucflz .COfluflmmmmo can . . n onso mmmmmmmwmm =.m:0HumHom mumuHHHZIHfl>HU mo hmoHomhfi o>HumummEOU d: .Emnxosq m m m we amerHHom GM onom hnmufiaflz wo mmoaommalt.MIv .oem szson> Hmoepnaom .m sou oumum cmwoumso Honcoaoouumom Ans .o been uoz oumum oumum Esfipmz canuouomnm .w :onoumsw any .h Houusou oumum somflnumw .o smflm o>euooflnnm .m BQH mfindusHICOnumz mums0flpsao>om Anv.v mEHducHsoowumz Monacoo anaconusao>ou humasnmumcou N Santos smflm nsoz Amv.v Houucou Houucoo swam pmnwmmm .m m>wuownno .H owumoshom pmucmsomum enroaucH ”030m Hosom U0 «LTTCEOM WHUUHHHE HH>HU 121 supposed to be apolitical; the military are not defined as a delibera- tive or policymaking body, but as skilled servants of other decision-makers. Second, civilian supremacy is typical. Welch, on the other hand, believes one can still usefully distinguish between civilian and military regimes, although he also sees subtleties of interplay between these two groups. Welch‘s 36- fold typology, reproduced in Figure 4—4, thus is based on a dis— tinction between civilian and military regimes as well as on the relative strengths of these two groups and the nature of the boundaries between them. If Welch is correct in holding that it makes a difference to know whether a civilian or officer heads a government, then his extended typology is superior to Luckham's. If this dis- tinction has no predictive value, it is superfluous at best and con- fusing at worst. By using the term "praetorian" to designate all types of military regimes, Welch has blurred the meaning of that con- cept as originally defined by Huntington (1968) and Rapoport (1963) and as used in Luckham's typology. Also, there is a paradox in Welch's typology because of his treatment of the distinction betWeen civilian and praetorian polities as as one dimension. 0n the one hand, Welch's purpose in drawing up his typology apparently is to define and, by the ingredients of that definition, to account for different military roles. On the other hand, however, the distinction between civic and praetorian polities amounts to using a judgement about military roles in the definition of those roles, so the argu- ment becomes circular. 122 .Amnma meme .no «.uosamom ono moeom Hmoeunsoo snmuneez mo smonomsa,-.aua .oem .mmoum husnxso “mumsuaom cuuozv masm can oaom umuwaflz .QUHEm .M usnuud pom ..Hb .SOHUB .m mosmao "monooma Houom Houom Houusou Honusoo ammofiuflamoo HMOHDHHom ucowswm mafiawbflo EMHHH>HU mm humpflaflz loosH mm wnmuaaflz m>auoomnsm o>fluomflno OH Mantras b «in» omcoo Hpfimm Hmmmz MHQEMN 30g nnuomnz oeflnamom Snowmen: namoarnm umooo suo>H sanooz sou ocoenons uoon>anm Hm conH anmocom roam Amoma .ov Anmma .Uv - .mom unoncnsoo mauum mousse nonm numoo 30a mango sumo ooexmz oneno summon annooz pesos Haunum undo neonH roam Romma .Uv Aomma .ov ashamed Momsuwu ocmasfim coach son msflucomum Ammmalvmmav moxuss HmmumH cowo3m Edwooz Awflm Aswansmom :he comes museum oomonw cons: uwn>om mouwpm nouns: son: popcommmum HmumouoH counomwmum HmumoucH coaucm monumwcdom mmflumonsom sumcwuum Ifloauumm Hmonunaom Hmowunaom season annuouonum sueeom on>no m.snoueee2 mo usouxm 123 There is another weakness in the work of Welch and Luckham. To a certain degree, both seem to argue the need to keep civilians and the military in their respective places. It is assumed that each must have a clear and separate identity. More, it is assumed that each must have strength to defend itself against the other, since they will be antagonistic. That is, a balance of power between military and civilian groups must be achieved in order for government to be stable. While I agree that there will inevitably be some tension between civilians and the military--at least to the extent that they are distinct groups-—there must be more than a relationship of reciprocal deterrence between them for government to operate. They must remain distinct if they are to achieve the special expertises and flexibility in government that are the advantages of a division of labor, and the boundaries between them must be clear and effective if the military are not to become over-involved in politics. But those boundaries must be more than hostility-lines; they must define a mutually constructive or symbiotic relationship. The distinction between the groups is not meant to isolate or protect one from the other but to define complementary roles. Even if the boundaries are clear and tight, they may not restrain the military from entering politics if the military see themselves as misused or dishonored or cheapened by the distinction between their roles and those of civilians. Worse, if boundaries isolate the groups or define dif- ferences in goals and legitimacy and ways of interacting in terms that are mutually exclusive and antagonistic, the grOups must conflict. 124 Knowing all this still does not allow a clear categorization of African countries into civil-military regime—types. Welch's typology classifies eight African states, but if the distinction between civic and praetorian polities is removed—-I have suggested above some problems arising from the inclusion of this dimension-- then those countries are grouped more closely together. Luckham suggests that most African states will be grouped in a couple of regime types (Post—Colonial Guardian State and Praetorian State), so his typology is not much help in differentiating politics in these states. In addition, very few of the variables in Luckham's theory can be systematically compared across the African states, largely for lack of data. There are few data available on the strength of civilian institutions, and even those data may be mis- leading. For example, Aristide Zolberg has argued (1966) that political mobilization and regime-support cannot be inferred from data on voting, since even overwhelming electoral victory may be based on a tiny and atypical percentage of voters. The strength of civilian institutions is thus often judged indirectly, by whether or not there is a successful coup d'état. But we have already argued, with Luckham and Welch, that the military may be highly influential even in so-called civilian governments. The estimate of the nature of military boundaries usually Suffers from the same indirect measurement. The strength of the military is somewhat easier to determine, since there are published figures on military sizes, expenditures, and firepower in most African states. Some of these data are 125 presented in Table 4-3. In addition, I have surveyed as thoroughly as possible the cabinet personnel of African states in 1964 and 1965, looking particularly for the presence of military officers. This search did not prove very helpful. Only one state (United Arab Republic) showed more than one military officer in its cabinet, and that state can be eliminated from my sample for other reasons. All the states shOWed very high rates of turnover in cabinet personnel, often a majority each year, so that whatever influence any one cabinet officer might have was likely to be short—lived. Also, I was some- times unsure whether military titles were purposely not being included with the names of cabinet personnel (I was concerned about this pos- sibility because I simply could not find any_military officers in cabinets outside the UAR). Finally, biographical information is not available on a very high percentage of African elites below the level of head of state, so information on military affiliation or experience or education or other possibly important factors could not be accumu- lated for the large number of countries being compared here. There is some evidence, however, that the military do not enjoy a very good image in Africa (Grundy, 1968). Table 4-3 presents data for a crude quantitative estimate of military influence in African governments in 1964. I have ranked 35 c0untries on three coercive and strategic variables suggested by Luckham, and then summed the ranks for each of the countries to obtain an overall measure of military strength. There is a sub- stantial range of Scores on each variable, especially size of army and size of army as percentage of population (the largest scores are 126 Table 4-3.—-Partial Indicators of Extent of Military Influence in African Governments, 1964 countries Ranked on Variables, with Tie Scores, Bracketed. Mi itary Size of Army Rank Expenditures Size of Army as a of Sum of Ranks as % of GNP Population 1 7.4 UAR 120,000 UAR .438 Tunisia 5 UAR 2 4.8 Morocco 48,000 Algeria .437 Algeria ll Algeria 3 4 . 3 Cameroon 34 , 848 Morocco . 415 UAR ll Morocco 4 4.0‘ Mauritania 30,000 Congo K. .343 Liberia 24.5 Somalia 5 4 . 0 Somalia 30 , 000 Ethiopia . 320 Libya 28 Ethiopia 6 3.3 Congo Brazza 20,000 Tunisia .268 Morocco 28.5 Libya 7 3.2 Algeria 11,000 Sudan .196 Congo K. 31 Guinea 8 3 . 1 Mali 8 , 000 Ghana . 195 Somalia 32 Ghana 9 2 . 5 Dahomey 8 , 000 Nigeria . 140 Guinea 34 Sudan 10 2.3 Ghana 5,000 Libya .135 Ethiopia 38 Congo K. 11 2.3 Guinea 4,800 Guinea .130 Gabon 38.5 Tunisia 12 2 . 3 Sudan 4 , 600 Somalia . 106 Ghana 41 Cameroon 13 2.2 Ethiopia 4,000 Ivory Coast .106 Ivory Coast 42 Mali 14 2.2 Libya 3,580 Liberia .084 Congo Brazza 44.5 Liberia 15 2.1 Gabon 3,100 Mali .084 Sierra Leone 48.5 Ivory C. 16 2.1 Malawi 2,900 Zambia .083 Sudan 50.5 Congo 8. 17 2 . 1 Togo 2 , 700 Cameroon . 080 Zambia 55 Zambia 18 2.0 Niger 2,600 Malagasy R. .073 Senegal 56.5 Dahomey 19 1 . 7 CAR 2 , 500 Kenya . 069 Mali 57 Mauritania 20 1.7 Malagasy Rep. 2,500 Senegal .055 Mauritania 58 Gabon 2]. 1.6 Senegal 2,000 Tanzania .052 Cameroon 58.5 Senegal 22 1.5 Zambia ‘2,000 Uganda .043 Dahomey 60.5 Malagasy R. 23 1.4 Ivory Coast 1,850 Sierra Leone .042 Malagasy R. 67 Niger 24 Chad 1 , 200 Niger . 037 CAR 69 Nigeria 25 1.2 Congo Kinshasa 1,000 Dahomey .036 Niger 69 Sierra L. 26 1.2 Liberia 1,000 Upper Volta .032 Burundi 76 CAR 27 1 . 2 Nigeria 900 Rwanda . 029 Rwanda 76 Malawi 28 1.2 Uganda 800 Burundi .027 Kenya 76.5 Uganda 29 Upper Volta 800 Malawi .027 Uganda 78 Kenya 30 1.0 Kenya 5700 Congo Brazza .021 Upper Volta 82 Upper V. 31 0.8 Sierra. Leone 600 Gabon .020 Malawi 84 Togo 32 0.8 Tunisia 500 CAR .020 Tanzania 87.5 Rwanda 33 0 . 7 Burundi 500 Mauritania . 015 Togo 88 Burundi 34 0.7 Rwanda 400 Chad .014 Nigeria 88 Tanzania 35 0.4 Tanzania 250 Togo .012 Chad 95.5 Chad 127 greater than the smallest by factors of 480 and 36, respectively). The ratio of highest to smallest scores for military expenditures as a percentage of GNP is 15.5, and the ratio for the sum of ranks is 19. If the UAR is removed from these distributions, the range is greatly reduced for all variables except size of army as a percentage of population. If all the North African countries are removed, the ranges on all variables are greatly reduced, and the ratio of highest to lowest scores on the sum of ranks is only 3.9 to one. One country stands out on Table 4-3 as heavily influenced by the military; it is the United Arab Republic. This is also the only African state that showed a concentration of military officers in its cabinet in 1964. Because the UAR was so heavily influenced by the military at the beginning point of this study, and because the data on this state show it to be generally atypical of Africa, it will not be included in my tests for differences between civilian and military foreign policies. Like the UAR, the other states of North Africa generally show a tendency to be heavily influenced by their military, even though these states Were not headed by military men in 1964. Algeria had recently emerged from a real war for independence, it is true, and its head of state in 1964, Ahmed Ben Bella, was a leader in that struggle; but Ben Bella had been much more remote from the military situation than other leaders (because of his incarceration in France), and he had disengaged from the military and moved toward creating rule by the party (FLN). Morocco, Tunisia, and Libya were certainly not military regimes in 1964. Overall, though, North Africa seems 128 somewhat atypical of African states in the rather heavy presence of the military at the beginning point of this study. Presumably, this heavy presence would mean that over the four—year period under study these countries would show a less pronounced change in foreign relations toward special military concerns than would other states which had a much smaller military presence in 1964. Because of the apparent importance of the military in North Africa, then, as well as for the other reasons discussed in Chapter II, my tests for dif- ferences between civilian and military foreign policies will be conducted both with the North African states included in the sample and excluded. A few other states in my sample demand special attention when the tests are conducted. These are states which maintained what appeared to be continuous civilian ruleithrough the four-year period under study, but started that period with such a high degree of military presence that their regimes may not properly be called civilian. In addition to North Africa, these states include Somalia, Ethiopia, Guinea, and the Sudan. If these states show a common pattern of foreign policy which is different from that of other "civilian" states and closer to that of the military regimes which appear after coups d'état, then the hypothesis of a special military influence on foreign affairs may tend to be strengthened, but the inclusion of these states in the "civilian" group may confuse tests for difference between civilian and military groups. That is, civilian regimes permeated heavily by military institutions may not be so clearly civilian as other regimes, but may require some third 129 classification. For the time, however, I will consider them "civilian“ until the foreign-policy data may suggest otherwise. The Military Outlook Eric Nordlinger (APSR, Dec. 1970) has suggested that the military's influence on politics comes out of two sources: the military's corporate interests and the military's values. The first sees the military as a "trade Union," the second as a set of per— ceptions and judgments. They are two sides Of the same coin, rein— forcing each other. For Nordlinger, either perspective on the military leads one to the conclusion that they will be conservative and self-interested, hindering "progressive" economic and social change. As a trade union, "the military act to maintain or increase their wealth and prerogatives even when these values conflict with the aspirations and interests of larger segments within the society" (p. 1134). Out of their near-universal military values-—the normative attachments to order, dignity and hier- archy—-with which most officers are strongly imbued, emerges an overwhelming concern for political sta- bility, and thus a keen sensitivity to any divergence from the status quo that centains the potential for unwieldy change. But Nordlinger acknowledges that counter-images of the military outlook exist. Lucian Pye has argued (1962) that the modern organi- zation, skills, and weaponry of the military make officers extremely sensitive to the needs of modernization and technological advancement. This kind of sensitivity bears little relationship to the command of physical Violence and tests of human endurance . . . . In con- sequence the officers often find that they are spiritually in tune with the intellectuals, students, 130 and those other elements in society most anxious to become a part of the modern world. Military men not only see this need for change; they are also "forced to look outside their society for their models" (page 77). By inference, it is not only a parochial, protective outlook, then, but also a cosmopolitan and developmental outlook. According to this view of the military, this sensitivity and desire for modernization is buttressed by the technical and managerial competence which is the military's training. John J. Johnson argues that this technical'managerial orientation probably makes it easier for the military to accept the shift in political power from the land-holding elite to the new "urban alliances" (1964). Horowitz agrees, seeing the elite of the armed forces in the new states as “no longer aligned with the traditional upper classes" (1966: 261). In fact, Nordlinger himself sees the argument that the military con- stitute a "meritocracy rather than an established class" (1133). Some analysts believe that the army provides a recruitment and socialization function which not only instills skills and devel- opmental orientation but also liberates men from their traditional social structures. There are two aspects to this notion. First, Edward Shils (1962) has noted that, in underdeveloped areas, the military tend to recruit "the brightest and most ambitious young men of the small towns and countryside . . . from the families of petty traders, small craftsmen, and cultivators of small holdings, (who), like their fathers, are aware of the distance separating them from the rich and the political elite" (page 17). The military career, 131 then, may create a counter~elite, "resentful against the established order and isolated from its leading spokesmen“ (page 18). Second, Belmont Brice (1966) argues that the military has a better chance of combatting tribalism than any other element of African society. The reason is that the duties of military service tend to divorce the man from other elements of society more than do other occupations: he is physically separated from family and friends for long periods; he is subjected, indeed conditioned, to a new discipline that is sup- posed to override all other influences, at least during the period of duty (page 59). Irving Louis Horowitz agrees, arguing that "when functioning properly"--that is, presumably, when recruiting and training in terms of technical and managerial skills and national service-—the military "quickly acquires a sense of the nation and becomes hostile toward vested interests and sectional enclaves . . . (and it) mini~ mizes the class base of society through a heterogeneous recruitment policy" (1966: 259). Horowitz goes on to argue that "the capacity of the armed forces to act as a nation—building instrument is inherent in its structure. It is often the most 'modernized' and most highly refined organization in Third World nations. This does not simply mean that it is technologically proficient." It makes for social reorganization, it has a mystique surrounding its power and efficiency, it may provide some continuity with the past in that at one and the same time it is oriented to modernist goals and still committed to controlled and stable change. In societies where everyone is tardy, the military is prompt. Where the population is ragged, the soldiers are neatly uniformed. Where indecisiveness 132 reigns supreme, the military can take direct action (Horowitz, 1966: 267). Nonetheless, the literature predominantly presents an image of regression from modernization under under military rule, generally for three reasons: (1) lack of political skills, (2) rigidity of mind, and (3) conservative or reactionary values. All the authors cited above agree on at least the first judgment: "The military, powerful enough to cancel democratic norms, is not powerful enough to maintain social order over a long period of time. By temperament and by training, the military is more capable of preventing the exercise of political rule than of exercising such rule itself" (Horowitz, 1966: 268). Glickman (1966) summarizes: The strengths of the military also account for its basic weaknesses. The use of weapons internally erodes the confidence of the masses. Discipline makes for order, but it is not helpful in eliciting a response from the people. Military efficiency can accomplish technical tasks, but it is not related to the calculation of social choices and the balancing of social interests. If the military endures as a government, it gradually must shift from management to bargaining, a pattErn that subjects officers to political pressures with which they are ill equipped to deal as military men (pages 72—73). Typical judgments of the military, probably with more reference to the "developed" countries than to Africa, pass even harsher judg— ment. Snyder and Furniss reduce "the most serious criticisms of the military mind" to: (1) rigidity in thought and problem analysis—-the rejection of new ideas and reliance on tradition rather than on lessons learned from recent experience; inadequate weighing of non-military factors in military problems and inability to understand complex police—military relationships; (2 v 133 (3) an authoritarian approach to most social issues and situations, accompanied by a disrespect and disregard for civilian authority; insulation from non-military knowledge or anything beyond what is narrowly defined as militarily relevant; (5) judgment of policy-goals and techniques primarily in terms of military force and total victory from total war (1954: 369). (4 V The problem remains how to translate the "military min " and military values into policy—preferences. This translation should take into consideration the conditions under which the military operate, or the roles the military play in any state. The various roles in African states are disaussed in the following section. Here, a partial translation of values into policy, taken from Huntington (1957: Chapter Three), sums up military values as an "ethic of conservative realism." Huntington sees this ethic as "concrete, permanent, and universal“ among military professionals (page 89): The military ethic emphasizes the permanence, irra— tionality, weakness, and evil in human nature. It stresses the supremacy of society over the individual and the importance of order, hierarchy, and division of function. It stresses the continuity and value of history. It accepts the nation state as the highest form of political organization and recognizes the con- tinuing likelihood of wars among nation states. It emphasizes the importance of power in international relations and warns of the dangers to state security. It holds that the security of the state depends upon the creation and maintenance of strong military forces. It urges the limitation of state action to the direct interests of the state, the restriction of extensive commitments, and the undesirability of bellicose or adventurous policies. It holds that war is the instrument of politics, that the military are the servants of the statesmen, and that civilian control is essential to military professionalism. It exalts obedience as the highest virtue of military men. The military ethic is thus pessimistic, collectivist, historically inclined, power-oriented, nationalistic, militaristic, pacifist, and instrumentalist in its View of the military profession (1957: 79). 134 Huntington‘s portrait hangs on his notion of "professionalism," as Finer points out (1962: 24), and this notion really begs the question of why the military sometimes intervene in politics. Finer argues, in fact, that "professionalism" sometimes forces the military into conflict with civilian leaders, rather than preventing it (pages 24-30). Aside from this disagreement, however, the Huntington summary of the military ethic provides a way of pulling together the various views of military values, in that it shows the pivotal importance of the notion of the nation-state. The military might be both conservative-—in its concern for power, its distrust of novelty, its concern for craftsmanship and competence but not for grand design ‘or ideology, its incrementalism and concern for control, etc.—-and open to, even desirous of, modernization and change, because in the new states modernization means the development and management of power, taking one's "rightful" and "dignified" place in the world of nations, achieving control over change, executing policies with pre- cision and competence, and so on. On the other hand, there might not be this link between conservatism and orientation to modernization. This is an empirical question to which my research is meant to speak in the area of foreign affairs. For the very newest states-—including the vast majority of Africa--there are no past nation-states and few national power- structures for the military to protect. Unlike Latin American states, the military in Africa are lost if not identified with the new "nations-in-becoming." As we have already seen, the African countries show, in the aggregate, the lowest levels of modernization, 135 socially, educationally, and politically, and so we infer that the military, in order to realize their own values and gain their own interests, are forced to push for modern development. In this view, the military are hypothesized to be implementers of change. We arrive at this inference for two reasons: First, the military see and understand the world through the lens of profes- sional, western-style military training, and, second, what they see are new states and processes greatly at variance with the standards their training has provided them. "As products of a relatively advanced technical apparatus in a peasant society, officers are natural competitors to politicians who preach progress and problem-~ solving on the basis of advancing technology" (Blickman, 1966: 72). William Gutteridge makes the argument, in several places, that African armies are "essentially products of the colonial period" (1970: 18), and "that imperial policies with regard to recruitment and Africanization have determined to a considerable extent their composition and that this has been a factor influencing status and attitudes" (1969: 7). Thus, it is possible that the African military outlook is as oriented to the goal of development as that of civilian regimes, though their styles and practices will undoubtedly differ as they try to reach the goal. Because development is dependent symbolically and materially on foreign relations, such differences in goal-orientation and style should be evident in the structure of external behavior. This is the inference my data will confirm or deny. 136 Military Roles in New States I have already argued in Section 2 that military roles are determined by the interaction of military and civilian strengths and the boundaries between these groups. In Africa, we predict certain roles. "New nations, even more than old, depend for their existence on the integrity of their armed forces" (John Slessor forward to Gutteridge book, 1962: v), because the military "is a heavy insti- tution" in these lands (Greene, 1966: 10). By definition, though, African military establishments create severe problems by their dependency abroad and their low integration at home. It would seem that the military influence would be felt most greatly in those countries with the largest and most fully-trained armed forces and in those with the greatest socio-economic potential for enhancing military values, that is in countries with a capacity and orientation to material and technical development in a stable manner. A number of generalizations may be made about the role of armed forces in Africa. First, armies have not been important to the recent independence struggle, except in Algeria. We may infer that armies are not seen by either civilian politicians or the masses as crucial to the development and definition of the new states. Those armies which exist after independence are, to a large extent, hangover from colonial organizations. Second, as remnants of the imperial past, armies are often feared and mistrusted, according to Kenneth Grundy (1968?). The reasons for this attitude are not hard to find . . . . Their chief assignment during the colonial period was to pacify the hinterlands, and once that was accom— plished (and certainly not without popular animosity 137 toward the military), they were called upon to suppress disturbances and internal uprisings and to protect European property . . . . In this respect, the armies were looked upon as armies of occupation, betraying the nationalist struggle, and oppressing the masses rather than defending their interests (p. 429). Third, there is apparently little need for border defense in most African countries. At least, in few cases have there actually been external threats to national security. The most obvious excep- tion is Egypt, involved in three wars with Israel and an unending arms race, but Egypt has been removed from my sample for other reasons. To a more limited extent, Somalia, Ethiopia, and Kenya have had a series of border disputes, mostly resulting from the attempt to repatriate Somalia from the other two countries. This dispute was largely resolved, on an official level, in mid-1968 (cf., Africa Report, February, 1969). To my knowledge, however, there has been no major external threat to governments in African countries. Fourth, in most African countries, armies are so small they could not handle real security problems anyway. Probably size of army is related to the lack of border threats, though which is cause and which is effect is unclear. It may be that armies are small because neighboring states are no threat, or it may be that neighbors can make little threat because their armies are small. In either case, it seems that in most countries there are no external security risks sufficient to be rallying points for building large defense forces, especially since such forces would be terribly costly in already weak economies. The result is that the army tends to look inward to domestic politics. 138 Fifth, the army becomes involved in domestic politics in two ways. On the one hand, as indicated above, the military have little external function to serve and so may dwell on domestic problems. Their outlook may well bring them into conflict with civilian leaders, conflicts which might be suppressed or overlooked if both faced an external enemy. On the other hand, the armed forces may appear to civilian leaders as readily available and competent instruments for domestic development; armies may be put to work building roads and hospitals, and conscription may be a convenient device for recruiting and socializing the masses to a national outlook and to national job— priorities. Israel is seen as a successful example of this develop- ment and use of armed forces, according to Gutteridge (1965: 441-3). These generalizations do not add up to the successful inte— gration of armed forces into the life of new African states. Though the armed forces are small, even the smallest have shown the capability of intervening in politics (Togo). Though there are tasks of domestic development for which they may be used, they are not armies which were developed with this orientation in mind; they are mostly "colonial armies," to use Pierre van den Berghe's term (1965), oriented to a function of border defense which largely does not exist. They are Often proud of their professionalism and concerned with their security job, which they see depreciated by participation in labor Projects at home. Finally, their Weapons, equipment, training, and organizational structure all reflect colonial origins and, hence, continued costly dependency abroad. These costs are especially difficult for African states to bear, since the most expensive 139 items—-foreign weapons, equipment, and training--feed very little into domestic development. The natural interests of these armies, then, clash with the priorities of development if not the goal of development. Hence, for the African states, it may be that the duality of images of the military mentioned by Lucian Pye (1962) is not a prob— lem. In Africa, there is no simple contrast between corrupt and enlightened images; the distinction between the incompetent authori— tarian, on the one hand, and the dynamic, self-sacrificing officer, on the other, is not really uSeful. The conservative attitude, the values of order and obedience, the nationalist view, the importance of the state, and so on, can all be brought together in an image of the African military which is still oriented to development and change. The problem is not the multiplicity of images held about the military, but the conflict and dilemmas within the military outlook. The military may both serve national integration and be very costly to national development. It is this conflict and dilemma within the military which leads to rival inferences about the role of the mili- tary in Africa. On the one hand, the professional soldiers may be implementers of modernization and development, but, on the other hand, they may be impediments. It is an empirical question. The conditions of the African states and the roles of the military in them compound the dilemmas for the military, even if we assume that the military are modern and developmental in orientation. These states have relatively little need, thus far, for modern armies to fight external enemies, though they may enjoy the prestige of such 7. -x- x, “4;." A...~,., —l- 140 armies. The military preference for development may be laudable, but also disjointed from problems of social mobilization and political management. In sum, the military outlook may not be corrupt, but it still may be at odds with other views of development. The military's developmental strategies will probably be materialist, technical, controlled, and cautious, whereas civilian politicians may be more sympathetic to symbolic, ideological, mobilizing, and bold strategies. Hypotheses on Military Influence on Foreign Relations Extent of Change After Coups d‘Etat My argument is that the military in Africa need not be seen as corrupt even though their priorities and style of orientation to development will be different from those of civilian elites. The question is whether the military will act on balance as implementers of or impediments to modernization and development, and how, particularly, their orientations will influence foreign relations. When the military take over governments through coups d‘état, there are three possible results for foreign relations: First, foreign relations may continue to develop along lines established by civilian governments; second, foreign relations may be restructured to reflect a special military outlook similar in all or most military govern— ments; and third, foreign relations may be restructured in idiosyn— cratic ways without much similarity among military regimes. The extent of restructuring of foreign relations, whether it reflects general military or idiosyncratic designs, will be dependent on how close the previous regime came to the military priorities 141 and style. It is assumed that civilian regimes are more likely to approximate the styles and priorities of the military the more that there is a heavy presence of armed forces. As in Table 4—3, we will measure military influence on civilian regimes in two ways: first, in terms of the relative burden of military expenditures which the civilian elite allows or supports; and, second, in terms of the actual size of the armed forces. It is assumed that a heavy level of military expenditures—-especially in poor African states with relatively few security problems--reflects unusually effective mili— tary lobbying, bargaining, or threat which the civilian government cannot overlook or may not wish to overlook. It may be that the civilian elite values military advice highly, or it may be that the civilians are maneuvered or forced into taking it, by threat of violence, by lack of equally well organized countervailing advice, by persuasion, or by the offer of rewards. In African countries where civilian elites hold power in a fragile manner, it may not take a very large military force to create a very effective military influence. Some African countries do have relatively large armed forces, hOWever, which must magnify the potential for influence which the military hold. The corporate interests of the military take on a variety and comprehensiveness of their own as the corporate body becomes large. Thus, even if they represent a smaller percentage of the total population than in a smaller country, when the armed forces reach a size in the tens or hundreds of thousands they may become a more complex and demanding interest group of a new order. As the 142 armed forces become large, they become institutionalized, with greater role-differentiation and organization, second—order requirements for supplies, and greater probability that a significant number of men will "eat and sleep" things military. Their interests may become articulated and aggregated in degrees that small armies cannot often achieve, and they are more unwieldy to control. Whether or not civilian regimes value their advice, the armed forces may reach a critical mass which cannot be overlooked. These two measures of military influence are presented in the figures on the following pages for the period before the rash of coups d'état in Africa. By establishing the extent to which mili— tary influence was felt before the coups, we may predict the extent to which foreign relations will be altered after coups. The measure of relative burden of military expenditures is given in Figure 4—5 by cross-tabulating military expenditures as per- centages of GNP against GNP per capita. This method of measuring "burden" is taken from the work of the United States Arms Control and Disarmament Agency. Their explanation is this: The most commonly used indicator for these purposes, the ratio of military spending to GNP, is inade- quate if used alone, and may be misleading. A major weakness of this ratio is that it fails to take account of the population factor and therefore of the level of economic strength as represented by per capita income. To offset this weakness, per capita GNP in each country is shown next to the ratio figure. Countries with low per capita incomes are more likely to have greater and more urgent resource scarcities. In these countries military spending competes directly with both consumption and the investment necessary for growth (World Military Expenditures, 1969: 5). Military Expenditures as % of GNP (1964—65 Average) HIGH BURDEN 5.0% and above to 4.0% 2.1 3.0% 1.1 2.0% 1.0% under GNP/Capita (1965) Under 70 $70-140 $140-210 3210—280 $280—350 $350 & up United Arab Republic Somali Republic Morocco Mali Guinea Congo Braz (Algeria) Cameroon Congo Kinsh Mauritania (Dahomey) Ethiopia Niger Libya Sudan Malagasy R. Senegal (Upper (CAR) Tunisia Volta) (Nigeria) Liberia Zambia (Ghana) Gabon Malawi chad Ivory Coast Uganda l'f‘ngg) (Burundi) Tanzania Rwanda Ken a (Sierra L.) y Low BURDEN Figure 4—5.--Relation Burden of Military Expenditures (Coup Countries in Parentheses). 144 Figure 4-5 summarizes the relative standing of all the countries in this study, taking account of all three relevant factors: military expenditures, GNP, and population. The figure shows where each of the countries stands in per capita GNP and in the ratio of military spending to GNP. Countries with the heaviest defense burden in terms of these criteria appear in the upper left portion of the figure; countries with the lightest burden, in the lower right portion. Two heavy diagonal lines have been drawn to separate the 35 states into three groups having respectively (1) a high burden, (2) a medium burden, and (3) a low burden of military expenditures. The ten "coup" countries are indicated by parentheses. Only one "coup" country, Congo Kinshasa, appears in the "high burden" category, while two "coup" countries appear in the "low burden" group, Sierra Leone and Ghana. The other seven “coup“ countries are distributed very similarly to the continent as a whole. Figure 4-6 cross-tabulates the three-way ranking of states by burden of military expenditures from Figure 4-5 with a three-way ranking of states by actual size of armed forces. Using the 1964 data from Table 4-1, the countries were ranked by army sizes, and all sizes above 10,000 were designated "high," those between 2000 and 10,000 were designated "medium," and those 2000 and below designated "low." For each of the two dimensions, the "high" category was given a weight 0f 3, the “medium" a weight of 2, and the "low" a weight of 1. Each of the nine cells of the figure has a combined weight which is the multiplication of its row weights by its column weights. The possible Military Burden Supported by Regime 1964—65 Hid HIGH (weighting = 3) 145 Army sizes 1964 MEDIUM (weighting = 2) LOW (weighting = 1) Influence High (3) Medium (2) Low (1) (combined weighting=9) (Congo Kinshasa) Ethiopia Morocco United Arab Rep. (combined weighting=6) Cameroon Somalia (combined weighting=3) (combined weighting=6) (Algeria) Sudan (combined weighting=4) Kenya Liberia Malagasy Republic (Nigeria) (combined weightinq=2) (Burundi) (Central African Congo Brazzaville (Dahomey) Tanzania Malawi (Togo) Mauritania Uganda Niger (Upper Rwanda Volta) (combined weighting=3) Tunisia (combined weighting=2) (Ghana) Ivory Coast Libya Senegal Zambia (combined weighting=l) Gabon (Sierra Leone) Low Influence Figure 4-6.——Estimates of Military Influence in Pre-Coup Period (Coup Countries in Parentheses) 146 combined weights are 9, 6, 4, 3, 2, and l: the higher weights indicate higher estimated military influence in the pre-coup period of 1964—65. Figure 4-6 appears to show a slight curvilinear relationship between level of military influence in the pre—coup period and sub- sequent coup d'état. Though the "coup" countries tend to concentrate in the lower-weighted cells of the figure, as-a percentage of the countries in each cell these “coup" states seem slightly more likely to appear at both extremes, the highest and lowest weights. Table 4—4 presents these data. Table 4-4.-—Distribution of Estimates of Military Influence for 35 States in Pre-Coup Period Weights Number of Number Percentage Coup States All States Coup States 9 l 4 25% 6 1 6 17% 4 l 4 25% 3 0 l 0% 2 6 18 33% 1 l 2 50% Presumably, those countries with the largest military influ- ence before coups d'état will need to reorient policies and practices less than those having least military influence. Thus, while all coup countries are hypothesized to show reorientations in the structure of foreign relations after coups, Congo Kinshasa should show less than 147 Sierra Leone, and all the other "coup" countries should show degrees or reorientation somewhere between these extremes. The extent of reorientation in foreign relations after coups d'état depends on something more than the baseline of military influence before coups, however. It depends first on the extent to which the military group which takes over fits the professional outlook discussed above and on the following pages. It also depends on the extent to which the country or governmental apparatus which is taken over has effective decision-latitude. As William Foltz argues in the course of his review of Edward Luttwak's Coup D'Etat: A Practical Handbook, In Africa, even more than elsewhere, the coup is likely to have a nonrevolutionary bias built into it, irre- spective of the intent of its planners. The coup aims to take over the governmental apparatus, and the most successful possible coup can capture no more than what the government possesses. In Africa that may not be very much. Numerous regimes, in effect, centrol little more than the capital city, some major provincial towns and communications routes and some uncertain part of the commercial exchange economy. Certainly there is no reason for a military coup, or the regime issuing from it, to be more effective in mobilizing the masses than was its predecessor (1969: 34-35). Foltz may be overly pessimistic. It is not inconceivable to me that a coup may be greeted with enthusiasm by the politically attentive elements among the masses, and by that enthusiasm be enabled to reorient policy and practice. 0n the other hand, Foltz would seem to have a useful argument to the extent that countries Poorly developed in modern materials and infrastructure may have little capability for reorientation. In other words, they may have little decision—latitude. 148 We might estimate this decision-latitude in two ways: first, by the actual level of social and economic development, and second, by the level of economic gigg. Development speaks of capacity for change; size speaks of resources for change. The figures on the following two pages rank the "coup countries first on urbanism, an indicator of development (Figure 4—7) and second on GNP, an indicator of economic size (Figure 4—8), and plot these indicators against estimated levels of pre-coup military influence (Figure 4—6). Interpreting both graphs, it is assumed that a combination of large size, or high development, and low pre-coup military influence will make for great shifts in the post—coup orientation of foreign rela- tions. Thus, the closer a country appears to the upper right hand corner of the graph, the greater the extent of change in foreign relations expected, and, conversely, the closer it appears to the axis of the graph in the lower left hand corner, the less the change expected. Quarter-circles are drawn through the points for each country so as to facilitate ascertaining the ranking and relative degree of change expected among them. Each graph predicts the same ordering for the four countries expected to shift most--though to different degrees—-and both graphs bunch together the last six countries in somewhat different order. Comparison of the orderings is presented in Table 4-5. Figures 4-7 and 4—8 agree least in their predictions for Congo Kinshasa and Togo, with Togo ranking much higher in probability of high change under a meaSure of development than under a measure of size. The Congo Kinshasa-—a very large country--fares much better 149 High Change‘ x w "'0 m a ”HR u v m u o 5:20 1:03 mg. m o c «33 355 u D -H d? ‘63.: A g o c 088 3,22 -v-l 8:3” Q:H o o m \\\Dahomey ._\\\>\ » \C‘AR\‘~‘ \\\‘\\MK“‘— > Upper Volta ._ ‘ ‘ Low 0 9’ 8. 7. 6. 5. 4. 3. 2. 1. . Change 0 Pre-Coup Military Influence Estimates Figure 4-7.-—Predictions of Degrees of Shift in Foreign Relations After Coups De'Etat Based on Social—Economic Development. Based on the findings of Russett EE_31., percentage population in urban areas is selected as the best single indicator of social-economic development. See their discussion on page 288, World Handbook of Political and Social Indicators, Yale University Press, 1964. M- Decision Latitude Estimates Based on GNP (1964-65) 150 High 3200 1600 800 400 . Sierra Leone Upper Volta 200 F 100 n L o J o o c Q o O o o ow 7 6 5 4 3 2 Change 8 1 0 Pre-Coup Military Influence Estimates Figure 4-8.——Predications of Degrees of Shift in Foreign Relations After Coups D'Etat Based on Social- Economic Size 151 Table 4-5.--Predictions of Degrees of Shift in Foreign Relations after Coups D'Etat Rank Development Size High 1 Ghana Ghana Change 2 Nigeria Nigeria 3 Algeria Algeria 4 Sierra Leone Sierra Leone 5 Dahomey Upper Volta 6 Togo , Central African Rep. tie . Congo Kinshasa 7 Central African Rep. 8 Upper Volta Dahomey 9 Burundi Togo Low Change 10 Congo Kinshasa Burundi under the size measure than under the development measure. In both graphs, Burundi measures very low in probable change. Directions of Change After Coups d'Etat Turning now to the substance of changes in foreign relations expected under military rule, we have argued that the sheer number and rapidity of coups in Africa tends to reinforce the notion that civilian leaders have held power with a minimum of structure, with a decision-latitude in policy-making related to the lack of strong countervailing pressures—-until the military came along--and to the size and development of the state. Policy-formation has thus been 152 largely a function of the personal experiences and outlooks which leaders bring to their offices. In this regard, the military have a common discipline, even a common ideology, which probably surpasses any common influence among civilian leaders. By their training as actors within a modern, technical, pro— fessional institution, the military learn new roles and gain clear—cut identities while being spared the need to create these roles. There is little ambiguity about the characteristics of a "good soldier." By contrast, though, civilian politicians are faced with the requirement to create and fill much less defined roles and to build, from scratch, new political institutions. Civilian leaders are more parochially influenced: they have a less common set of experiences, training courses, and outlooks than the westernized military elite. Civilian policies reflect this difference by being more concerned with articu- late, ambiguous, symbolically-expressed and ideologically-justified foreign policies than the military. If the military have clearer and simpler roles, they will be focused on derivative, tangible and managerial requirements, and so they would tend to interpret symbolic argumentation about policies as a sham and a waste of time, and political conflict as a diversion and an ailment. Janowitz emphasizes the military ethos of public service and national identification, allowing the inference that political instability and weakness are intolerable for the military, whose activities, goals, and identification assume a stable nation (Janowitz, 1964). Cowan expects a "trend away from the inordinate concentration on politics and political organization," "greater attention to 153 immediate problems of development, with a much firmer 'get on with the job' attitude," an effort to "eliminate corruption and increase efficiency, with a strongly reforming and moralistic tinge,“ and a "strong emphasis on standards of public honesty and social discipline" (1968: 22). Following this line of argument, we can see an overlap of several observations made. First, the picture which has been developed of the military in Africa suggests that this group will tend to develop pragmatic, "national-interest," foreign policies-- Zartman's terms-—rather than "ideological“ policies. Second, the dilemmas of foreign policy, as formulated here, would tend to be resolved by the military in favor of tangible short—term needs. Thus, the military tend to see development primarily in respect and economic terms, and are interested in foreign relations largely for "rational status" and for the material advantages of trade and aid. The military are less likely to covet "non-alignment" or to seek relations with either the communist or the non-communist world, because these are embroiling political questions which seem to them more or less unrelated to the issues of development and national loyalty. Of course, their training and indoctrination, largely from the former metropole, probably provided something of an anti— communist bias, and their roles in the new states may reinforce this bias. In fact, the military cost-accounting approach to foreign relations would tend to constrain their international activities generally, and I expect military regimes to be more chary in the 154 numbers of diplomats sent out and more parochial in choosing where they send them. Specifically, then, we hypothesize the following changes in foreign relations when the military take over a government: a greater intensity of trade and aid, concentrated in relations with fewer countries, with less ideological concern for both alignment and nonalignment . Expanding the table of foreign relations areas from Figure 3-9, we get the following table of changes predicted: Trade Aid Diplomacy Intensity Higher Higher lower ,4 a Extensity LOWer Lower Lower B :3 Alignment : '3 with Africa Lower lower Lower Communist Lower Lower Lower West Higher Higher Unchanged Intensity (Not appropriate) Higher lower H m gExtensity (Not appropriate) (Not appropriate) lower «1 :Alignment: j: with Africa (Not appropriate) (Not appropriate) lower g Communist (Not appropriate) (Not appropriate) lower West (Not appropriate) (Not appropriate) Unchanged Fig. 4-9.--Predicted Changes in Foreign Relations after Coups D'Etat .- .__._ -ll..,. - V _ ' _ V . H H”. _ 1,, - CHAPTER V THE DATA AND THE RESEARCH DESIGN Selection of Variables and Data Collection This dissertation is based on library research. All the data used here have been published previously or have been calculated by me from previously published data. Only in this latter sense can I say that I have generated new data. Despite this dependence on previously published data, it is not valid to classify this research as secondary analysis. Nearly all the data used here have appeared in quantitative yearbooks in unanalyzed form, and I have not relied on any other research projects as sources of data. To my knowledge, there simply are not any other studies involving the comparison of the foreign- affairs behaviors of so many countries over such a long period. I have had to accumulate the data for this purpose. The variables in this study were chosen for several reasons. First, a very broad range of kinds of variables was desired in order to allow the least-biased discovery of foreign-relations patterns. I have argued in earlier chapters that the severity of problems in foreign affairs and the novelty to most African states of dealing with these problems, plus the lack of theory pertaining to the foreign-affairs behavior of these states, requires that the researcher begin by 155 156 throwing a data—net as broadly as possible so as not to miss any important aspects. To this end, data were sought on nearly sixty variables affecting foreign relations, although the completed research reports data on only 48 variables. Some variables were eliminated because there were too many gaps in the data and others because they correlated so highly and positively with one or another variable that they could be said to be measuring the same phenomenon. The 48 vari- ables of the completed study are listed in Appendix I. Second, the difficulty of obtaining any measures of foreign relations for the African countries which would be reasonably reliable and comparable across 34 countries and four years is a very great one, necessitating, frankly, a certain amount of taking what one can get for data. Thus, availability influences the choice of variables. There are both strengths and weaknesses in such choice. An advantage of using published statistics from standard sources is the likelihood that such variables have an established meaning which facilitates valid and reliable measurement. Also, I assume that the compilers of international handbooks such as those published by the OECD, the World Bank, and the United Nations—-all with considerably more resources for data-gathering than I--have over the years established standard defini- tions and procedures for gathering data. There are disadvantages, also, in the reliance on handbook data. The major weakness which shows up here is that handbooks tend to concentrate on attribute-data and to slight behavioral data. I am not satisfied that I have enough measures of political behavior, especially of political behavior easily changed by top decision—makers such as statements of policy and aspects of personal diplomacy. Such 157 events-data relevant to international affairs simply are not reported in statistical handbooks. For one thing, "events" are not so easily categorized as other data. The extensive discussion of event—data analysis presented in Chapter Three of the second edition of the World Handbook of Political and Social Indicators says, simply, that "an event is an occurrence or noteworthy happening" (1972:60). That is not much of a definition of events. For another thing, researchers depending on events-data have tended to focus on political violence and irregular changes of governmental executives (e.g., Collins, 1971; Morrison & Stevenson, 1971; Rummel, 1966; Tanter, 1966), subject—areas about which governments tend not to boast or facilitate data—collection. Finally, there is a great deal of work involved in collecting events—- data, along with serious methodological difficulties, and these efforts often do not produce great quantities of data. In any one year in Africa, there simply may not be enough reports of comparable noteworthy events in foreign affairs of the various countries to allow statistical analysis. Thus, the aspects of foreign behavior which might be most sensitive to regime-changes-—and therefore of interest to this study-- are likely to be events difficult to conceptualize and for which adequate data are not available. Third, variables were chosen not only to measure a broad range Of substantive areas of foreign—affairs behavior, but also to measure behavior on three analytical parameters which I have developed. I have trichotomized the substantive areas of foreign behavior into trade, aid, and diplomatic categories, whereas the analytical parameters are intensity, extensity, and alignment. In many cases, variables were created by manipulating handbook data into forms which would better 158 indicate performance on these parameters. For example, the raw data presented in Direction of Trade annuals were used to get indications of the intensity and concentration (extensity) of trade of each African country. By calculating the percentage of gross national product represented by imports and exports, I believe I have a measure of trade-intensity. By noting the numbers of states with which 10% or more of total trade is carried on, I believe I have a measure of trade extensity, or concentration. By computing the differences in the percentages of countries' trade with the United States and the Soviet bloc, I get a measure of trade-alignment. Similarly, data on aid and diplomatic interactions were manipulated to create variables explicitly designed to measure performance on these three analytic parameters. Possible Errors and Reliability Problems Two kinds of errors are possible in the data I have gathered: errors of validity and errors of reliability. Validity involves the question, is the measure devised a satisfactory measure of the property intended for measurement? In this research, we are primarily concerned with measuring accurately the foreign-affairs behavior of African states. That is, we are concerned with defining variables which really reflect what we have called foreign—affairs behavior, at least all its major dimensions. Earlier I have argued that there is vir— tually no theory of foreign relations to guide this research, especially as we are concerned with Africa and the possibility of distinctions between military and civilian governments' orientations on that contin- ent. This lack of theory means there is little guidance in the 159 literature on choosing variables and therefore little basis for judging the validity of variables constructed. H I have responded to this problem in a number of ways. First, have included as many measures of foreign-affairs behavior as I could imagine and also get data on, trying to minimize the possibility that my view of African foreign affairs might be too narrow or biased if I depended on only a few variables. Second, I have postulated some abstract parameters of foreign—affairs behavior in order to establish measures with maximum validity. Because we have no well-tested theory of foreign-affairs behavior and no other general study of the foreign- affairs behaviors of African states against which to compare the results of my work, we have no way--at least in the short run--of answering the question of the substantive validity of my variables. However, my abstract parameters can be evaluated for their logical validity, if not substantive validity. Thirdly, I have cautioned against inferring causal relationships among the variables, since this study is a multi- national overview limited to presenting statements of association or correlation. Fourth, I have tried to control for possible ecological fallacies—-that is, inferring uncritically that continental patterns are applicable to individual countries—-by using nonparametric statistics which are not affected by skewed data, by presenting means for variables, and by presenting the individual data for each country so that any significant deviants can be identified. Data—reliability is another problem with several aspects. Reliability—errors-might creep in because of different definitions of a concept in different countries, different procedures for data— collection, or differently or inadequately trained data—collectors in 160 different countries. As explained in Section I above, I have tried to offset these problems by drawing most of my data from the handbooks of international orgainzations which, I assume, have worked to minimize such errcrs. Even so, we must recognize the limits to the influence of the United Nations and other organizations in attempting to standardize definitions and procedures of data—collection. For one thing, virtually all the data in United Nations handbooks are not collected by UN per- sonnel, but rather by people employed by national governments. States and governments may not be able to resist the temptation to portray themselves in a more flattering than accurate light, and so may alter data to some degree before passing them on to the United Nations. We might fear that this temptation would be especially great in countries such as those of Africa which tend to show the lowest levels of devel- opment in the world. A perhaps even more important consideration is that many of these countries do not have really good statistics to give out even if they wanted to do so, and even if there is perfect understanding of definitions of variables and procedures for collection. Gathering data is expensive and difficult, especially in countries without sufficiently large and well-trained bureaucracies to do so. For young countries like those of Africa, censuses have not been done very often, so there is little precedent for doing the job and little historical context for estimating validity. As an unusual thing, gathering data may cause fear among persons questioned and therefore poor cooperation. Also, the data published may become political footballs domestically, with national governments having to justify their rule in terms of the data released. In Nigeria, for example, a simple head—count of the tial efforts on the part of the government to convince the citizens that the census of 1973 would be accurately conducted and reported, there still is great discussion by ordinary persons and influential politicians alike about the acceptability of the figures. 161 population is cause for great political concern, and, despite substan— The best control and recognition of data with such inaccuracies would come from the international bodies that regularly aggregate the information. Appendix III lists the sources of data for all the vari- ables in the study, showing my dependence on such international bodies. I have further attempted to get reliable figures by often cross-checking 3 data from more than one source; for these cases the multiple sources are listed in Appendix III. The yearly data on each variable for each i country were examined in order to spot any wide fluctuations which might suggest reliability problems. Whenever wide fluctuations were discovered, as compared to the average changes of all countries or the average of the country in question, an attempt was made to verify the data from another source before a figure was accepted. Through these methods, a commonsense judgment of reliability of the data was estab- lished, though I could not compute numerical estimates of reliability. One advantage of a longitudinal and comparative study such as this, however, is the somewhat reduced importance of absolute accuracy in every datum, since reliability can be estimated in the multiple measures assembled over time, and since rankings of countries on the variables can probably be taken as accurate even if the data are not exact. In this study there is no problem with possible sampling error, since very nearly all the countries in the universe under investigation 162 (Africa) are included. This means two things. First, we have a full description of the African countries with respect to the phenomena of interest, and, second, we cannot properly generalize the findings for Africa to other areas of the world. Formation of Indices of Intensity, Extensity, and Alignment In-order to analyze the performance of states on my proposed parameters of intensity, extensity, and alignment of foreign-affairs behavior, composite indices were computed for each country on each parameter for each of the years under study. Each index—-whether for intensity, extensity, or alignment—-is made up of several variables, as indicated in Figures 5-1, 5-2, and 5-3. Substantive Variable Area Number Variable Description Trade 2 Exports/GNP 39 Official aid from multilateral sources Aid 42 Grand total of all aid from all sources 43 Grand total of all aid/GNP 22 Rate of participation in UN roll—call voting Diplomacy 29 Number of countries to which diplomats are sent 33 Total number of members in all IGO's of which the country is a member Figure 5—1. Variables Included in Indices of Intensity 163 Substantive Variable Area Number Variable Description 7 Number of countries to which exports are sent 10 Percentage of total imports coming from the one Trade country sending the largest amount 11 Percentage of total exports going to the one country receiving the largest amount Aid 41 Largest amount of aid from any one country 29 Number of countries to which diplomats are sent Diplomacy 34 Average number of member states in IGO's of which the country is a member Figure 5-2. Variables Included in Indices of Extensity Substantive Variable Area Number Variable Description . 14 Percentage of imports from all African countries Africa . 15 Percentage of exports to all other African countries Trade USA- 46 % of imports from USA minus % from Communist Bloc USSR 47 % of exports to USA minus % to Communist Bloc Africa NA NA Aid USA— 48 Money—value of official aid from USA minus USSR Communist Bloc aid 23 Rate of voting against African majority in UN 32 % of African members in IGO's of which the Africa country is a member 35 Total number of African members in all IGO's of which the country is a member Diplo. “qu USA- 26 Difference: rate of voting against USA minus USSR rate of voting against USSR Figure 5-3. Variables Included in Indices of Alignment 164 Several rules guided the selection of variables to be included in the indices. First, variables from all three substantive areas-— trade, aid, and diplomacy--were included in each index so as to obtain a full picture of performance on each parameter. Second, variables were excluded which correlated very highly with others in the index, so as to minimize computational work and so as to obtain an index in which each variable contributed similar amounts of information. The result of this selection is that nearly all the variables within any one index correlate at less than 0.5. Third, variables were excluded which showed little variation among countries, for the same reasons as rule two above. Fourth, variables with any missing data were excluded, since it was desired to get indices for every country for each of the four years under study. Fifth, with one exception, variables were selected for inclusion in only one type of index so as to maximize the differences among the types of performance being measured by the parameters of intensity, extensity, and alignment. A slightly larger selection of variables was included in the trend-indices computed for the two groups, the coup and non—coup countries. All the variables included are identified by number in the appropriate tables in Chapter Six. The additional variables were included in the trend-indices for these two groups of countries because it was judged that variables with small amounts of missing data could be tolerated when overall group measures were being com- Puted. Also, as Tables 6—20 and 6-22 of Chapter Six show, the trend— indices for intensity, extensity, and alignment have been subdivided into various substantive areas, including bilateral versus multi— lateral relations, and trade, aid, and diplomatic areas. Alignment 165 has been further subdivided into measures of alignment with the rest of Africa, with the communist bloc, and with the United States. A large number of variables were thus required in the trend—indices so as to allow these subdivisions. Method of Analysis General Aspects of Design As Campbell and Stanley point out, "good experimental design is separable from the use of statistical tests of significance. It ["good" design] is the art of achieving interpretable comparisons," and "in all such cases, the interpretability of results depends upon control over the factors we have been describing (which jeopardize validity)" (1963:22). An ideal research design here would therefore be one in which a group of countries, or more accurately several national governments, were each given an equivalent "treatment" or presented with a similar disruptive situation, and the changes in behavior of these governments were noted. These noted changes would then be compared with theoretically-predicted changes in order to evaluate theory. Ideally, there should be a control group of govern- ments, demonstrably equivalent to the experimental group, whose behaviors were also monitored over the same time period but which did not undergo the experimental treatment or disruption. In this research, we have an approximation of such an ideal design for analysis. We have two groups of countries whose behaviors are compared, i.e., an experimental group of coup countries all of whose governments have been replaced as a result of military coups d'état, and a control group of countries whose governments have ..-a.:..._; “ham- .x’ . 166 remained civilians with the same persons in office over the whole period under study. These two groups essentially exhaust the popula- tion of countries under study, Africa, and I have argued that the groups appear to be substantially equivalent because nearly all African countries share many important characteristics determining governmental performance, such as colonial histories, recent indepen- dence, very low levels of social and economic development, ethnic bonds across states but great social heterogeneity within states, and so forth. However, I cannot prove equivalence between the groups. Mem- bers in the groups are, in a sense, self—chosen, not randomly assigned by me. Group-sizes are quite different. On the other hand, as I have shown in Chapter Two, these groups do not seem significantly different on most social and economic attributes; they differ importantly only on population—density and agricultural land per capita, among the attributes measured, and these seem to be more aspects of size than of social-political performance. I am interested in comparing the foreign-affairs behavior of African military and civilian groups. Furthermore, baselines of behavior for the two groups are drawn up for the period immediately preceding the military coups d'état, when all African governments were civilian. We have a kind of multivariate design, but not as this is usually understood. Here we have really only one independent variable, the dichotomy between military and civilian governments, although this is a rather complex distinction. Military coups d'état do not simply influence foreign policies, they actually change the personnel and structures of governments. We have several dependent variables, 167 however, all of which are aspects of foreign relations, and these variables are grouped into substantive categories of trade, aid, and diplomacy and into analytic parameters of intensity, extensity, and alignment. With pre-test and post-test measures on two groups of govern— ments, we have a quasi-experimental design most like the nonequivalent— control-group type identified by Campbell and Stanley (1963:47—50). Here the exact equivalence of the groups cannot be established, and there is no control over assignment to groups to facilitate estab- lishing equivalence. Nonetheless, the groups are reasonably similar and their behaviors are subjected to the same measures over the same periods, so the design controls for the main effects of history, maturation of the countries, and testing-procedures, which might otherwise confound interpretation of results. Analysis of Yearly Data for Baselines Data on all the variables have been collected for each of the thirty-four countries in the study for each of the four years from 1964 through 1967. Before analyzing the foreign-affairs behavior of military governments directly, baselines of foreign-affairs activities were established for all these countries in the period immediately preceding the coups d'état. 1964 is this baseline year. All thirty- four countries had civilian governments in that year, most being headed by the nationalist leaders who had taken control at independence a few years before. Ten countries changed governments through military coups d'état in the period from June 1965 to January 1967. Four variables were selected for detailed analysis to establish baselines in 1964. These variables are exports/GNP (#2), all aid received/GNP (#3), the rate of voting against the United States in the United Nations minus the rate of voting against the Soviet Union (#26), and the total number of memberships in international governmental organizations other than the United Nations (#31). These four variables were selected from among the forty-eight variables for which data were gathered because it was judged that they represented the substantive and analytic areas of interest in the study and because analysis of individual country scores was facilitated by concentrating on a minimum number of variables. In the later analysis of military and civilian groups, changes in behavior from 1964 baselines are described on all forty-eight variables, though only with respect to each of these groups as a whole, not with respect to individual countries. Baseline data are necessary for at least three reasons: First, the data for 1964 provide a starting point before the coups d‘état to measure changes in foreign-affairs behavior, for both the continuous civilian governments and the new military governments which arose sub- sequently. Second, scrutiny of the performance of individual countries in the "pre—test" year, 1964, allows the identification of states with unusual or very deviant behavior, as compared with others in the sample. It is important to discover any extreme "deviants“ in the sample at this pre-test stage in order to interpret properly the changes in foreign-affairs activities in later years. Third, analysis Of baseline activities provides a picture of overall continental behavior as a context within which to compare the coup and non-coup groups of countries. 169 In addition to analyzing the performances of individual states on the four variables mentioned, baselines of activities have been established on the three analytic parameters. Countries were ranked on their indices, and countries with the most extreme behavior on the parameters have been plotted on my theoretical cube of international actor types. Analysis of Continental Trends, 1964-1967 All the countries in the sample are analyzed for their behavior in 1967 on the four variables mentioned above in order to assess the continental pattern of behavior in the last year under study and the pattern of change from the first year under study. In addition to analyzing the rankings of the countries on the variables and parameters for 1967, the rankings of countries are described and evaluated on the amount, percentage, and direction of changes in performance on the selected variables. Countries showing the most extreme changes in performance on each of these four variables are discussed at length in order to clarify patterns of change. Also, the relationships among the four variables, as measured by changes in their intercorrelations, are assessed both numerically and qualitatively, in order to discover the most important shifts in continental patterns of foreign behavior. Thus, we have not only the 1964 baselines against which to measure the performances of the coup and non—coup groups, but the overall contin— ental patterns of change-—rate, amount, and direction of changes in Substantive and analytic areas--against which to measure the changes of each group. Analysis of Two-Year Trends for Assessing Effects of Military Coups on Foreign Behavior In order to assess as carefully as possible the changes in foreign-affairs behavior brought about by military takeovers of govern— ments, the behaviors of states in the coup group were measured from one year before the coup to one year after. For those states having coups in 1965, that means measuring their behaviors from 1964 through 1966; for those states having coups in 1966, it means the period 1965 through 1967. The "control" group of non—coup countries was subdivided into two groups, one studied for the period 1964 through 1966 and the other for the period 1965 through 1967, in order to enhance the validity of comparison with the coup states. Since 30% of the coup states had coups in 1965, and therefore were studied for trends over the period 1964 through 1966, 30% of the "control“ group were randomly assigned to the control subgroup being studied for that period and the remainder to the subgroup studied for the period 1965 through 1967. In fact, of course, the analysis of yearly data does not allow the assessment of foreign—affairs behavior from precisely one year before a coup to precisely one year after. Coups do not occur at one time-point in any year; they occur over a period of months, yet my behavioral data are available only in a form aggregated for calendar years. Thus we have the slight anomaly of countries with coups six months apart in l965——Algeria in June and Dahomey in December-—being' a part of the group whose foreign—behavior trends are assessed over the period 1964 through 1966, and the greater mis-match of states with coups one year apart (Nigeria in January of 1966 and Togo in January Of 1967) being included in the group whose trends are assessed from 171 1965 through 1967. Unfortunately, the foreign-activities data simply are not available in smaller time—slices, whether monthly or weekly or daily, so the individual state behaviors cannot be more strictly com— pared from exactly one year before to one year after the coups. However, this is not as much of a problem as might initially be feared. Because the data are aggregated for all the days within any one year, the trends derived here-—whether for the period 1964 through 1966 or 1965 through 1967—~are based on yearly data which include the periods of one year before the coup to one year after for all the coup countries. Thus, each country's trends are assessed over the period from the calendar year before the coup to the calendar year after the coup, although that usually does not coincide with the periods of 365 days before and 365 days after the coup. The two—year period was selected for assessing trends in for- eign-affairs behaviors because this period was judged to be broad enough to allow for significant behavioral changes to occur as a result of changes in government through military coups, but not so broad as to include many changes for reasons extraneous to the coups. If there is a peculiar orientation of military governments in foreign affairs, that ought to be visible throughout all the days after military rule is consolidated, whether one month after a coup, one year, or several years. However, foreign affairs are patterned by several influences. Not only are the peculiar orientations of particular regimes important, but also the general domestic conditions of the states and their situations in international space and time. I judge that at a time quite distant from a military coup, say five or ten years later, it would be difficult to sort out these influences on foreign affairs, 172 whereas in the first year or so after a coup the greatest changes in the pattern of foreign—affairs activities are likely to be the result of the abrupt change to a military government through a coup. The analysis of longer time—periods would be useful to the research objective here in that such longer periods would facilitate the separation of peculiar regime—influences on foreign affairs from peculiar situational influences. The primary objective here, however, is establishing a distinction, if any exists, between the orientations of two types of regimes, military and civilian, and this objective seems better served by study of a narrower period. The two—year period is the narrowest possible for analysis by the full range of variables assembled here. Nonparametric Testing In order to facilitate evaluation of the differences in per— formance of military and civilian governments, statistical tests of significance have been computed on all the trend variables and on the trend indices measuring the parameters of intensity, extensity, and alignment. In addition, the rankings of countries on yearly data——both for substantive variables and for analytic indices-—have been tested in order to determine probabilities of discriminating between the coup and non—coup (i.e., military and civilian) groups. Throughout the study, one of the so-called nonparametric tests, the Mann-Whitney U test, has been used. A nonparametric test has been dictated because of the form of the data. It is fortunate that all the variables in the study except for the distinction between military and civilian (i.e., between coup and non-coup) could be 173 measured on at least an equal—interval scale, and usually on a ratio scale. Despite the fact that these data allow full arithmetic manipu— lation, the form of the data is invariably not a normal distribution. Thus, parametric tests, such as analysis of variance, cannot be used, since they require that data fit a theoretical distribution, usually the normal distribution, with roughly equal variances and means among the samples. The Mann-Whitney U test is chosen because it is a powerful test for comparing two independent samples on variables measured on at least an ordinal scale. Although our data are measured on an equal-interval scale or better, the distribution on the variables are nearly always highly skewed, that is, they have considerably more extreme cases in one direction than in the other. This skewness would distort para— metric tests using means, and we are interested in comparing our samples on their central tendency. In this regard, the Mann-Whitney test is probably the most powerful nonparametric test available. If we were interested in comparing our two groups on some other parameter, such as their dispersions, then another test might be more powerful, for example, th Wald—Wolfowitz runs test, or the Kolmogorow—Smirnov test. Siegel (1956:126—27) and Blalock (1972:260—62) cite studies which con- clude that for moderate-size samples, such as those in this study, the Mann—Whitney test is approximately 95 per cent as powerful as the Eftest, which would be the alternative among parametric tests if the Eftest's assumptions could be met by the data. The Mann—Whitney test is essentially a test for comparing the distributions of ranks in two sets of scores. When comparing samples where at least one sample has more than twenty cases, U scores can 174 be transformed into 2 scores, and the normally distributed 2 scores can be used to calculate probabilities for the tests of significance. In most instances where tests of significance are presented in this study, the U scores and z scores are presented along with the resultant probabilities of discriminating between the two samples. In addition, in tables of trend variables and indices presented in Chapter Six, all variables and indices are listed with their probabilities, instead of presenting only those arbitrarily judged significant. It seems as important to know on which variables and indices the military and civilian governments differ not one whit as to know where they differ quite significantly. By presenting the full list of variables and indices on which the samples were measured, there is no need to select arbitrarily a significance level. All the variables and indices are included in the interpretative discussion, and that discussion is enhanced by comparison of all their probabilities or significance levels. CHAPTER VI ANALYSES OF DATA Baselines of International Activity, 1964 Looking first at Table 6—1, we get some picture of the inter- national activities of 35 countries on selected measures of trade, aid, and diplomacy. Each of the four variables displayed in the table is an indicator of maximum political variability among the countries. Thus, the variables of Exports/GNP and Aid/GNP are given, rather than measures of exports and aid alone, since these latter two variables would be distributed very nearly like GNP, as their intercorrelations indicate, and I assume that variation in GNP is much less amenable to political manipulation than are rates of export and aid. Several observations are possible based on these data for 1964. First, the "coup countries" are widely dispersed on each of the vari— ables. Second, each variable shows a substantial range. The difference in scores from top to bottom in each ranking involves at least a factor of 25, yet the difference in scores for countries ranked adjacently is small. Thus, there is enough variation to make analysis important, and there are no extreme scores to distort analysis. Third, there are very few similarities in the distributions of coup countries as against non— COup countries across these four variables. 175 176 Table 6-l.—-International Activities in 1964 (Coup countries in parentheses) Rate of Voting in UN Total Number Exports/GNP All Aid/GNP Against USA—~Rate Against IGO Memberships (Variable 2) (Variable 43) USSR (Variable 26) (Variable 31) Zambia .883 Congo Braz. .338 Mali 469 Morocco 25 Libya . 854 Somalia . 273 (Algeria) 455 UAR 23 abon .686 Malawi . 210 Kenya* 450 Tunisia 19 Liberia . 655 UAR . 175 Zambia" 394 Sudan 12 Mauritania . 425 (Algeria) . 160 UAR 367 (Ghana) 11 Ivory Coast .377 (CAR) .140 Morocco 359 9 Uganda . 349 Tanzania . 131 Tanzania 341 (Nigeria) 9 (Sierra L.) .322 Kenya .130 (Ghana) 334 Cameroon 8 Congo Braz. .310 Guinea .114 Uganda 307 (Congo K.) 8 Tanzania . 303 Mauritania . lll (Burundi) 291 Guinea 7 (Algeria) . 26S (Togo) . 111 Guinea 291 Liberia 7 (Burundi) . 250 Mali . llO Tunisia 282 Malagasy 7 Malawi . 225 (Dahomey) . 108 Ethiopia 269 Mali 7 Somalia . 224 Gabon . 092 Sudan 230 Mauritania 7 (Congo K.) .222 Senegal .091 Somalia 218 Senegal 7 (T090) . 222 Liberia . 087 Chad 192 Gabon 6 Cameroon .207 Chad .078 (Nigeria) 192 Niger 6 (CAR) . 195 Malagasy . 07B (Dahomey) 180 Tanzania 6 (Ghana) .193 Tunisia .077 Libya 180 (Algeria) 5 Morocco .182 Uganda .073 (Upper V.) 180 (CAR) 5 Senegal .172 (Upper V.) .070 Congo Braz. 177 Chad 5 (Nigeria) .171 (Congo K.) .068 Mauritania 177 (Dahomey) 5 Kenya .167 (Burundi) .065 (Togo) 166 Ivory Coast 5 Malagasy .150 Rwanda .061 (Sierra L.) 154 (Upper V.) 5 Sudan .148 Niger .057 Niger 153 (Sierra L.) 4 Tunisia .137 Ivory Coast .050 Senegal 139 Congo Braz. 3 Guinea . 136 (Ghana) . 047 Malawi* 120 Ethiopia 3 UAR -. 125 Cameroon . 046 (CAR) 116 Kenya 3 Chad .124 Morocco .044 (Congo K.) 115 Somalia 3 Ethiopia .097 (Sierra L.) .036 Cameroon 90 Uganda 3 Niger .082 Zambia .034 Ivory Coast 89 (Togo) 2 (Dahomey) . 078 Libya . 022 Gabon 77 Malawi 1 (Upper V.) .055 Sudan .019 Rwanda l3 Rwanda 1 MEli . 050 Ethiopia . 016 Malagasy 0 Zambia 1 Rwanda . 001 (Nigeria) . 01 5 Liberia —39 (Burundi) 0 *Data. for 1965. U=ll4 U=107 U=121 U=103 Z = -.402 z = —.658 z = -.146 z = -.804 '0 = .3439 p = .2554 p = .4220 p = .2109 177 We may infer that none of these measures of trade, aid, and diplomacy says much about how to predict the occurrences or probability of coups d'état. Of course, these data are drawn for 1964, before any of the coups, and it may be that when these measures are taken one to two years before coups they are too remote in time to be associated with the coups. Secondly, the African countries show widely varying degrees of international activity on each measure, and the score on any one measure is a poor predicator of the score on any other. We see this latter fact in the matrix of intercorrelations in Table 6-2. For 1964, the highest intercorrelation among the four vari— ables is 0.434, and the other five are below 0.300. If it were proper to speak of significance levels here, a correlation of about 0.330 would be significant at the 0.05 level for this "sample" of 35 states. Thus, only one correlation of the five, that between total number of IGO memberships and the rate of voting against the USA minus the rate of voting against the USSR, could be considered significant. In addition to these variables, four other variables have been added to the correlation matrix in Table 6—2 in order to show some strong patterns. These additional correlations show that the total amount of trade——imports plus exports--is very highly correlated with GNP (0.919). One would hardly expect it to be otherwise. More unex— pected are the high correlations (0.776) between the total amount of aid received and GNP, and the similarly high correlation (0.747) between the total amount of aid and total amount of trade. Apparently, the larger economies not only carry on a greater volume of trade than do smaller economies, but the larger also receive more assistance from 178 Table 6-2.——1964 Intercorrelations Variable—numbers in Parentheses (5) (42) (43) (26) (31) (20) (21) Exports/GNP (2) .176 —.148 -.096 -.283 —.144 -.149 .623 Imports + Exports (5) 1.000 .747 —.151 .345 .544 .919 .433 Total Aid (42) 1.000 .282 .457 .456 .776 .199 All Aid/GNP (43) 1.000 .060 —.156 -.114 -.089 Voting USA-USSR (26) 1.000 .434 .420 .039 IGO Memberships (31) 1.000 .644 .337 GNP (20) 1.000 .168 GNP/capita (21) 1.000 aid—giving sources. Intuitively, this relationship seems reasonable in that larger economies can absorb and utilize more money than can smaller economies and that donor countries may expect bigger pay—offs from donations to bigger countries; it seems less reasonable if we argue that smaller economies may need greater external assistance in order to generate and sustain their growth and development. The larger economies also show their strength in the rather high correlation (0.644) between GNP and the number of intergovern— mental organization memberships; the greater the GNP, the more inter— governmental organizations the state is likely to belong to. ‘Finally, extrapolating from the correlations, it appears that GNP alone predicts the number of memberships in IGO's much better than does the other measure of economic strength given, GNP/capita, though the correlation between GNP/capita and number of memberships in IGO's is still » e ’7 -)V Ariel's-r 179 substantial, or "significant," at 0.337. Curiously, though, GNP/capita predicts Exports/GNP better than any other variable in the table (0.623), suggesting that in 1964 high per capita wealth was a function of an economy which produced a high rate of exports. In order to get some perspective on the African countries‘ scores on these variables, I have computed scores on one of the vari— ables, Exports/GNP, for 22 other arbitrarily chosen countries, for comparison. This variable was chosen for illustration in part because data was readily available for such a heterogeneous set of non—African states. Also, illustration of scores on the other variables is either irrelevant or less useful. Many of these non-African states are aid— givers rather than receivers, so scores on Variable 43 would not be very useful. Eyeballing the data on the other two variables suggests these two general conclusions: First, the non-African countries of the world are distributed over a broader range of scores on United Nations voting than are the African states. As has been reported elsewhere (Alker and Russett, 1965), voting-blocs can be discerned in the United Nations, and the African states have their own pattern of voting. In general, one aspect of their pattern of voting is a tendency to vote much more often with the Soviet Union than with the United States . Second, eyeballing suggests that African states tended in 1964 to have fewer memberships in international governmental organizations than did other states. This does not seem a surprising observation, since so many African states were new, small, and/or poor at that time. Table 6—3 shows that Exports/GNP for these quite different 22 non—African countries show scores distributed in a fashion quite Table 6.3-—Exports/GNP for Selected Non-African Countries, 1964 Hong Kong .7321 Saudi Arabia .6416 Iraq .4402 Venezuela .4201 Netherlands .3332 Denmark .2302 Ceylon .2066 Bolivia .1790 Thailand .1379 United Kingdom .1368 Israel .1303 Australia Argentina France Japan Brazil Indonesia Lebanon Turkey United States India Soviet Union .1272 .1066 .0934 .0766 .0641 .0611 .0545 .0433 .0386 .0341 .0200* *Subect to much greater error because of great uncertainty in estimates of both trade and GNP. Export data are taken from Direction of Trade, Annual 1963—67; GNP data taken from the United Nations Statistical Yearbook, 1969. 181 similar to the African states studied here. If there is any appreciable difference between the distributions of the African and non—African states, it is in the slight tendency for the African states to export a larger portion of their GNP than do the other states. Foreign Behavior on Analytic Parameters, 1964 Looking next at Figure 6—1 for country rankings of the para— meters of intensity, extensity, and alignment, we find that in 1964 the "coup countries" are spread over nearly the full range of rankings of all parameters. Mann-Whitney U tests reported at the bottom of each column on the table show clearly that there is no difference between the distributions of "coup countries" and "non-coup countries" on any of the parameters. Just as the four variables of trade, aid, and diplomacy considered in Table 6—1 did not differentiate the coup from the non—coup groups, neither does ranking on these analytic parameters. It does appear, however, that there is some tendency for countries which are larger in population and economy to be more intense, more extensive, and more aligned. Plotting five "coup countries" and eight non—coup states with the most extreme rankings on all the parameters, we get the pattern on Figure 6-2. No other states could be satisfactorily represented on this Figure because at least one parameter score was near the middle, Where plotting would have been quite ambiguous in this two-dimensional representation of a three—dimensional cube. The Figure does allow some valid interpretations, however. including three First, there is a distinct cluster of states, large countries from the "coup" group, at corner A, which I have HI \DGJQO‘U'Ifi-UJNH H H H H H H H m m A w m H o 17 Figure 6—1. Intensity Rwanda Zambia Burundi Kenya Malawi Sierra Leone Upper Volta Niger Dahomey Uganda Ethiopia Chad Gabon Libya we gig Guinea Somalia Mauritania Sudan Tanzania Malagasy Rep. Cameroon Liberia Mali Congo Brazza Ivory Coast Nigeria Congo Kinshasa Morocco Senegal Tunisia m Algeria U = 115 z = -.l89 p = .4251 182 Extensity Rwanda Malawi Malagasy Rep. Burundi Dahomey Zambia Niger Chad Cameroon Tanzania Sierra Leone % Upper Volta Gabon Congo Brazza Algeria Ivory Coast Senegal Kenya Liberia Uganda Mauritania Somalia Guinea Togo Tunisia Congo Kinshasa Morocco Mali Libya Ethiopia Nigeria 911129 Sudan U = 116.5 2 = —.l323 p = .4474 Alignment Rwanda Zambia Upper Volta Malawi Niger 292 Chad Dahomey Malagasy Rep. Mauritania Senegal Congo Brazza Sierra Leone Kenya fl Gabon Cameroon Tunisia Ivory Coast Tanzania Libya Sudan Nigeria Uganda Burundi ‘glgeria Congo Kinshasa fiiberia Somalia Guinea we Morocco Mali Ethiopia U = 118 z = —.0756 p = .4699 1964 Country Rankings on Parameters ("Coup countries“ are underlinea 183 Consumer Vulnerable F G L 0 Rwanda Alignment I Mala“ Malagasy I Rep. I 7 I z I Protege l Hig l Burundi B C Vassal Low | I l | Upper I Volta Senega? Independent E/L‘ ——————————————— ./ Fe 1e p / if, / / c E / m Nigeria Guinea C go K. Ethio ia Hi F ’Eh Morocco Mali P : ana D Calculating algh Intensity Low Aggressrve Figure 6-2.——Typology of Selected Countries in 1964. 184 labeled "aggressive." The most aggressive is Ghana, measured here a little over a year before Kwame Nkrumah was deposed. These states show high intensity, high extensity, and high alignment in foreign affairs, and in all cases their alignment was a preference for the Soviet Union over the United States. By and large, they appear to be among the biggest states of Africa, with the capability that sheer size of popu- lation and economy allows. The exception in the group is Mali. At the opposite corner of the cube, we find Rwanda and Malawi, which I would label "vulnerable." Clearly, these are small, weak states having few links with the rest of the world. Among the other cOuntries, Malagasy appears to be the strongest “consumer," Senegal the most "independent," Burundi the clearest "vassal," and Ethiopia the most "calculating." None of the African countries comes close to being "feeble" or a "protege." Figure 6-3 is a plotting of the ten "coup countries" on the analytic cube. These countries appear to fall, by and large, along a diagonal line running between the extreme corners A and G. None of the ”coup countries" comes close to corners B, D, E, F, or H. Thus, none can be classified as protege, calculating, independent, consumer, or feeble. Only Burundi comes close to corner C, indicating a vassal stance. Country-size seems to make a difference in accounting for actor-style: the four biggest of the states all tend to be aggressive actors, especially Ghana, Congo Kinshasa, and Nigeria. There is not a Perfect relationship between the smaller "coup" states' sizes and their positions in the cube, however. 185 Consumer Vulnerable F G Low Alignment | I Dahomey l Protégé I Burundi Hig I B 1 C Vassal Low I I Upper I Volta | CAR | I Sierra L. I Independent E ) ________ a ________ Algeria //// Fe 1e >‘ / P Togo a / 5 / g /' Nigeria m Coygo K. / High / Ghana A _ D Calculating High Inten51ty Low Aggressive Figure 6-3.--Typology of Coup Countries in 1964. 186 International Activity, 1967 Table 6-4 ranks the African countries on the same variables as did Table 6—1, but the data here are for 1967. Some similar observa- tions may be made for both tables. As in Table 6-1, the data in Table 6-4 show that the "coup countries" are dispersed throughout each of the variables. Again, each variable has a substantial range, with enough variation to make analysis important, but no extreme scores to distort analysis. There is again little similarity in distributions of countries across the four variables. In some ways, however, the data in Table 6-4 are differently patterned from the data in Table 6-1. First, there is less variation in scores among countries in Exports/GNP and All Aid/GNP for 1967. No states reach such high rates of exports nor such high rates of aid in 1967 as some did in 1964. Second, most African countries voted sub— stantially more heavily against the United States in the United Nations in 1967 than they did in 1964. The exceptions to this trend are seven countries——Ghana, Uganda, Gabon, Malawi, Sierra Leone, Kenya, and the Ivory Coast, in order of voting more favorable to the United States--which voted in patterns more favorable to the United States as compared to their voting in 1964. Of these seven, three——Ghana, Gabon, and Uganda-~voted more often with the United States than with the Soviet Union. The other four states which showed a voting pattern more favorable to the United States than in 1964, plus all the other twenty—eight states scored, still showed in 1967 an overall pattern more favorable to the Soviet Union than to the United States. Because Of these changes in this variable, this pattern of United Nations Voting shows a substantially greater range of variation in 1967 than 187 Table 6—4.‘v-Internationa1 Activities in 1967 (Coup countries are in parentheses) Rate of Voting in UN Total Number Exports/GNP All Aid/GNP Against USA-~Rate Against IGO Memberships (Variable 2) (Variable 43) USSR (Variable 26) (Variable 31) Liberia .638 Liberia 11:64 Mali 616 UAR 29 Libya . 632 Malawi . 140 (Burundi) 571 Morocco 28 Gabon .606 Congo Braz. .123 Morocco 550 Tunisia 28 ambi .454 Somalia .114 (Algeria) 545 Ivory Coast 25 Mauritania .444 (CAR) .108 UAR 535 (Nigeria) 25 (Congo K.) .357 Tunisia .107 Congo Braz. 522 Cameroon 22 Ivory Coast .304 Rwan a .104 Guinea 500 (Algeria) 21 Congo Braz. .290 (Dahomey) .099 Libya 500 Tanzania 20 (Algeria) . 259 Mauritania . 093 (Togo) 487 (Ghana) 19 Tanzania .849 (Upper Volta) .077 Sudan 478 Kenya 19 Uganda . 243 Zambia . O77 Tanzania 475 Uganda 19 Malawi .216 Chad .076 Mauritania 445 (Upper V0.) 19 (Sierra L.) .184 (Togo) .070 Zambia 445 Chad 18 Senegal .182 iger .068 Tunisia 442 Malagasy 18 (Burundi) .161 (Burundi) .065 Somalia 441 Ni er 18 Cameroon . 158 Gabon . 065 Chad 4 29 Senegal 18 Morocco .157 (Congo K.) .064 (CAR) 406 (Congo K.) 17 (Togo) .156 Senegal .062 (Congo K.) 395 Sudan 17 (CAR) .155 Malagasy .059 Cameroon 381 (CAR) 16 Tunisia .153 Cameroon .050 Senegal 379 Congo Braz. l6 Somalia . 14 5 Mali . 047 Kenya 378 (Dahomey) l6 enya .144 Ke ya .044 (Nigeri 341 Gabon 16 Malagasy .141 Tanzania .04]. (Upper Volta) 341 Mauritania 15 (Nigeria) .138 (Algeria) .034 Et iopia 334 (sierra L.) 15 Sudan . l 38 (Ghana) . 033 nda 258 (Togo) 13 Guinea .118 Ivory Coast .032 Niger 250 Libya ‘12 UAR .114 Guinea .031 (Dahomey) 217 Malawi 11 Chad .102 Uganda 027 Malagasy 103 Ma i 10 Niger .079 oro .026 Ivory Coast 85 (Burundi) 9 (Dahomey) .072 (Nigeria) .023 Liberia 71 Rwan a 8 (Upper Volta) .069 Sudan .023 (sierra L.) 48 Zambia 8 Ethiopia .065 (sierra L.) .020 Malawi 0 Guinea 7 Mali . .044 Ethiopia .019 Uganda -28 Ethiopia 6 Rwanda . O36 UAR . 014 Gabon -126 Liberia 6 (Ghana) . 002 Libya . 001 (Ghana) -133 Somalia 4 U=100 U=121 U=112 U=l32 2 = -.9131 z = -.1461 z = —.4748 2 = .2557 p = .1806 p = .4420 p = .3175 p = .3991 _ .,..x." V... ,rxl...r , 188 in 1964, in contrast to the reduction in ranges for the first two variables. It appears, then, that for the continent as a whole, economic transactions with the rest of the world became more restrained over the years from 1964 to 1967, whereas diplomacy in the United Nations became active and more divisive among the African states. With regard to memberships in intergovernmental organizations, the range of variation is the same in 1967 as in 1964, although all but two of the states--Guinea and Liberia—-show memberships in more IGO's. Most of the states show a substantial increase in memberships, although the states which showed the highest number of IGO memberships in 1964 tend to show a smaller increase by 1967 than do the other states. .This phenomenon seems reasonable, since there is a finite number of IGO's and there must be a point of diminishing returns for these relatively small African states as they calculate the benefits of memberships in more IGO's. Again in 1967, Mann-Whitney U tests show that the "coup countries" are not clustered significantly differently from the other countries on any of the variables. However, the changes in probabili— ties of difference between the two groups shows that the "coup group“ is appreciably more distinguishable as a group in 1967 in Exports/GNP and in United Nations voting with the big powers. The coup and non— coup states are less distinguishable as separate groups in Aid/GNP and ISO memberships. Table 6-5 displays the intercorrelations among the same vari— ables as did Table 6-2, though for 1967. Many relationships and Patterns discernible in these 1967 correlations are similar to those in 1964. Total amount of trade——imports plus exports-~remains very 189 Table 6-5.—-l967 Intercorrelations Variable—numbers in Parentheses (5) (42) (43) (26) (31) (20) (21) Exports/GNP (2) -.121 -.100 -.109 .134 -.051 -.l35 —.065 Imports + Exports (5) 1.000 .621 -.533 .241 .448 .859 .577 Total Aid (42) 1.000 —.l39 .143 .524 .692 —.024 All Aid/GNP (43) 1.000 —.189 -.362 —.567 —.260 Voting USA—USSR (26) 1.000 .039 .227 .052 IGO Memberships (31) 1.000 .559 .033 GNP (20) 1.000 .171 GNP/capita (21) 1.000 highly correlated with GNP (0.859), as does the high correlations (0.692) between the total amount of aid received and GNP, and the similarly high correlation (0.621) between the total amount of aid and total amount of trade. Thus, the intercorrelations show that the larger economies continue to carry on a greater volume of.trade and receive a greater amount of aid than do the small economies, although the relationship is slightly lower in 1967 than in 1964. Also, GNP shows a high correlation with memberships in IGO's again (0.559), although the relationship between GNP/capita and IGO memberships drops substantially to 0.033. Among the four activities—variables displayed in Table 6—4, the intercorrelations are even lower in 1967 than in 1964. As in 1964, the highest of these is barely "significant," at —O.362. The other five intercorrelations among these four variables are below 190 0.200. Overall, these intercorrelations show less association among these four variables in 1967 than in 1964.' Patterns of Analytic Parameters, 1967 Figure 6—4 displays the rankings of the countries on the analytic parameters of intensity, extensity, and alignment, for 1967, and should be compared with Figure 6—1. As in 1964, the "coup countries" are spread over nearly the full range of rankings of all parameters in 1967. Again, the Mann—Whitney U tests show that there is no significant difference between the distributions of "coup countries" and "non-coup countries" on any of the parameters, although the coup group clusters somewhat more clearly in 1967, especially in extensity. The rankings for 1967 suggest, however, that there is a slightly less systematic tendency for countries which are larger in population and economy to be more intense, more extensive, and more aligned. Thus, variables other than those of sheer size apparently are operating more forcefully in 1967 to determine the intensity, extensity, and alignment of each state in foreign affairs. Figure 6-5 shows a plotting of the same five "coup countries" and eight non-coup states as did Figure 6—2. No cluster appears as clearly in 1967 as it did in 1964, although there are again more states near the "aggressive" type of actor than at any other corner of the cube. Overall, it appears as if the states all tended to become more moderate on these parameters of foreign affairs. None Of these states comes really close to the actor—types relabelled protege, vassal, independent, consumer, vulnerable, or feeble. The most aggressive states are still the largest states in terms of population Figure 6—4.——1967 Country Rankings on Parameters (Coup Countries Underlined) Intensity Extensity Alignment Low 1 Rwanda EAR Rwanda 2 Guinea Upper Volta Upper Volta 3 Burundi Niger Dahomey 4 Sierra Leone Chad Malawi 5 Libya Malawi Senegal 6 Mali Congo Brazza Niger 7 Gabon Cameroon Ivory Coast 8 Togo Malagasy Rep. Malagasy Rep. 9 Uganda Algeria Sierra Leone 10 Somalia Gabon Gabon ll Malawi Dahomey Zambia 12 95% Senegal Chad 13 Chad $233 Uganda l4 Dahomey Burundi Nigeria 15 Ethiopia £§vory Coast 2222 16 Zambia Liberia Mauritania 17 Upper Volta Somalia Tanzania 18 Congo Brazza Mali Cameroon 19 92223 Zambia Libya 20 Niger Mauritania E53 21 Mauritania Uganda Kenya 22 Malagasy Rep. Kenya Ghana Figure 6—4.—-Continued 192 Intensity Extensity Alignment 23 Tanzania Rwanda Congo Brazza 24 Sudan Sierra Leone Algeria 25 Ivory Coast Congo Kinshasa Somalia 26 Kenya Tanzania Tunisia 27 Liberia Morocco Congo Kinshasa 28 Algeria Ghana .Liberia 29 Congo Kinshasa Tunisia Morocco 30 Cameroon Guinea Sudan 31 Morocco Libya Burundi 32 Nigeria Nigeria Guinea 33 Senegal Ethiopia Ethiopia High 34 Tunisia Sudan Mali U = 128 U = 136 U = 126 z = .3024 2 = .6049 2 = .2268 p = .3812 p = .2726 p = .4102 193 Vulnerable Consumer G 1" Lo [ Upper Volta Alignment I I Malawi l I Protégé ’ Hig I B 41 Vassal Low l l Rwanda sqnegal l Malagasy R. l Burund Independent E l__ ________ L _______ H / / F ble >1 / .fi / g / . . ; 0 / Nigeria i2 / m 1 / Congo K. ‘ / Ghana Mofocco \ ngh Ethio ia _ A . D Calculating High Intensity Low Aggressive Figure 6—5.——Typology of Selected Countries in 1967. 194 and GNP. Guinea is the only of these states which has clearly moved much closer to one of the corners of the cube; it has become the most calculating actor on the continent. Figure 6—6 is a plotting of the ten 'coup countries" on the analytic cube, comparable to Figure 6—3. To the extent that there is still a discernible pattern in their plotting, these countries appear to fall again along a diagonal line running between the extreme corners A and G. However, it is also clear that these countries have moderated their behavior. None of these countries is as aggressive as in 1964, and none is really vulnerable. Indeed, none of the "coup countries" comes really close to any of the corners of the cube. Changes in Selected International Activities, 1964—1967 Table 6—6 displays the rankings of African countries in terms of the amounts of change calculated over the period 1964-1967 on the international activities discussed in Tables 6-1 and 6-4. Tables 6-7 and 6-8 complement Table 6—6 by presenting rankings by percentages of change and rankings by amOunt and direction of change, respectively, on these activities. Studying the three tables, we may arrive at several conclusions. First, there is appreciable change in the behavior of African states on all the variables from 1964 to 1967. Second, there is generally a higher degree of change evident on the two diplomatic variables (United Nations voting and IGO memberships) than on the two financial measures. Third, of the two diplomatic variables, one (IGO memberships) discriminates well between the coup and non-coup 195 Consumer Vulnerable F G Low I Upper Volta I Ali nment g I Dahomey I CAR I Protégé I Big I B I Vassal I Low I | Algeria I I I I Serra L. Independent )“ ‘— __._.__Bnrundi - / Togo / / 5‘ / . . '$ / Nigeria 5 / .U a /' Congo K. / Ghana h / Hi 9A D Calculating High Intensity Low Aggressive Figure 6-6.-—Typology of Coup Countries in 1967. .s-s-W‘; , v' ....._ ' "7 '7 Table 6-6.——Country Rankings by Amounts of Change Selected International Activities, 1964—67 (Coup Countries Underlined; Tied Scores Bracketed) te of Voting Against Total Number Exports/GNP All Aid/GNP USA--Rate Against USSR IGO Memberships Low Change 1 Niger -.033 Burundi 0 Ivory Coast -4 Guinea 0 2 Mali —.00__ C a —-.002 Da omey 37 Somalia 1 3 Dahomey —.033 Ethiopia .003 Zambia 51 Liberia -1_ 4 Algeria —.0 Sudan .004 Ethiopia 65 Morocco 5 Malawi -.O(fi Congo K. —.004 Kenya —72 Mali 3 6 Malagasy R. —.OO£I Cameroon .004 Algeria 90 Libya :I 7 Sudan -.01_0— Upper Volta .007 Niger 97 Ethiopia 3_ 8 Senegal .OLQI Nigeria .008 Malagasy R. 103 Sudan 5 9 Upper Volta .014 Dahomey -.009 Sierra Leone —106 Zambia ’il 10 Tunisia .016 er .Oll Liberia 110 Rwanda 7 11 Liberia —.017 na -.014 Malawi —120 Mauritania g] 12 Guinea —.018 Sierra L. —.016 Tanzania 134 Ghana 13 Mauritania .019 Morocco -.018 Mali 147 Tunisia 9 14 Congo Brazz -.O20 Mauritania — 018 Nigeria 149 ng . 9 15 Chad —.022 Ivory Coast -.018 Tunisia 160 Burundi 16 Kenya -.023 Malagasy R. —.019 Upper Volta 161 Malawi 10- 17 Morocco —.025 Lib a —.020 Morocco 191 abon 10 18 Ethiopia —.o32 -.o27 bon —203 To 0 11' 19 Nigeria -.033 Senegal —.029 Guinea 209 Sierra Leone ll 20 Rwanda .035 Tunisia .030 Somalia 223 Senegal 11 21 Egg —.O4O CAR -.032 Chad 237 Malagasy R. 11 22 Cameroon -.049 _'I‘o_go —.041 Senegal 240 Dahomey ll 23 Tanzania -.054 Zambia .ofl Rwanda 245 9133 11 24 T090 -.066 Rwanda .043 Sudan 248 Niger 12 25 Ivory Coast -.O73 Uganda —.046 Mauritania 268 Congo Brazz l? 26 Somalia —.O79 Mali -.063 Congo I\. 28?} C ad 13 27 Gabon —.080 Malawi —.070 Burundi 280 Upper Volta 14 28 Burundi -.089 Liberia .077 g]: 290 Tanzania 14 29 Uganda -.106 Guinea —.083 Cameroon 291 ameroon 14 30 Congo K. .135 Kenya —.086 Libya 320 Uganda 16‘ 31 Sierra L. -.138 Tanzania -.090 Togo 321 Nigeria 16 32 Ghana -.l9l Algeria -.126 Uganda -335 Kenya 16 33 Libya -.222 So alia -.159 Congo Brazz 345 Al e 16 34 Zambia —.329 Congo Brazz -.215 Ghana —467 Ivory Coast 20 High Change U=33 U=l38 U=157.5 z = .945 z = 3.280 2 = .680 z = 1.420 p = .1724 p = .0006 p = .2483 p = .0778 groups on all three tables, whereas the other variable (United Nations voting) consistently does not discriminate significantly between the groups. The two financial variables show different patterns of change with respect to activities. For most of the countries, the rate of exports changes less than the rate of aid received, and shows a lower percentage change than any of the other variables. Over the period from 1964 to 1968, then, it appears that the African countries were slowly adjusting their rates of exports, generally decreasing those rates. Over this period, the rates of aid received also decreased, though these rates of aid dropped more rapidly than did the rates of exports. From these data alone, we might infer that over this period the African economies turned inward somewhat. They tended to receive less and less aid from abroad, and they tended to sell a greater percentage of their products and services at home. The countries which experienced successful military coups d'état during this period show a significantly greater drop in the rate of exports than do the countries with continuous civilian rule. The "coup" and “non-coup" groups show no significant difference in the drop in rates of aid received, although the "coup" group shows a very signifi— cant tendency to change much less in the amount, though not the rate, of aid received. Overall, African countries seem to thrust themselves more into international affairs diplomatically than economically in this four- year period of the mid—19605. Also, while these countries vary substantially in the amounts and rates of change on these variables, no country is completely stagnant. No country shows insignificant 198 change on all variables. Exact measures of association among these distributions have not been calculated, but it is clear that changing performance on one variable does not very fully predict performance on another. Some countries have very high rates of change on all variables (e.g., Rwanda), some have moderate to very low rates of change on all variables (e.g., Ethiopia), and others Show greatly different patterns: very high and very low rates of change. Algeria shows a very low change in its rate of exports, a very high change in the rate of aid received, a low change in the pattern of United £ Nations voting, and a relatively high change in IGO memberships. Zambia shows a very similar pattern, although its scores are more extreme, with the exception that it shows a high rate of change in exports. Liberia shows the same pattern as Algeria on the financial variables, but an opposite pattern on the diplomatic variables. Congo Kinshasa and, to a more limited extent, Burundi show patterns of high and low change directly opposite the pattern of Algeria. Looking back at Figure 6—2, which plots the countries with the most extreme combined scores on the parameters of intensity, extensity, and alignment, we see some interesting links to changing activities as displayed on Tables 6-6 through 6—8. Some of the most aggressive actors of 1964 had changed their performances dramatically by 1967, some had not, and some showed a mixture of change and non—change.) None of these aggressive actors showed a great change in the rate of aid received or the number of IGO memberships. Congo Kinshasa shows a very high change in its rate of exports and its pattern of voting in the United Nations, hOWever, and Ghana shows a similar pattern. These two countries are different completely in the direction of these changes, 199 however. The Congo shows a great increase in the rate of exports and a great increase in UN voting more favorable to the Soviet Union than to the United States, whereas Ghana shows a great decrease in the rate of exports and a great increase in voting more favorable to the United States. The other aggressive actors show much more moderate patterns of change. Both of the very small former Belgian colonies, Rwanda and Burundi, show very dramatic changes in activities, perhaps in an effort to ameliorate their vulnerable and vassal statuses, respec- tively. Rwanda shows a very high rate of change on all the variables in Table 6—7; indeed, its scores on Exports/GNP and United Nations voting are so high as not to be comparable with the other states. In 1964, it was the least aggressive and most vulnerable of the African states. That is, Rwanda showed the least intense, least extensive, and least aligned pattern of foreign—affairs behavior at that time. Its high scores on Table 6—7 can partly be attributed to its low starting— point in 1964. Nonetheless, these scores show that by the end of 1967 Rwanda was much more intensely involved in foreign affairs, with a vast increase in its rate of exports, a great increase in its memberships in IGO's, and a fantasticlly more aligned pattern of UN voting (aligned with the Soviet Union). It did suffer a great drop in the rate of aid received, though. Though not a powerful country in 1964, by any means, Burundi had a higher rate of exports than Rwanda and also received a higher rate of aid. Thus, it was more intensively and extensively involved in international financial affairs. Diplomatically, like Rwanda, it had virtually no memberships in IGO's, but, unlike Rwanda, Burundi 200 Table 6—7.—~Country Rankings by Percentage of Change Selected International Activities, 1964—67 (Coup Countries Underlined, low % Change at Top) Rate Voting Against Total Number Exports/GNP All Aid/GNP USA—‘Rate Against USSR IGO Memberships 1 Algeria -2.26 Burundi 0 Ivory Coast —4.49 Guinea 0 2 Liberia —2 59 Chad —2.56 Zambia 12.94 Morocco 12.0 3 Niger —3.66 Congo K. —5.88 Kenya -16.00 Liberia ~14.3 4 MalaWi -4.00 Dahomey —8.33 Algeria 19.78 bya 33.3 5 Mauritania 4.47 Cameroon 8.69 Dahomey 20.55 Somalia 33.3 6 Senegal 5.81 Upper Volta 10.00 Ethiopia 24.16 Sudan 41.7 7 Malagasy R. -6.00 Mauritania —16.22 Mali 31.34 ali 42.9 8 Congo Braz -6.45 Ethiopia 18.75 Tanzania 39.30 TuniSia 47.4 9 Sudan —6.76 iger 19.30 Morocco 53.20 Ghana 72.7 10 Dahomey --7.69 Sudan 21.05 Tunisia 56.74 Ethiopia 100.0 11 G bo —11.66 93 -22.86 Niger 63.40 Co go 112.5 12 TuniSia 11.68 Malagasy —24.36 Sierra L. -68.83 Mauritania 114.3 1 Mali —12.00 Gabon -29.35 Guinea 71.82 Malagasy 157.1 14 Guin —l3.23 —29.79 Nigeria 77.60 Senegal 157.1 15 Morocco —13.74 Senegal -31.87 Upper Volta 89.44 Gabon 166.6 16 Kenya —13.77 Malawi ~33.33 Burundi 96.22 Cameroon 175.0 17 Chad —17.74 Ivory C. -36.00 Malagasy R. 100.00 Nigeria 177.8 18 Tanzania —17.82 Togo —36.94 Malawi —100.00 Niger 200.0 19 Nigeria -l9.30 Tunisia 38.96 Somalia 102.29 _R 220.0 20 Ivory C. -l9.36 Morocco -40.91 Sudan 107.83 Dahomey 220.0 21 _A_R —20.51 Sierra L. -44.44 Uganda —109.12 Tanzania 233.3 22 Cameroon —23.67 Nigeria 53.33 Chad 123.44 Chad 260.0 23 Upper V. 25.45 Mali -57.27 Ghana -139.82 Sierra L. 275.0 24 Libya -25.99 Somalia -58.24 Mauritania 151.41 Upper V. 280.0 25 Togo —29.73 Uganda —63.01 Senegal 172.66 Algeria 320.0 26 Uganda -30.37 Congo B. —63.61 Libya 177.77 Ivory C. 400.0 27 Ethiopia —32.99 Kenya —66.15 To 0 193.37 Congo 8. 433.3 28 Somalia -35.27 Tanzania —68.70 Congo B. 194.91 enya 533.3 29 Burundi -35.60 Rwanda 70.49 Congo K. 243.48 Uganda 533.3 30 Zambia -37.26 Guinea —72.81 CB 250.00 Togo 550.0 31 Sierra L. —42.86 Algeria —78.75 Gabon —263.64 Rwanda 700.0 32 Congo K. 60.81 Liberia 88.50 Liberia 282.05 Zambia 700.0 33 Ghana -98.96 Libya -90.91 Cameroon 323.33 Burundi 900.0 34 Rwanda 3500.00 Zambia 126.47 Rwanda 1884.61 Malawi 1000.0 High % Change at Bottom U=71 U=l64 U=120 U=84 z = -1.85 z = .166 z = 0 z = —1.36 p = .0322 p = .4341 p = .5000 p = .0869 . . ... ..._ 201 was quite highly aligned with the Soviet Union in United Nations voting. That alignment increased over the four years studied here, so that by the end of 1967 Rwanda and Burundi were voting in a more similar pattern in the UN. In 1964, Ethiopia was classified as the most calculating of the African actors, and that classification may be verified by the data in Table 6-7, which show that Ethiopia had the overall pattern of least change on the selected international activities. By 1967, Guinea had become the most calculating of the actors, achieving that classification through a moderate drop in its rate of exports, a substantial drop in the rate of aid received (though having diversified its sources of aid), and through appreciably increased alignment with the Soviet Union in United Nations voting. More general conclusions can be drawn from Tables 6-6 through 6-8. First, some of the variables disciminate very well between the coup and non-coup groups when measuring amount or percentage of change; but taking the direction of change into account, as in Table 6-8, largely destroys that discriminatory power. The exception to this observation is the measure of amount and direction of change in IGO memberships; but this exception may be seen as an artifact of the fact that the distribution of this variable on Tables 6—6 and 6-8 are almost ' identical, reflecting the fact that only one African country decreased its IGO memberships OVer the four—year perid. Thus, in general, amount of change is a much better discriminator between the two groups than its direction of change. Second, looking at Table 6—7, we see that two variables discriminate very well between the coup and non—coup groups: 202 Table 6-8.—-Country Rankings by Amounts and Direction of Change Selected International Activities, 1964-67 (Coup Countries Underlined, High Positive Change at Top) Rate Voting Against Total Number Exports/GNP All Aid/GNP USA--Rate Against USSR IGO Memberships 1 Congo K. .135 Liberia .077 Congo Brazz 345 Ivory Coast 20 2 Rwanda .035 Rwand .043 Togo 321 Algeria 16 3 Mauritania .019 Zambia .043 Libya 320 Kenya 16 4 Tunisia .016 TuniSia .030 Cameroon 291 Nigeria 16 5 Upper Volta .014 Niger .011 _C_A_13 290 Uganda 16 6 Senegal .010 Nigeria .008 Burundi 280 Cameroon l4 7 Niger -.003 Upper Volta .007 Congo K. 280 Tanzania 14 8 Mali —.006 Cameroon .004 Mauritania 268 Upper Volta 14 9 Dahomey —.006 Su an .004 Sudan 248 Chad 13 10 Algeria 006 Ethiopia .003 Rwanda 245 Congo Brazz 13 ll Malawi -.009 Burundi 0 Senegal 240 Niger 12 12 Malagasy -.009 Chad -.002 Chad 237 C13 11 13 Sudan -.010 Congo K. -.004 Somalia 223 Dahomey ll 14 Liberia -.Ol7 Dahomey —.009 Guinea 209 Malagasy R. 11 15 Guinea -.018 Ghana —.014 Morocco 191 Senegal ll 16 Congo Brazz -.020 Sierra Leone —.016 Upper Volta 161 Sierra Leone 11 17 Chad -.022 rocco —.018 Tunisia 160 Togo 11 18 Kenya —.023 Mauritania —.018 Nigeria 149 Gabon 10 19 Morocco —-.025 Ivory Coast -—.018 Mali 147 Malawi 10 20 Ethiopia —.O32 Malagasy R. --.019 Tanzania 134 Burundi 9 21 Nigeria -.033 Libya —.020 Liberia 110 Congo K. 9 22 93 -.040 -.027 Malagasy R. 103 Tunisia 9 23 Cameroon -.o49 Senegal —.029 Ni er 97 Ghana 8 24 Tanzania -.054 EA}: -.032 Algeria 9O Mauritania 8 25 o -.066 Tog —.041 Ethiopia 65 Rwanda 7 26 Ivory Coast -.073 Uganda -.046 Zambia 51 Zambia 7 27 Somalia -.079 Mali -.063 Dahome 37 Sudan 5 28 Gabon —.080 Malawi —.070 Ivory Coast —4 Ethiopia 3 29 Burundi —.089 Guinea -.083 enya —72 ibya 3 30 Uganda -.106 Kenya —.08 Sierra Leone —106 Mali 3 31 Sierra L. —.138 Tanzania —.090 Malawi —-120 Morocco 3 32 Ghana —.19l Algeria —.126 Gabon —203 Somalia 1 33 Libya —.222 Somalia -.159 Uganda —335 Guinea 0 34 Zambia —.329 Congo Brazz -.215 Ghana —467 Liberia -1 high Negative Change at Bottom U=109 U=132 U=126 U=159 z = -.4159 2 = .4537 z = .2268 z = 1.4745 p = .3387 p = .3251 p = .4102 p = .0702 203 Exports/GNP and Total Number of IGO Memberships. The other two vari— ables on this table show no such discriminatory power. It appears, however, that those countries which show the greatest percentage of change on all the variables of the table have one of two characteristics which may predict that change. Each of the six countries having the highest percentage of change on the four international activities either experienced a successful military coup d'état, or was found among the most extreme rankings on the variable in question in 1964, or has both characteristics. In general, then, we may infer one of two explanations for these countries' having undergone such extreme rates of change on the variables: Either the new military regimes significantly reoriented policy in the area, or there was a reaction to the extreme (in relation to the rest of Africa) pattern of activity that took place in 1964. Actually, the explanations are not quite that simple, even considering the data on Table 6—7 alone. It is true that all the countries showing the highest percentage of change on all the variables either experienced successful military coups or had extreme rankings in 1964. Also, those countries which show a very high rate of change on a variable but did not experience military coups nearly all show change which goes a long way to moderate their extreme rankings of 1964. But in two cases, countries which ranked at an extreme in 1964 actually had magnified the extremity of their positions greatly by the end of 1967. The most dramatic example of a country showing this phenomenon of "reaction against extremes" is Rwanda. In 1964, it showed the lowast rate of exports and very nearly the fewest IGO memberships. By 1967, it showed the highest rate of increase in exports/GNP and 204 nearly the highest percentage gain in IGO memberships. Zambia showed the same pattern for its IGO memberships, but became less extreme with regard to its rate of exports. In 1964, Zambia had the highest rate of exports of any African state, but by the end of 1967 it showed one of the highest percentages of decrease in exports. Figure 6—7 classifies each of the six countries showing the highest percentage of change on each of the international activities according to whether it experienced a successful military coup or had an extreme ranking on the variable in 1964. There are three exceptions to the patterns among the 24 classifications. In two of these, countries did not react against their extreme rankings of 1964, but actually magnified the extremity of their positions. In 1964, Libya received one of the very lowest rates of aid of the African countries, but by the end of 1967 had experienced the greatest decrease in the rate of aid received of any country on the continent. In 1964, Gabon was one of the least—aligned countries in terms of voting in the United Nations. This meant that it was one of the four states voting most often with the United States rather than with the Soviet Union. The data show that by the end of 1967 Gabon had bucked the very sub— stantial continental trend in UN voting which led most African countries to align more with the Soviet Union, and Gabon had become one of two African countries which voted more often with the United States. Thus, its "extreme" position in UN voting was magnified. The last exception fits least well the classification in the table. Liberia shows the third highest percentage of change in rate of aid received, yet it neither experienced military intervention nor was ranked at an extreme in rate of aid received in 1964. Figure 6-7.--Classification of Countries Showing Highest % of Change Selected International Activities, l964~67 Exports/GNP All Aid/GNP Rwanda Zambia (reaction against extreme) (reaction against extreme) Ghana Libya (military coup) (extreme magnified) Congo K. Liberia (military coup) (?) Sierra L. Algeria (military coup) (military coup) Zambia Guinea (reaction against extreme) (reaction against extreme) Burundi Rwanda (military coup) (reaCtion against extreme) United Nations Voting IGO Memberships Rwanda (reaction Cameroon (reaction Liberia (reaction Gabon against extreme) against extreme) against extreme) (extreme magnified CAR (military Congo K. (military coup) coup) Malawi (reaction against extreme) Burundi (military coup) Zambia (reaction against extreme) Rwanda (reaction against extreme Togo (military coup) Uganda (reaction against extreme) Table 6-9 presents another way of interpreting the continental changes in international activities. Comparing the correlations dis— played on Tables 6—2 and S-S, this table lists the correlations showing the largest changes over the period 1964 through 1967, ranked by the amounts of change. Looking down that ranking, we see (1) that while the rate of exports was rather highly related to per—capita GNP in 1964, by 1967 that relationship had dropped to insignificance. (2) Total amount of aid received is highly related to GNP in both years, but Total Aid/GNP moves from very low negative to rather high negative correlation with GNP. If we consider the variable All Aid/GNP as rate of aid, then rate of aid becomes, over the four-year period, inversely related, at a substantial level, to GNP. Thus, although a country with a large GNP gets more aid in absolute terms than a country with small GNP, by 1967 the small-GNP states tend to get more aid per unit of GNP. (3) As a corollary to the previous change, we see that the rate of aid in 1964 is moderately correlated to total aid, but by 1967 becomes somewhat negatively related to total aid; the direction of the relationship has become reversed. (4) In 1964, the higher the rate of exports, the less tendency to vote more against the United States in the United Nations than against the Soviet Union. By 1967, that relationship has been reversed. (5) In 1964, the more inter— governmental organization memberships, the more tendency to vote against the United States than against the Soviet Union, but by 1967 that relationship has also been erased. On an aggregate basis, then, Africa showed these general trends in international activity for the period 1964 through 1967: The larger economies consistently carried on greater trade and received more aid 207 Table 6—9.—-Ranking of Changes in Correlations Rank Variables 1964 1967 Change 1 Exports/GNP X GNP/capita .623 —.065 ‘ 688 2 A11 Aid/GNP X GNP —.114 -.567 453 3 All Aid/GNP X All Aid .282 —.139 421 4 Exports/GNP x Voting against USA minus against USSR -.283 .134 417 5 IGO Memberships x Voting: USA—USSR .434 .039 395 6 Imports-Exports X All Aid/GNP —.151 -.533 382 7 All Aid X Voting: USA-USSR .457 .143 314 8 IGO Memberships X GNP/capita .337 .033 304 than did the smaller. However, the smaller, poorer countries began to do better in spite of their few economic resources; that is, per unit of GNP, they attained greater rates of trade and aid. Over the whole continent, though, the average rates of trade and aid decreased signifi- cantly by 1967, no doubt reflecting, among other things, the substantial fall—off in aid from the superpowers. In other words, the period shows a weakening of all foreign aid to Africa. In addition, we see a tendency for two different groups of states to emerge, distinguished by the modes of their primary inter- national activities. We can call them the Exporters and the IGO Diplomats. These groups do not coincide with the coup and non—coup groups. The Exporters are evident in both 1964 and 1967 as countries with very high rates of exports but ranked quite low in IGO member- ships. These Exporters included Zambia, Gabon, Liberia, and Mauritania. 208 The IGO Diplomats, on the other hand, are states which show many IGO memberships but low export rates. These IGO Diplomats included the Maghreb countries, the UAR, Nigeria, and Ghana. In 1964, Exporters tended to align themselves with the United States in United Nations voting, while IGO Diplomats tended to vote much more heavily against the United States and with the Soviet Union. By 1967, these relationships had changed only marginally. Of the countries showing change, Ghana became a voter much more favorable to the United States, and Libya and Mauritania became much less favorable. Also, by 1967, a third group had emerged which showed high activity both in rates of exports and in IGO memberships. These states included Algeria, Tanzania, the Ivory Coast, and Uganda. In this group, voting patterns in the United Nations include both extremes: Algeria and Tanzania voted heavily against the United States and often with the Soviet Union, while, especially in 1967, the Ivory Coast and Uganda voted much more favorably with the United States. Changes in Patterns on Analytic Parameters, 1964—1967 Figures 6—8 through 6—10 present rankings of 34 countries on changes in international behavior measured by the parameters of inten- sity, extensity, and alignment. Only one of the eight rankings in these tables discriminates between the coup and non—coup groups in a way that approaches significance. That is the ranking on the amounts of change in extensity. Looking at all these rankings, there appears to be no simple characteristic which can be used as an explanation of the distribution. Unlike the specific changes in performance on variables reported in 209 Figure 6-8.--Country Ranking by Amounts of Change on Parameters, 1964 to 1967 (Coup Countries Underlined; Tied Scores Bracketed) Intensity Extensity Alignment Low 1 Uganda 0 Congo Kinshasa 1.5 Upper Volta 0.5 Change 2 Chad 0.5 Ivory Coast 255 3 Mauritania Liberia Congo Kinshasa 4 Senegal 1.0 Libya 2.5 Guinea 1.0 5 Tanzania Mauritania Liberia 6 Sierra Leone 2.0 Nigeria 4.5 Algeria 7 Ivory Coast 4.0 Senegal 6.0 Malagasy Rep. 2.0 8 Congo Kinshasa 4.5 -Eabon 7.0 Malawi 2.5 9 Malagasy Rep. 5.5 Morocco Somalia 5.0 10 Burundi 8.0 Somalia 7.5 Mauritania 6.5 11 Morocco 9.5 Ethiopia 8.0 Cameroon 7.0 12 Liberia 10.0 Sudan Sengal 8.0 13 gpnisia Cameroon 9.0 Kenya 8.5 14 Sudan 10.5 Uganda 9.5 Dahomey 9.0 15 Ethiopia 11.0 —Kenya 10.5 Mali 9.5 16 Dahomey 11.5 Malawi Niger 10.5 17 Gabon 12.0 Algeria 11.5 Rwanda 11.0 18 Malawi 14.5 Tunisia 12.0 Sierra Leone 12.0 19 Nigeria Congo Brazzaville 17.0 Tunisia 20 Eggg' 15.0 Malagasy Rep. 17.5 Ethiopia 12.5 21 955 16.0 -Euinea 18.0 Sudan 13.0 22 Cameroon 16.5 LSierra Leone Morocco 15.0 23 Libya 24.0 _Bahomey Burundi 15.5 24 Rwanda 26.5 Niger 22.0 Libya 16.0 25 Congo Brazzaville 27.0 éggyz Chad 16.5 26 Somalia 27.5 Chad 23.00 Gabon 2? Upper Volta 29. 5 SEES 23.0 3319 18.5 28 Niger 36.0 -Burundi Tanzania 19.0 29 Algeria 47.0 LLanzania 27.5 Nigeria 24.0 30 Guinea 48.5 Mali 29.0 Congo Brazzaville 24.5 31 £2233 52.0 Upper Volta 30.5 Uganda 25.0 32 Mali 58.0 Zambia 34.0 Ivory Coast 29.5 33 Zambia 59.0 953 41.5 Ehggg 30.0 High 34 Kenya 65.5 Rwanda 63.0 Zambia 41.5 Change U = 108 U = 82 U = 136 z = -.4537 2 = -l.4367 z = .6049 p = .3251 p = .0754 p = .2726 210 Figure 6—9.--Country Ranking by Amount and Direction of Change on Parameters, 1964 to 1967 (Coup Countries Underlined; Tied Scores Bracketed) Intensity Extensity Alignment Less 1 Mali CAR EEEEE 2 92223 Upper Volta Ivory Coast 3 Guinea Mali Uganda 4 Algeria gpgpg Nigeria . 5 Somalia Chad Tanzania % 4 6 Congo Brazzaville Niger Gabon i 7 Libya Tpgg Libya 8 CAR Congo Brazza Morocco 9 Tpgg Algeria Ethiopia 10 Gabon Cameroon Sierra Leone 11 Malagasy Rep. Somalia Dahomey 12 Ivory Coast Gabon Senegal 13 Sierra Leone orocco Cameroon 14 Senegal Senegal Somalia 15 Uganda Ivory Coast ' fAlgeria 16 Chad Liberia lMalagasy Rep. 17 Mauritania Congo Kinshasa Guinea 18 Tanzania .Libya Upper Volta 19 Congo Kinshasa Mauritania EAR 20 Burundi Nigeria Congo Kinshasa 21 Morocco Ethiopia Liberia 22 -Liberia Sudan Malawi 23 _Tynisia Uganda Mauritania Figure 6—9.--Continued 211 Intensity Extensity Alignment 24 Sudan Kenya Kenya 25 Ethiopia Malawi Mali 26 Dahomey Tunisia Niger 27 “Malawi Malagasy Rep. Rwanda 28 Nigeria Tiuinea Tunisia .E:._____ 29 Cameroon Sierra Leone Sudan 3O Rwanda Dahomey Burundi 31 Upper Volta Burundi Chad 32 Niger Tanzania Tpgp 33 Zambia Zambia Congo Brazza More 34 Kenya Rwanda Zambia U = 135.5 U = 145.5 U = 133.5 2 = .5860 z = .9641 z = .5104 p = .2790 p = .1675 p = .3049 212 Figure 6-10.--Country Rankings Aggregated Across Three Analytic Parameters (Coup Countries Underlined) Ranks by Amount of Change on Parameters, 1964;1967 Ranks by Amount and Direction of Change on Parameters, 1964-1967 1 Congo Kinshasa 92322 2 Mauritania Algeria 3 Liberia Gabon 4 Senegal EAR 5 Malagasy Republic Mali 6 Morocco Ivory Coast 7 Malawi Somalia 8 Ivory Coast Libya 9 Somalia Senegal 10 Cameroon Uganda 11 Sierra Leone Morocco 12 Ethiopia Congo Brazzaville 13 Sudan Tpgg l4 Uganda Guinea 15 Tunisia Upper Volta 16 Libya Nigeria 17 Gabon Sierra Leone 18 Algeria Cameroon 19 Nigeria Chad 20 Chad Malagasy Republic 21 Dahomey Tanzania 22 Guinea Ethiopia Figure 6-10.——Continued 213 Ranks by Amount of Change on Parameters, 1964—1967 Ranks by Amount and Direction of Change on Parameters, 1964-1967 23 CAR 24 Upper Volta 25 Tanzania 26 Burundi 27 Kenya 28 Niger 29 Togo 30 Congo Brazzaville 31 Rwanda 32 Mali 33 Ghana 34 Zambia U = 113 z = -.2646 p = .3956 Congo Kinshasa Liberia Mauritania Niger Dahomey Malawi Sudan Tunisia Burundi Kenya Rwanda Zambia U 146 z = .9830 .1628 "C II 214 Table 6-7, which could be associated, at least in part, with the pre- sence or absence of military coups and/or with extreme highs or lows of activity in 1964, the rankings of changes on the analytic parameters do not seem to relate to military coups or to extreme rankings on these parameters in 1964. Nor does any ranking appear to relate to country- size or economic development. States with large and small population, GNP, and GNP/capita appear at both ends of each of the distributions. Remembering that each of these parameters ranks countries according to composite scores on half a dozen variables of trade, aid, and diplomacy, or more, it appears that the complexity of the parameters makes the identification of clusters of states much more difficult. Thus, we turn to the analysis of parameters within each of the three issue-areas of international relations: trade, aid, and diplomacy. Mann-Whitney U Tests of All Trend—variables North and Tropical African Countries Tables 6-10 and 6-11 present the results of analyzing the impact of military coups on foreign-policy patterns in the African countries. Data were collected on 46 variables of international rela— tions, representing trade, aid, and diplomacy. For the group of ten countries which underwent successful military coups d'état during the period from early 1965 to early 1967, the trends in foreign affairs behavior were calculated, measuring the change in each variable from one year before the coups d'état to one year after. These average changes for the "coup group” were compared with the average changes among the non—coup group during the same period. Mann—Whitney U tests 215 were computed to determine the probable significance of difference between the coup and non—coup groups' scores. It should be noted that the term "test" is used loosely here. On the one hand, the sample of countries so nearly approximates the size of the full population under study that the probabilities computed for the variables become statements about the occurrence of the distribu- tions by random chance, not inferences about the population. On the other hand, my theory of military orientations is not so precise that it allows definite predictions of change on all my variables. I have therefore freely manipulated the variables, sometimes using the Mann— Whitney U computations as one-tailed tests and sometimes as two—tailed tests, in search of relationships. These results are thus information which allow us to refine our images of the military and foreign rela— tions, but they do not constitute formal tests. Such formal tests would not allow this exploratory manipulation of data, but would require precise predictions before the data were gathered. Thus, for— mal tests of significance must be the task of other studies, and I believe my results would greatly aid in formulating subsequent formal testing, although I cannot claim to have achieved such test here. Percentages of change, plus or minus, were calculated for each group on each variable. Countries were then ranked according to these percentages, so that Mann-Whitney U tests could be calculated. Para— metic tests were disallowed because distribution assumptions could not be met. The probabilities given are those associated with the U values obtained. The group sizes are large enough for transformation of the U values into 2 values which are practically normally distributed, according to the formula given by Siegel in Non-Parametric Statistics. 216 Thus, the probabilities are those associated with values of z normally distributed, and the lowest values are those most significant. The highest possible probability is .5000, which indicates absolutely no difference between the distributions of the coup and non-coup groups. In the cases of a few variables—-numbers l4, 6, 23, 19, 16, and 22-—a curvilinear relationship was clearly evident, such that one group of countries appeared at both the high and low extremes of percentage change and the other groups appeared in the middle. In these cases, the countries were reranked according to absolute amount of change—- plusses and minusses were dropped——and Mann—Whitney U's re—calculated. In each of these cases, this curvilinear test proved much greater significance for the variable than the linear test, so probabilities for the curvilinear tests are given and the means are described as high or low change. Table 6-10 shows at least ten variables that I would consider significant. It also indicates that some variables differentiate between coup and non—coup groups virtually no better than change. Six of the ten most significant variables are measures of trade, only one is of aid, and three are of diplomacy. (USA "educational and cultural" H expenditures are considered "diplomacy, since the largest portion is used to send American entertainers and others for "person~to—person" diplomacy, although a good sum sponsors Fulbright-Hays lecturers and researchers for which the African states presumably realize direct benefit.) Three of the variables displayed in Tables 6—1 and 6-4 are among the least significant discriminators between the coup and non—coup groups. Thus, in rates of exports, rates of IGO memberships, and rates of aid received, the military regimes are no different from 217 Table 6—lO.——Discrimination Between Coup and Non—Coup Countries: Trend Data Ranked by Probabilities From Mann-Whitney Scores Prob- Non—Coup ability No. Variable Description Coup Mean Mean .0039 18 Exports minus imports/GNP —4l.l7 -23. 68 .0143 14 % imports from other African countries low change high .0196 6 Number countries from which receive imports low change high .0222 28 USA educational & cultural expenses 47.21 —l6.19 .0262 3 Total imports —3.39 16.23 .0336 39 Total Multilateral aid -23.62 62.99 .0427 13 Second highest % exports to any one country 15.63 —4.41 .0465 5 Imports plus exports —0.61 14.88 .0582 23 Rate of voting against African majority in UN High low change .0582 35 Number African members in common 00' 171.70 18 . .0606 40 Total communist bloc aid 0.03 27.69 .0793 27 Number scholars & artists brought to USA by USA 33.59 —l9.42 .0885 l Imports/GNP —9.83 —1.83 .1075 26 Rate voting against USA minus rate against USSR 46.76 43.39 .1131 10 Highest % of imports from any one countr ry —7.78 6.57 .1151 33 Total number members in common 100' s 191.32 149.42 .1210 25 Rate voting against USSR in United Nations 14.8 25.78 .1251 19 Number USA diplomats assigned resident low change high .1379 12 Second highest % imports from any one country 33.55 7.24 .1711 16 t of imports from communist countries low change high .1711 29 Number countries to which diplomats are sent 3.96 9.23 .1711 41 Largest amount aid from any one country 5.48 —5.46 .1894 34 Number members in common IGO's/number common IGO's 25.47 57.87 .1920 45 Exports to the United States 0.15 35.02 .2087 48 USA aid minus communist bloc aid —67.97 -6.27 .2206 30 Number countries from which diplomats received 10.21 16.70 .2327 44 Imports from the United Sta es —5.23 8.37 .2451 15 % of exports to all "Other Africa" 16.68 11.47 .2451 38 Total official United States aid 88.36 36.75 .2530 4 Total exports 2.32 49.88 .2546 37 Total official OECD aid 12.53 13 37 .2912 42 Grand total all aid -2.06 34.28 .3050 36 Number African states in common IGO's/Number IGO's 18.62 29,95 .3526 47 % of exports to USA minus % to communist b 0c .3 —15.94 .3840 22 Participation rate in UN roll—call voting low change high .4270 7 Number countries to which exports are sent 1.90 7.56 .4330 32 Number "African only" IGO's/Number all IGO's —15.36 —1.62 .4562 24 Rate of voting against USA in United Nations 14.00 13.58 , .4590 11 Highest % of exports to any one country _ 1.96 1.88 .4748 2 Exports/GNP ‘ —l4.81 35.66 .4801 17 % exports to communist countries 15.83 36.11 .4860 31 Total number of memberships in IGO 3 90.48 95.62 .4860 43 Grand total all aid/Gm —9.85 12.81 —91.93 320.29 .5000 46 % of imports from USA minus % from communist bloc 218 overall continental trends. In terms of voting—alignments in the United Nations, however, the military regimes show a tendency to vote more heavily against the United States than do the civilian regimes. The group means given in the Tables are somewhat misleading, since they mask the extremity of the distribution of scores: in both coup and non-coup groups there are often extreme scores which exagger— ate the means. Nonetheless, the means accurately show the directions of changes——whether the group is moving up or down on the measures—— and the more significant the variable, the more reliable the mean group scores. Taking this into consideration, we see these trends: In imports and total trade (imports plus exports), the military regimes show a slight drop, while the civilian regimes show substantial increase. Most significant of all, however, is the measure of balance of trade per unit of GNP (exports minus imports/GNP). Here, both groups show decreasing means--they show more favorable balances of trade per size of economy—~but the military regimes are apparently much more vigorously overseeing the rebalancing of trade. At the same time, the significant variables 14, 6, and 13 show that the military regimes show stability in the numbers of other states traded with and a sub— stantial tendency to increase exports to their second—largest trading partners. The overall trade picture, then, for the coup countries, is of somewhat diminished but better-balanced trade, spread somewhat more evenly among a stable number of countries. In aid relationships, only one variable shows clearly signifi— cant difference between the two groups. It indicates that, after coups, the military regimes lose substantial amounts of aid from multilateral sources, that is from the World Bank, the various United 219 Nations sources, and other international governmental organizations. This variable is especially significant because, at the same time, the civilian regimes are greatly increasing their receipt of aid from these sources. The coup group shows nearly a 24% decrease, while the non—coup group shows about 63% increase, on the average, making an absolute difference of 87%. Almost significant is the measure of communist— bloc aid, which shows virtually no communist aid to the military regimes and a substantial increase in communist—bloc aid to civilian governments. The other measures of aid are much less useful discri— ‘ minators, having probabilities of .1711, .2087, .2451, .2912, and .4860. If the group means for these variables have any value, however, it would seem that the military regimes received their aid from fewer sources, especially more from the United States, while the civilian governments are decreasing their dependence on one wealthy country. Overall, there seems to be little significant difference in the amounts of aid, but the patterns of dependence are different in the two groups. Measures of diplomatic relations show a very strong rise in United States interest in the countries which undergo successful coups d'état, along with a strong decrease in interest in the group of non— coup countries. Variable 28 shows more than a 47% increase in United States educational and cultural expenses in the coup group, but over a 16% decrease in the non-coup group, relative to 1964 levels. The nearly significant variable 27 is similar in pattern, showing over a 33% rise in the numbers of scholars and artists brought to the United States on State Department grants from the coup group, but over 19% reduction in numbers from the non-coup group. 220 The interest of the United States in the new military regimes is apparently not balanced by a commensurate alignment of these regimes with the USA in United Nations voting. Although they show significance levels which are at best only suggestive, it appears that both coup and non-coup groups show markedly increasing rates of voting against both the United States and the Soviet Union, while the non—coup group shows a greater tendency to vote against the Soviet Union, and so, overall, the military regimes turn out to be somewhat more antagonistic toward the United States than the continuous civilian regimes. Both groups are remarkably independent of big power politics in the United Nations, however. In 1966 and 1967, it was indeed rarely that the United States voted with the African majority. At the same time, while the Soviets joined the Africans in voting about issues of colonialism, this affinity was more than counter-balanced by disagreements on other issues. In relations with the rest of Africa, the military regimes showed little change in trade relations but significant difference in diplomacy. Variable 35 shows that the coup group tended to join inter— national governmental organizations less populated by African states than did the non—coup group. This variable and others, including numbers 33, 34, and 31, show that both groups very greatly increased the number of international organizations they joined, and thus vastly increased their exposure and presence in world organizations, but still there was a tendency for the coup group to be more independent of their regional colleagues in this process. That independence is also demon— strated by the significant variable 23, which shows that relative to 1964 levels, the military regimes show a tendency to vote much more 221 often against the african majority in the United Nations than the non— coup group. As Table 6—12 shows, not every military regime showed an increase in voting against the African majority, but those states which did increase their independence from continental voting patters did so at an overwhelming rate. Trgpical African Countries Alone Table 6—11 presents information similar to Table 6—10, but only for the reduced sample we labelled Tropical African countries. Algeria, Libya, Morocco and Tunisia have been removed from consideration in rank— ing the countries for Mann-Whitney U tests here. Only one of these states (Algeria) is a member of the coup group, and so we have nine coup states compared with 21 non-coup states. The distribution of probabilities for the variables in Table 6—11 is very similar to that in Table 6—10, as might be expected. Table 6‘13 ranks the variables by amount of change in probability from Table 6—10 to Table 6—11. About half of the variables show higher probabilities in analy— sis of the reduced sample of countries, and half show lower probabili— ties. Of the ten significant variables on Table 6-10, six remain significant on Table 6-11, including the five most significant of Table 6-10. The other four drop in significance, although in no case does the drop put a variable at the other extreme of the rankings. Among the variables that drop below probability of .06, and also those variables that reach that probability or better in Table 6—11, one can find measures of trade, aid, and diplomacy. Evidently, removing the North African countries does not differentially affect the rank- ings of the substantive areas at the most significant levels. 222 Table 6-11. ——Discrimination Between Coup and Non-Coup Countries of Tropical Africa: Trend—Data Ranked by Probabilities From Mann—Whitney U Sco PtO— Non—Coup ability No. Variable Description Coup Mean Mean .0069 14 % imports from other African countries low change high .0113 18 Exports minus imports/GNP —76.87 —27.27 .0119 6 Number countries from which receive imports low change high .0250 28 USA educational & cultural expenses 53.53 —14.60 .0351 19 Number USA diplomats assigned resident low change high .0375 3 Total import -3.38 17.78 .0562 40 Total communist bloc aid 0.00 15.53 .0571 27 Numbe er scholars & artists brought to USA by USA 42.32 ~20.45 .0630 46 % of imports fro m USA mi nus % from communist bloc -79.81 361.84 .0708 14 % imports from other African countries 5.44 ~9.31 .0778 5 Imports plus exports —0.73 15.11 .0918 1 Imports/GNP 9.49 0.11 .0918 23 Rate of voting against African majority in UN high low change .0985 29 Number countries to which diplomats are sent 1.63 10.03 .1075 10 Highest % of imports from any one country -8.70 7.28 .1075 16 A of imports from communist countries low change high .1170 39 Total multilateral aid —11.95 68.87 .1251 25 Rate of voting against USSR in United Nations 22.98 32.19 .1251 26 Rate voting against USA minus rate against USSR 45.37 41.88 .1446 44 Imports from the United Stat as -7.64 20.06 .1446 13 Second highest % exports to any one scountry 9.49 -1.48 .1660 31 Total number of memberships in IGO' 78.32 109.98 .1788 35 Number African members in common IGO' s 166.82 ' 202.77 .1788 23 Rate of voting against African majority in UN 103.19 58.89 .1788 34 Number member in commo IGO' s/ number common 160‘ 5 27.08 66.02 .1894 41 Largest amount of aid from any one country 11.26 2.77 -1327 48 USA aid minus Communist bloc aid —74.11 26.60 .2033 12 Second highest % imports from any one country 16.81 5.78 .2643 38 Total official United States aid 103.35 52.15 .2709 37 Total official OECD aid 19.10 23.53 .2877 4 Total exports 2.08 55.71 .3264 15 % of exports to all "other Africa" 20.57 13.95 .3264 17 % of exports to Communist countries 3.92 45.35 .3594 32 Number "African only" IGO's/Number all IGO's —9.57 0.31 .3660 47 t of exports to USA minus % to Communist bloc 30.73 —12.50 .3707 33 Total number of members in common IGO's 18.67 170.85 .3707 42 Grand total all aid 5.82 46.72 .4052 43 Grand total all aid/GNP -2.55 22.32 .4207 30 Number countries from which diplomats received 12.93 10.83 .4562 11 Highest % of exports to any one country 9.49 —1.48 ' .4621 22 Participation rate in UN roll-call voting -0.72 —5.22 .4761 36 Number African states in common 1% 's/Number 100' 5 20.15 28.28 .4920 7 Number countries to which exports are sent 4.17 6.80 .4920 24 Reta of voting against USA in United Nations 14.19 13.33 .4920 45 Exports to the United States 4.48 72.63 mam uoz noauflm mo xopcw on» umcu mmocoo we Scum oumwumon . muwamcu mflum> can no usages: one .muoz . a Son o o Hm>< 002 so now» dump E w an H woowocfl won as com: moan WQHMOHWGH .floz ommwmahufimhmm CH gogm 0H6 uuo«>mcom cmflouom mo mowmucsoo mm mowumm Hoowmoue com cuuoz In. I ousuam momme >n oomsono mmofiocH noose co mofluucsoo msoolcoz new moon cooSuom coaumcfleflnomflo Ha m . moms. u a Ammo pmms m whoa. n a nose m Ammo onH. n o ammo .55200 I .l 1 momo. n a Aomo moaned . . v vmmH. n m Amev .¢.z .¢.Z VBHmZmem m ommo. u o Avm+mm+ma+omo ommo. n a love .a.z meHmzmezH m oooo. u a omao. n m 1om+mm+smo Hmom. n a Aooo 1om+mmo some . ommo. n a .m .¢.z oooo. u m Ammo 1mm+oao s a pmflcsssoo U 1 oomm. n o v 1 .<.z .a.z 1sH+oHo m aonma W .ezmsonom q osmo. u m Hams. u m Ammo Amauos-o+o+m+oo oooa. n m 1om+mmc meHmzmexm owe. n a 1Hm+svo oooo. n o 1m+m+ao meHmzmazH NUdZOQmHQ DHfl MQ¢MB 224 Table 6-12.-—Rates of Change in UN Voting After Military Coups: Coup Countries voting Against African Majority, 1964 vs. 1967 Dahomey 266,66% Congo Kinshasa 53.09% Ghana 222.90% Burundi —12.03% Sierra Leone 195.16% Togo —33.74% CAR 151.61% Algeria —36.36% Nigeria 141.59% upper Volta -56.50% Among the variables which show the greatest changes in pro- babilities, we find more diplomatic variables than other types, although some measures of trade are also found. This probably reflects the fact that the North African countries are among the states with the highest levels of involvement in IGO's, so that removal of these states from the sample allows Tropical states to have greater effect on the rank- ings of IGO variables. Significance Tests of All Trend Indices (Parameters) North and Tropical African Countries Figure 6—11 presents the results of analyzing the impace of military intervention on the intensity, extensity, and alignment of foreign affairs of the African countries. The table also displays the results in terms of the substantive areas (trade, aid, and diplomacy), and in terms of bilateral or multilateral arrangements. This more detailed conceptualization of foreign relations allows a more precise identification of the areas in military influence was felt and those in which it was not. 225 Figure 6—11 shows some general trends after military coups. First, significant realignment is apparent on some measures of trade, aid, and diplomacy, but not on all. In bilateral relations, intensity and extensity of trade is more significant than intensity and extensity of aid and diplomacy. In multilateral relations, the parameter of intensity quite highly discriminates between the coup and non—coup groups. Four of the five most significant indices deal with alignment. To a large extent, however, these alignment indices show changes in foreign attitudes toward African states rather than African—initiated changes in relations. As was seen in Tables 6-10 and 6-11, the data suggest that the African military regimes were increasingly courted diplomatically by the United States after coming to power, and aid to them from the USA rose substantially. But these states showed no reciprocal special alignment with the United States in trade or diplomatic relations. Indeed, as a group, they became less favorably disposed toward the United States than the non-coup group. It is not as if these relations were unimportant; diplomatic intensity is greater for both military and non-coup groups, even if total trade and aid intensity is down slightly for the coup group. Table 6-14 presents the same data as Figure 6-11, but with the indices ranked according to significance, and with groups means pro— vided. Overall, the measures of realignment tend to be more signifi- cant than those of intensity or extensity, although this observation is somewhat confounded by the fact that there are more measures of alignment than of the other two analytic parameters. From the point of View of the military regimes, the realignment is more in relation 226 Table 6-13.-—Changes in Discrimination Probabilities After Removing North African Countries Change No. Change No. Change No. .0028 1 -.0222 27 —.0808 43 -.0028 11 -.0260 48 .0813 15 .0028 28 .0313 S .0834 39 .0041 25 .0336 23 .0881 44 —.0044 40 .0347 4 —.0900 19 -.0056 10 .0358 24 .1019 13 -.0074 14 .0550 7 .1206 35 -.OO77 6 -.0636 16 —.1537 17 —.0106 34 .0654 12 .1711 36 —.0113 3 —.0726 29 .2001 30 .0134 47 .0740 18 .2556 33 .0163 37 —.0746 32 .2900 45 ..0176 26 .0795 42 -.3200 31 .0183 41 .0801 22 —.4370 46 .0192 38 227 oo.He mm.mm suamcoucH one am oe+mm ooov. om.ms oo.eH omms .ucmsconaa cacao spans em moms. om.moa oo.wv. possessoo .uamsqmmmc moons Hm nH+oH omme. mm.om om.m 1m.ooHo unmeeomac omena mums: om+mm ommm. mm.om om.om poms .uamscmaam one mm mm Hmem. ma.mu oo.oH mosses .ucmecomH< womua Hm mH+oH oomm. mo.nm me.mm summcmuxm cacao mums: em gown. med: we...“ hummcmuxm v.2 mm Hg 3:. eo.mm oH.eH summcmuxm o summewucH cameo am om+om oooa. mm.mm ow.ea ummcsssoo .ucmeeomac omena spas: mm cams. mm.me om.ov .ssoouumms .ucweaomac cameo spans om moom. Hm.oe om.mmu suamcmucH moons Hm m+m+H oooo. om.oH om.mm summcouxm oomue Hm HHIOHIo+m+s+o osmo. Ho.mvm mo.omm summcmuam oamno nudes mm+Hm+mm omso. oo.mm no.0 uwmcsssoo .ucwscomaa 6mm Hm ow oooo. 30H omcgo cog ZD .ucwficmfiam 0.3.3 spas: mm mwmo. om.me mo.m- poms .ucwecmmaa moons am mv+oe ooeo. oo.mo mo.mmu suemcwucH one spans om ommo. oo.aeu oa.oo “mos .ucmeaomam oaamo mm om+sm+om omao. com: com: componuommo xmoaH xmoaH an sausage mooolcoz 9:00 moanmflumkw Inoum mounumm HMUHQOHB can suuoz "mouoom mo swnuaas-:amz buns omumnoomma mmaunamnmaonm cam mmumoaH ocmuaxx.ealo warms 228 to the rest of Africa in diplomacy than in relation to the big powers. What realignment is shown vis—a—vis the big powers may be mostly a function of big-power initiatives, in fact. Each of the analytic para— meters shows somewhat equivocal results. Intensity of multilateral diplomacy is more greatly increased for the coup group than for the non—coup group, but multilateral aid and all three measures of bilateral intensity show the non—coup group becoming more active than the military regimes. Both measures of diplomatic extensity Show somewhat more parochialism in diplomatic relations for the coup group, but in i bilateral trade and aid this group is less parochial (more extensive) than the non—coup group. Tropical African Countries Alone Figure 6—12 presents the same data as Figure 6—11, in this case with the probabilities calculated on the reduced sample we called Tropical African countries. The only statistically significant measures are three variables of bilateral alignment, and these show essentially the same probabilities of discrimination or even greater significance than on Figure 6—11. Generally, though, the indices show less signifi— cance on this table. Indeed, 14 of the 19 indices have higher pro— babilities than on Figure 6—11, indicating less likelihood of discriminating between the coup and non-coup groups. There is no clear explanation for this change in significance of most indices between the tables. In part, it may reflect the difficulties of achieving statist— ical significance in a smaller sample. At this quite small sample— size, any reduction in sample-size rather substantially affects the power of tests. On the other hand, though the changes in probabilities 229 .mumo he scum manmaww>m uoz no cowumnw Hmsumoocoo we scum mumwumoummd uoz umnuwo ma xmoco on» pogo mmumomooo .2.2 .mommnucmuom cw canon mum mmowpco ms» cw owns moanmfluo> on» NO oneness one "0002 .mwwuucooo 0m mooumm HMUHQOHB "Hoo>msmm cowouom mo mmmha so cmmoouw monUCH ozone co mmohucuoo QSOUICOZ one @500 cmmsbmm coaumcwsflnomwnlu.mauw musmflm omoo. u a “ems Hmma. u m mmumum cocoon $3 W some. n a Ammo .1 _ OOHm “mu—”£5500 m oops. u a 1mmo moms. n o STAB was“: .22 8.2 9282on.2 m mama. u o some .8.2 .2.2 weHmzmexm wove. n m 1mm+am+mmo omHH. n a Ammo .2.2 weHmzmezH mono. u m 1om+sm+oao mwom. u 2 ions mmHo. u o 1me+voo mmaaem omaHzo .<.2 mono. n a Aoeo ommm. u m 1mH+omo comm emHzozzoo a. :22 4.2 .38. u m 131.: m aonmm ezmzzomqa mm vooH. u a some. n 2 havo AHHuoano+o+s+oo meHmzmsxm ommm. u a 1om+omo mmoe. n o 1oe+mmo Hmmfl. u m 1m+m+Ho waHmzmozH wocsoquo on momma 230 may reflect a real, substantive and systematic influence of the North African countries on continental patterns of foreign affairs behavior. Tables 6—15 and 6—16 display the changes in probabilities of discrimination of the indices, in precise terms. Composite Change on All Indices It appears that the amount of change in foreign relations after military coups d'état is not predicted by the estimates of pre—coup military influence and regime decision latitude developed in Chapter Four. At least, the predictions developed in that chapter are not borne out by the data on changes in foreign relations activities, though there is a definite relationship between the overall extent of change and economic size. Table 6217 lists gross national product and GNP per capita in 1966 for the coup countries and a composite score of amount of change in foreign relations. This composite score is the addition of percent— ages of change on all trend indices for each state. Spearman's rank correlation coefficients were computed between each of the economic indicators and amount of foreign relations change. Size of gross national product is related to amount of change——significant at slightly better than the .05 level-~but inversely. That is, the smaller the GNP, the greater percentage change in patterns of foreign behavior after military coups d'état. Per capita GNP shows no relation to amount of change in patterns of foreign relations. Table 6~18 lists the coup countries by three rankings. Two rankings are taken from Table 4—5, where the countries are ranked by predictions of degrees of shift in foreign relations after coups 231 Table 6—15.--Trend Indices: Probabilities Associated With Mann— Whitney U Scores North and Tropical Africa Tropical Africa Only 35 Countries 30 Countries Probability Variables in Index Variables in Index Probability .0156 19+27+28 19+27+28 .0125 .0336 39 44+45 .0158 .0480 44+45 40 .0562 .0582 23 39 .1170 .0606 40 25 .1251 .0720 22+3l+33 26 .1251 .0770 6+7+8+9+10+1l 1+2+5 .1251 .0880 1+2+5 23 .1788 .1075 26 34 .1788 .1210 25 41 .1894 .1660 29+30 6+7+8+9+10+11 .1894 .1711 41 29+30 .2358 .1894 34 16+17 .2578 .2206 l4+15 38 .2643 .2451 38 l4+15 .3015 .3330 32+36 37+40 .4052 .4130 16+17 22+3l+33 .4404 .4562 24 32+36 .4562 .4880 37+40 24 .4920 232 Table 6—16.--Changes in Discrimination Probabilities After Removing North African Countries Change No. Change No. Change No. .0020 l9+27+28 .0192 38 -.0828 37+4O .0041 25 —.O355 44+45 .0834 39 —.0044 40 .0358 24 .1124 6+7+8+9~lO—1l -.0106 34 .0371 l+2+5 .1206 23 .0176 26 .0678 29+3O .1232 32+36 .0183 41 .0809 l4+15 —.1552 16+17 .3684 22+3l+33 d‘état. They are ranked, first, according to a prediction based on estimates of pre-coup military influence and of decision—latitude as determined by economic size. They are ranked, second, according to the prediction based on pre—coup military influence and decision—latitude as determined by per capita economic development. The third ranking Table 6—18 is the ordering of countries according to composite change on all trend-indices. It should be noted that the two methods of predicting degrees of shift in foreign relations are based partly on the measures of gross national product and per capita GNP presented in Table 6—17, but also on measures of pre-coup military influence in government, as developed in Chapter Four. Thus, these rankings of countries by predictions of change are very similar to the rankings by GNP and GNP/capita, and that similarity is also evident in the correlation coefficients of the two tables. Table 6-18 shows again that the extent of reorientation of foreign relations after military coups d'état is not positively related 233 to the level of pre-coup military influence in government or to the economic estimates of decision—latitude. Comparing the correlation coefficients in Tables 6-17 and 6—18, we can infer that including the estiamte of pre—coup military influences——developed in Chapter Four—— in the technique for predicting degrees of shift in foreign relations was not very helpful. That conclusion is based on the similarity of correlations in the two tables. The correlation between GNP and compos— ite change in Table 6—17 is almost identical to that between composite change and the predicted degrees of shift based on size given in Table 6-18. In the latter table, of course, the predicitions based on size used GNP as the measure of size. The correlation between GNP/capita and composite change, given in Table 6-17, is somewhat more distinct from its companion correlation in Table 6-18, that is, the correlation between composite change and the predicted degrees of shift in foreign relations based on development and decision—latitude. While the correlation in Table 6—17 of 0.0061 seems of no consequence whatever, the companion correlation of —0.1636 in Table 6—18 is not statistically significant either. This latter correlation is negative, like the other corre— lation in the table, suggesting a general conclusion. Governmental decision—latitude, as predicted from economic size (GNP), is signifi— cantly but negatively related to the extent of reorientation of foreign relations after military coups d'etat, but that reorientation is not related to per capita development and only very slightly, and negatively, related to the extent of military influence in government before the coup. Thus, countries having very small economies, though not necessarily those having the least per capita income, tend to 234 Table 6—17.--Relations Between Country Economic Measures and Composite Change on All Trend Indices Composite % GNP Change on Country (millions) GNP/Capita All Indices Nigeria 4603 77 737.47 Algeria 3040 243 1356.81 Ghana 2492 314 1288.56 Congo Kinshasa 1730 108 908.38 Sierra Leone 361 150 1398.97 Upper Volta 245 49 1299.28 Togo 209 124 1554.88 Dahomey 193 80 1872.87 Central African Rep. 187 133 1022.24 Burundi 154 47 2071.72 Spearman's rho —.6121 .0061 Table 6-18.—-Relationship Between (1) Predicted Country Rankings of Degrees of Shift in Foreign Relations, and (2) Actual Changes. (1) Predicted Shifts, Ranked (2) Actual.Changes, (a) Development (b) Size Ranked Ghana 1 l 7 Nigeria 2 2 10 Algeria 3 3 5 Sierra Leone 4 4 4 Dahomey 5 8 2 Togo 6 9 3 Central African 7 6.5 8 Rep. Upper Volta 8 5 6 Burundi 9 10 1 Congo Kinshasa 10 6.5 9 Spearman's rho —.1636 -.6151 235 show the greatest amount of change in patterns of foreign relations, and countries with the largest economies the least change. Of course, these countries with small economies remain less intensely and less extensively involved in foreign affairs than the larger countries; the actual changes in foreign relations reported in Tables 6-17 and 6—18 are percentages of changes from their pre—coup patterns. CHAPTER VII DISCUSSION OF RESULTS General Continental Trends, 1964—67 Selected International Activities In order to get an overall view on African international activites, Chapter Six presented data for all of the countries in this study on four quite different foreign-affairs behaviors: exports per GNP, aid per GNP, bloc-voting in the United Nations, and member- ships in international governmental organizations. The data showed that on these variables the African countries varied a great deal in their activities, both in 1964 and in 1967. The coup countries were not significantly different from the non—coup countries in their per— formances, and there was little similarity in the distributions of countries across these four variables. This was reflected in the generally low correlations among these variables in 1964, and even lower correlations in 1967. Total amounts of trade and aid were highly correlated with gross national product, as were the numbers of memberships in IGO's, in both years, though EEEEE of trade and aid-- that is, trade and aid divided by GNP--were not similarly correlated with size of economy. Rates of exports-—that is, exports divided by 236 237 GNP-—were highly correlated with GNP's per capita in 1964, but not in 1967. Over the period from 1964 through 1967, many patterns remained the same, though a few changed significantly. The distributions of countries on each of the four variables looked very much the same in 1967 as in 1964; there were very few countries which greatly changed their rankings on these variables. There were general continental shifts in all four kinds of activity, however. Nearly all countries showed drops in the rates of exports and in the rates of aid received, and nearly all showed increases in memberships in intergovernmental organizations and clearer alignments with the Soviet Union, as against the United States, in United Nations voting. On these latter two diplomatic variables, not only were activities increased, but the amount and percentage of change from 1964 to 1967 was significantly greater than on the economic variables. V In 1964, the countries having larger economies were, as a group, distinctly more active in international affairs than those having smaller economies. The larger economies carried on a greater amount of trade, they got a greater amount of aid, they were members of more intergovernmental organizations, and they were more boldly aligned in the United Nations, voting more often against the United States and for the Soviet Union. It is true that the larger economies did not show greater rates of trade (exports/GNP) or greater rates of aid received (aid/GNP) than the smaller economies, but they nonetheless dominated trade and aid activities of African states by sheer volume. By 1967, the African economies were under greater pressure internationally, doing less trade and receiving less aid. We can say 238 that the economies were turning inward, since all were growing, though to greatly different degrees, yet they tended to receive less and less aid from abroad, and they tended to sell a greater percentage of their products and services at home. The larger economies consistently carried on greater trade and received more aid than did the smaller throughout the four—year period under study. However, the smaller, poorer countries began to do better in spite of their few economic resources; that is, per unit of GNP, they attained greater rates of trade and aid by 1967. The relationship of this general change in the African economic situation to African diplomacy is not clear, or at least not consistent for all the countries. Nearly all the countries increased their international activities from 1964 through 1967. One example of this increase is their greater numbers of memberships in international governmental organizations. In addition, most countries became more hostile to the United States, and less so to the Soviet Union, in their voting in the United Nations. We cannot conclude, however, that the African states expressed greater hostility to the United States in the United Nations because they blamed the USA for contributing to their difficult economic situation. For one thing, the Soviet Union also decreased its overall aid to Africa during this period. For another thing, of the seven countries (Ivory Coast, Kenya, Sierra Leone, Malawi, Uganda, Ghana, Gabon) whose UN voting showed increased align- ment with the United States as opposed to the USSR, all seven showed a drop in rates of exports and in rates of aid received, some (Ghana, Uganda) showing some of the most severe decreases in these economic relationships and yet the greatest degrees or rapprochement with the 239 USA in UN voting. On the other hand, some of those countries (Congo Brazzaville, Togo, Libya) which most strikingly turned against the United States in UN voting were among the countries suffering the most extreme decreases in rates of exports and rates of aid received. United Nations voting came to show more variation among the African countries by 1967, and it became signficantly more divisive among them. That is, on this variable, in contrast to the other three, there is much less evidence of a general continental shift over the four years. A few generalizations emerge from this analysis of selected international activities. First, the African continent generally showed reductions in the rates of exports and rates of aid received, allowing the inference of a deteriorating international economic situation for them. Second, Africa generally showed an increase in diplomatic activity as indicated by more numerous memberships in international governmental organizations. Third, African states showed sharper alignments with the big powers in United Nations voting, most becoming more hostile to the United States but a few signifi- cantly more aligned with the USA. Fourth, the countries which showed the most extreme changes in behavior on these selected activities also showed one of two characteristics: they had either undergone military coups, or they had recorded extreme rates of activity in 1964. Thus, their great changes in behavior apparently came about either because of new decisions by new military leaders or because of desires or needs to moderate previously extreme behavior, or for both of these reasons. In international relations over this period, African states acted much more homogeneously in international economic affairs than 240 in diplomatic. It is interesting to note that international economic relations are not fully under the control of African governments. Rates of exports and aid depend as much or more on the decisions of foreign businesses and governments as on the decisions of African governments. Thus, it may be more accurate to say that foreign countries treated Africa as a class of states in economic affairs, than to say that Africa behaved as a class on these affairs. In diplomatic affairs, African countries acted in part as a single class of states, (e.g., nearly all increased their memberships and activities in international organizations), and in part as a heterogeneous group, showing a general tendency to become more aligned with the Soviet Union in United Nations voting, but with a few states clearly bucking that trend. Three types of actors seemed to emerge on the African continent, as determined by performances on the selected variables. In both 1964 and 1967, we can distinguish two mutually exclusive but not exhaustive groups, which I have called the Exporters and the IGO Diplomats. These groups do not coincide with the coup and non—coup groups. The Exporters are countries with very high rates of exports but ranked quite low in IGO memberships. These states, which included Zambia, Gabon, Liberia, and Mauritania, tended to align themselves with the United States in United Nations voting. The IGO Diplomats, on the other hand, are states which show many IGO memberships but low exports—rates. These IGO Diplomats included the Maghreb countries, the UAR, Nigeria, and Ghana, and they tended to be much more aligned with the Soviet Union in UN voting. 241 By 1967, a third group had emerged which showed high activity both in rates of exports and in IGO memberships and strong alignment in the United Nations. The direction of voting alignment was not consis— tent across this whole group, however; some of these states voted heavily with the Soviet Union, and other states clearly favored the United States. I shall call these states the Actives. With regard to the four selected international activities, they are the most "aggres— sive" states according to my definition in Section 5 of Chapter Three. That is, they show highly intensive and highly extensive activities, and they tend to be highly aligned with one of the superpowers. These four variables do not provide enough measures to judge their actor-type overall, however, especially with regard to extensity, and, in fact, discussion of total performances on the analytic parameters shows that the Actives are not the most "aggressive" states. Nonetheless, on the selected variables, the Actives emerge as a distinct subset of African states, and, internationally, they are probably the most influential of the groups identified. Analytic Parameters Data on the performances of African states on the analytic parameters of intensity, extensity, and alignment show that the "coup" countries are not significantly different from the "non—coup" countries in either 1964 or 1967. The coup countries do differ significantly on only one measure of change from 1964 to 1967: that is, in the amount of change in the extensity of their relations. However, the rather greater change in extensity is not the same for all the coup countries; some became much more extensive in their relations, and some much less. 242 Over the whole continent, it is difficult to identify any simple shift in overall intensity of international activites from 1964 through 1967. Changes in the intensity of economic activities are less clear on a continental scale than changes in the intensity of diplomatic activities. Though there are general, if small, reductions in the rates of trade and aid for most African countries, total imports and total aid do slightly increase. Continental averages on these economic activities are too gross for drawing conclusions, however, since they mask a great deal of variation among the countries in the amounts and directions of trade and aid relations. Changes in diplo— matic intensity are much more consistent across the continent, with nearly all the states becoming more active, sending diplomatic missions to more countries and joining more international organizations. The continental shifts in extensity of international relations are more clear than for intensity. In both trade and diplomatic aspects, most African countries became more extensive in their relations from 1964 through 1967. That is, most of these states exported goods' to a greater number of countries, they sent diplomats to more foreign countries, and they joined more international organizations having large memberships. Nonetheless, in some ways foreign relations generally became less extensive and more concentrated. This concentra— tion was especially apparent in sources of aid, where there seemed to be two patterns. The countries having continuous civilian rule received more and more of their aid from multilateral sources, whereas states experiencing military coups found that multilateral and communist bloc sources dried up and they came to rely on a few western countries. Lastly, even though most African states developed trade with more 243 countries over the four years, a greater percentage of that trade became concentrated with one or two trading—partners for many of them. Continental shifts in alignment are about as clear as shifts in extensity. In trade relations, nearly all states showed a greater percentage of trade with other African countries. This increased economic orientation to other members of the continent is not so clearly evident in diplomatic affairs, however. For one thing,, although diplomatic relations within the continent continued to grow over the four-year period, a greater percentage of the memberships of African states in international organizations grew in organizations located primarily outside Africa. That is, the African states joined more and more organizations having a predominance of non-African mem— bers and having home offices located outside Africa. In relations with the superpowers, most African states established clearer preferences with either the United States or the Soviet Union. This change is especially clear in their voting in the United Nations, but it is also evident in trade and aid relations, which became less evenly balanced between East and West. Diplomatic and economic alignments were not always consistent, however. Diplomatically, at least in the United Nations, a majority of African states beCame much more clearly aligned with the Soviet Union. Economically, however, more African states showed an increased concentration of relations with the United States than with the Soviet Union. Overall, though, we can say that alignments became sharper in this period, and, thus, nonalignment, so important in African rhetoric in the early 19605, became less practiced. Plotting the behavior of countries on all three analytic para— meters on the analytic cube shows that no single parameter dominated 244 in the patterns of shifting relations from 1964 through 1967. Fewer countries showed extremes of behavior on all three parameters in 1967 than did in 1964. Thus, even though alignments generally seemed sharper in 1964 than in 1967, fewer countries could be plotted on the analytic cube as "aggressive" actors in the later years. We can see intuitively how this happens when we remember the definitions of "alignment" and "extensity." I have defined alignment in terms of preference for either of the superpowers or for Africa in a set of relations, and I have defined extensity by the even spread of relations across many other states. Thus, as a state becomes more aligned it tends to make its behavior coincide with that of one of the superpowers, or with the majority of the African countries, and, in doing so, it may tend to concentrate its relations with the power or set of states with which it has become aligned. Alignment may increase concentration rather than extensity in relations, therefore, though this is not necessarily so. Table 7-1 shows that the three parameters are quite significantly intercorrelated in 1964, and Table 7—2 shows that in 1967 two of the three correlations have dropped to insignificance. Table 7—l.-—1964 Intercorrelations Among Parameters Intensity Extensity Alignment Intensity 1.000 .528 .499 Extensity 1.000 .710 Alignment 1.000 245 Table 7-2.--1967 Intercorrelations Among Parameters Intensity Extensity Alignment Intensity 1.000 .035 —.039 Extensity 1.000 .447 Alignment 1.000 We can draw these conclusions from these tables. In 1964, a country with a high score on one parameter is likely to have a similarly high score on both other parameters, and, conversely, a low score on one parameter is associated with low scores on the others. Thus, a highly aligned actor in 1964 was likely also to be an intense actor, and one with relations in many parts of the world. Ghana was the best example. By 1967, however, behaviors on the three parameters did not go together so neatly. In fact, at this later date there was essen— tially no correlation between intensity and the other two parameters. Extensity and alignment continued to be significantly correlated, but markedly less so than in 1964. Another way of describing this change in the relationships of the parameters may be to say that, by 1967, aggressive African actors discovered that they simply could not do everything in foreign affairs. They could not be intensive, extensive, and aligned in foreign affairs at the same time. We may infer that their resources were spread too thin and their alignments confused by their attempts at extensive relations, and so something had to give. Table 7—3 presents correlations among the parameters in their amounts and direction of change from 1964 through 1967. These 246 Table 7—3.—-Intercorrelations of Country Rankings in Amount and Direc- tion of Changes on Parameters, 1964—67 Intensity Extensity Alignment Intensity 1.000 .368 .107 Extensity 1.000 -.089 Alignment 1.000 correlations amplify our understanding of the continental shifts in behavior on these parameters. They show two insignificant correla— tions—~those between intensity and alignment (.107) and between exten- sity and alignment (—.089). Thus, on a continental basis, changes in alignment were not systematically associated with changes in either the intensity or extensity of foreign affairs. One correlation is statis— tically significant, that between intensity and extensity, where r = .368. The probability of finding a correlation this large in a sample of this size is approximately 0.02. The correlation means that, on the whole continent, there was a tendency for increases in extensity of foreign affairs to accompany increases in intensity. This is intuitively reasonable, because one way in which a foreign-affairs behavior becomes measured as more intense is to find that behavior extended to more and more countries over time. Despite the statistical significance of the correlation between the changes in these parameters, it should be noted that the correlation itself is only a moderate one, allowing only 135% of the variation in one parameter to be accounted for by the other parameter. Thus, it is possible for these two parameters to Show an appreciable positive relationship in their 247 changes from 1964 through 1967, and still end up with an insignificant correlation between them at the end of 1967. In some respects, then, African states behaved as a class in their changes in behavior from 1964 through 1967, and in some respects they did not. Nearly all showed increases in diplomatic intensity and in the extensity of both trade and diplomacy. Most showed decreases in the intensity of trade and aid. Also, most showed increases in alignments, although the direction of alignment was not the same across all countries, nor even always the same across trade, aid, and diplo— matic affairs for the same country. In reviewing the overall continental behavior of Africa from 1964 through 1967, it appears that there was appreciable change in the selected variables and in the analytic parameters. It appears that these states behaved as a class in some respects, but also their behaviors as a class and in their individual state variations suggest that each state was searching for its own formula which would improve its foreign contacts, influence, and benefits. .Finding such formulae was not easy. Diminishing rates of trade and aid suggest that African states did not find themselves in strong bargaining positions economic— ally. A goodly number of these states found themselves in the awkward position of crossing the United States diplomatically while being importantly tied to the USA financially. Conversely, it should be noted that the continent did not succumb en masse to difficult times or outside pressures. 248 Confirmation or Disproof of Hypotheses of Military Influences Extent of Change After Coups d‘Etat I have hypothesized that the amount of change in policies after military coups d‘etat would be related to two factors: the extent of military influence in the government before the coups, and the decision- latitude of the new government. Military influence was postulated to be a function of the burden of military expenditures, the size of the army, and the percentage of the population recruited to the army. Decision—latitude was postulated to be a function of overall social— economic development and size of economy. The data presented in Chapter Six show that actual changes in foreign relations after military coups are related to decision- latitude measures and army-influence measures, but not as originally predicted. The composite percentage of change in foreign relations after military coups shows no relation to the burden of previous military expenditures, nor to development measured by GNP per capita. Composite change does show a significant negative correlation (—.612) with gross national product, one possible indicator of decision- latitude, and an even greater significant negative correlation (-.736) with army-size. These numbers indicate that the greatest rates of reorientation of foreign affairs come when small armies take over countries with small economies. Obviously, gross national product is too simple an indicator Of decision-latitude, for the largest economies (those with presumably the greatest decision—latitude) generally show the smallest percentages Of change in foreign affairs, while the smallest economies show the 249 largest rates of change. Also, it makes no sense to argue that governments with restricted decision—latitude show greater degrees of change in foreign policies than governments enjoying broader decision- latitude. Yet we cannot overlook that fact that there is a relationship between degrees of change in foreign relations and size of gross national product. Apparently, a large economy does not grant a government great latitude in decision-making, as compared to the latitude available to a government ruling a small economy. I see two hypotheses which might be explored in another study to explain this relationship. On the one hand, it may be that the larger the economy the further it is from governmental control, and thus new governments are not so likely to be able to reorient the foreign relations of larger economies as are new governments trying to reorient smaller economies. On the other hand, it may be that larger economies are not so vulnerable to foreign influences as smaller economies. Thus, when military regimes take power by coups d'état and, perhaps, thereby scare foreign trading partners and investors, smaller economies are likely to show greater effects of the consequent decisions of foreign governments and businesses. In any case, it is clear that the countries with small economies show a much greater rate of change in foreign-affairs behaviors after military coups than do large economies. The second general factor that was predicted to influence the extent of change in foreign affairs after coups d'état was the overall influence of the military on government before coups. The notion was that countries showing a high degree of military influence on govern- ments before coups would probably not show very great reorientations of policy after coups, because the military would already have 250 influenced policy substantially. Conversely, states showing little military influence before coups would be expected to show much greater degrees of reorientation of foreign-affairs behavior after coups. There is a relationship between the identified factor and the percentage of change in foreign behaviors, and that relationship does confirm the proposed hypothesis. Taken separately from estimates of decision-latitude, the overall pre-coup influence of the military correlates negatively, but without statistical significance, with the extent of post-coup changes in foreign—affairs behavior, at r = -.380. This overall pre-coup military influence was calculated from three factors: influence estimated by the rate of expenditures on the mili— tary, by the size of the army, and by the percentage of the population in the army. Taken separately, two of the factors show no correlation with the extent of post-coup change; whatever the meaning of military expenditures as a percentage of GNP and of the percentage of the population in the army, these factors have no relationship to post- coup changes in foreign behavior. However, size of army correlates negatively and quite significantly with post—coup changes, at r = -.736. Here we must reinterpret a factor related to the degree of post- coup change. Originally, I argued that there were several facets to military influence, army-size being only one. The data show, however, that it is only army—size which is related. These data do confirm the general hypothesis. That is, the correlation allows us to argue that countries with small armies show greater degrees of change in foreign behavior after coups than do states with large armies, perhaps because small armies were less able to influence policies before coups. Once in power after coups d'état, small armies gain an opportunity to 251 change policies greatly. While sound as far as it goes, I do not feel this argument is conclusive. Could we not also argue that larger armies will have greater demands than small armies? Also, whatever their influence before taking over governments, there is no reason to suspect that the demands of larger armies will grow at a smaller rate than those for a smaller army. Finally, it is not always true that larger armies have greater pre—coup influence on governments than do smaller armies. The correlation between army size and rate of expenditures on the military is not significant for the ten coup countries, for example. I infer that small armies effect greater changes in foreign behavior after coups than large armies not only because smaller armies are making up for less pre-coup influence, but also because small armies are more likely to make decisions with a single mind. I believe we can expect smaller armies to have fewer factions than large armies, and to have fewer officers with personal, non—policy interests to satisfy. When a small army takes over a government, probably a greater proportion of it is involved in the coup than when a large army takes power. This active proportion already represents a kind of consensus about reorienting the government. If the same proportion of men in a large army had to be involved for a successful coup as in a small army, it is likely that so many men would be aware of the coup plans that secrecy would evaporate and the coup would be countered. In sum, I argue that a small army is more likely than a large army to achieve consensus within the army on bold new policies, and thus a small army can be expected to effect greater changes when it comes to power than a large army. Further, small armies taking over small, 252 and therefore more readily manipulable, economies are likely to be able to bring about the greatest percentage of change in foreign behavior. Directions of Changes After Coups d'Etat As in the previous section, the data shew some surprises with regard to my hypotheses on the directions of changes in foreign-affairs behavior after military coups. I believe, however, that the data allow us to refine the image of the military developed in Chapter Four rather than overturn it. Figure 7—1 displays all the substantive and analytic areas for which hypotheses were developed in Chapter Four. Using the data from Tables 6-20 and 6—21, I have identified in Figure 7-1 the areas in which the coup countries have most clearly reoriented the foreign behavior and become distinct from the non—coup group. I have included in the Figure all the areas where the difference in behavior between the two groups would occur by random change fewer than 10 times out of a hundred. There are eight such areas. States experiencing military coups d'état show distinctlv less intense trade, especially with the United States, but more extensive trade relations than states with continuous civilian governments. The coup group shows significantly less aid from the Communist Bloc and from multilateral sources. Finally, the coup group shows much greater diplomatic activity in international organizations, greater diplomatic activity with the United States, and greater change in diplomatic relations with the rest of Africa than the non-coup group. My hypotheses did not expect these changes in trade-patterns, but I believe that is because I did not foresee the extent to which 253 Trade Aid Diplomacy Intensity Lower * * H (.0880) _g xtensity Higher * * a (.0770) g lignment : With Africa * * * m Communist Bloc * Lower(.0606) * With U.S.A. Lower(.0480) * Higher(.0156) Intensity N.A. Lower(.0336) Higher(.0720) H m 'gExtensity N.A. N.A. * 0 . fiAlignment H _ . High change in H 3 Wlth Africa N'A' N' both directions(.0582) g Communist Bloc N.A N.A * With U.S.A. N.A N.A. * Figure 7-1. Actual Changes in Foreign Relations After Military Coups D‘Etat *indicates no signficant difference between coup and non~coup groups. Changes are indicated here only in those areas where the differences between groups probably would occur randomly in less than .10 of the cases. N.A. indicates Not Appropriate. 254 trade was beyond the control of the new military governments. I am sure the new military regimes did not desire less trade, but that is what they got. Undoubtedly, the abrupt changes in governments by coups d'état frightened foreign traders and investors and disrupted marketing procedures, at least in the first year after the coups, when my measurements were taken. The coup group may have responded to this pinch in trade by seeking new markets; at any rate, it became signifi— cantly more extensive in trade relations than the non—coup group. This increase in extensity of trade suggests that the military do see the practical and tangible importance of good foreign economic rela- tions, as I hypothesized, especially if they suffer general reductions in imports from abroad. Referring back to Table 6—16, which lists the individual variables according to their powers of discriminating between the coup and non-coup groups, we see further evidence for this interpretation. It is in the imports side of trade that the coup countries show significant decreases, not in exports. Total exports and exports/GNP do not change significantly for the coup group as compared to the non—coup states. Furthermore, the variable which most significantly discriminates between the groups is a measure of critical balance of payments (exports minus imports/GNP), and the coup group shows a distinctly greater tendency to curtail excessive imbalances in international payments. Of course, this change in balances of payments is influenced by the drop in imports discussed above, but this change can also be said to be evidence for the controlled, cost— accounting style of behavior I hypothesized for the military. In aid, the coup states generally suffer, receiving signifi— cantly less aid from both the Communist Bloc and from multilateral 255 sources than the non-coup group. The period from 1964 through 1967 was one in which the continuous civilian governments were working to diversify their sources of aid, becoming distinctly less dependent on any single country for aid and more dependent on multilateral sources. The coup group showed completely opposite changes. Appar- ently, communist aid was decreased for ideological reasons, as the Soviet bloc interpreted the new military governments as more reac— tionary than the civilians they overthrew, especially in the bigger states (Ghana, Congo Kinshasa, Nigeria, Sierra Leone). On the other hand, multilateral aid apparently dropped because the multinational bankers who controlled the purse strings had to await the formal recognition of the new military governments by their organizations, and because they generally believed strongly in stable government as a prerequisite for utilizing aid effectively. Average aid from the United States actually increased more for the states in the coup group than those in the non—coup group, although the variations in aid-relations within both groups are so great that USA aid is not a distinguishing variable between them. Certainly, we cannot simply interpret the overall lower receipt of aid by the coup group as an indication that military governments are uninterested in aid. Aid, even more than trade, is determined by foreign suppliers rather than by African governments, and we know that these suppliers are notoriously sensitive about the appreciation and stability of recipients. ’ Finally, the new military governments show some important changes in diplomacy which distinguish them from civilian governments. On the one hand, they show significant increases in the intensity of multilateral diplomacy, by their increasing memberships and activities 256 in international organizations. On the other hand, they find themselves courted with greater intensity by the United States, which signifi— cantly increases its expenses within the coup states to further educational and cultural relations, mostly by sending more American educators and groups of artists, and which significantly expands the program of bringing African elites to the United States on “get- acquainted" and "good—will" tours. It is interesting to note, however, that this bilateral attention on the part of the United States does not bring about greater pro-U.S. alignment by the African military govern- ments in multilateral affairs, such as voting in the United Nations. Therefore, there is no simple alignment of the military governments; this tends to confirm my argument in Chapter Four that the military are likely to make alignments on a pragmatic basis rather than because they covet the purity of non—alignment or because they have clear 1 ideological leanings. On the other hand, my hypothesis that the l military would tend to constrain diplomatic relations is disconfirmed. As a group, the military governments do not differ from the non-coup governments in the changes in numbers of diplomats sent out and received, nor in changes in the number of memberships in international organizations. For both the military and civilian groups, there are general increases in these activities. General Implications of Military Influences in African Foreign Relations Overall, this research allows us to clarify, to some extent, our image of the military in government and their impact on foreign affairs of African states. First, military governments do not shrink from substantial involvement in international affairs in essentially 257 nonmilitary ways. In some aspects of both trade and diplomacy, the African military governments showed greater activities in the first year after coups than before them. Second, while some of their greatest differences from the civilian governments may imply withdrawal or incrementalism or con- servatism in foreign affairs, some of the most distinguishing features of the military's foreign affairs suggest boldness and aggressiveness. Reductions in the intensity of trade and in the supply of aid from the Communist Bloc may suggest conservatism, but, as I have interpreted it above, these changes, along with higher extensity, higher diplomatic contact with the United States, and lower trade with the United States, suggest that much of foreign affairs is out of the control of these African governments. Despite lower trade, military governments have not allowed their economies either to collapse or to become much more dependent in trade on a few foreign partners. Rather, they have moved to balance international payments more favorably and to extend trade relations to more countries. Third, these military governments show no systematic ideological preference in international affairs, but neither do they show great concern for nonalignment or noninvolvement with ideological blocs. We might see reduced aid from the Communist Bloc as an indication of military preference for the Western Bloc, but such a simple inter- pretation clearly does not hold up under scrutiny. Aid from the United States does not significantly increase for the new military regimes to take the place of communist and multilateral aid. In fact, trade with the United States drops significantly. On the other hand, diplomatic initiatives from the United States increase significantly, f — —- < _ _-.~ " .V n v 258 and the military governments do not disparage them. At the same time, these same military governments tend, along with the rest of the continent, to become less and less aligned with the United States in United Nations voting. Overall, then, in their dealings with the ideological blocs and with the rest of Africa, the military regimes seem genuinely pragmatic, being quite active, taking what they can get, and not being incapacitated by worry about balancing relations between East and West. Also, there is no evidence that these military govern- ments have gone to great lengths to expound general political principles or to identify their ideological predilections. Fourth, while these conclusions tend to confirm that the military places more emphasis on national-interest than on ideological foreign policy, these two emphases in foreign policy obviously are not mutually exclusive. The pattern of military foreign relations is, I believe, realistic, adaptive, and inductive, but it is not neces- sarily modest, conservative, non—ideological, or compromising. Above all, military regimes are highly active diplomatically, suggesting a more intense concern for external affairs than the naive "national- interest" image implies. And, in some respects, the military are purists, especially in their own demands for liberation of African colonies and for Africans' control of their own affairs. I think it is clear that the foreign affairs of African military governments fit much better the image of national-interest than that of ideological foreign policy, but not to such an extent that the national-interest image implies that military governments will be weak or mindless or vacil- lating or shrinking violets in world affairs. 259 Fifth, despite the conclusions above, it is clear that in the short run, at least, military coups d'état tend to interfere with national development. Coups are likely to have the effect of drying up aid and disrupting trade relations. This is not evidence that the military are less interested in economic development, but the military must take the consequences of economic disruptions and squeezes which usually follow coups. Also, despite the immediate interest of the United States, it is clear that the new military governments must take some time to establish foreign contacts, to gain recognition and respect, and to make clear, at home and abroad, their policy— orientations. During this time, domestic development and foreign aid generally slow down, and it is not clear from this research whether the long-term achievements of military governments effectively justify the disruptions caused by their abrupt seizures of power. Sixth, there is no evidence that military governments control events abroad any better than civilian governments. Military regimes suffer problems in foreign relations (e.g., diminished rates of aid and imports) despite their diplomatic efforts and their pragmatism. When they take over large and relatively more complex countries, their reorientations of foreign affairs are likely to be less pronounced than when they take over small countries. This is evident in our earlier observation that the greatest changes in foreign affairs come when small armies take over small economies. In sum, of the two images of the military provided by Lucian Pye at the beginning of the 19605, this research has found in Africa in the mid-19605 no evidence in foreign-affairs behaviors to support the view that the African military goverments differ from civilian 260 governments by being distinctly incompetent, inactive, and reactionary. This research has found a good bit of evidence of military activism and pragmatism in foreign affairs. None of the alternative images of the military described in this study is fully confirmed, however. There are two major reasons. On the one hand, this research shows that many events that help to determine the foreign affairs of African states are beyond the control of those states, and that in the first year after coups military governments therefore suffer losses which they might later recover, particularly in trade and aid. On the other hand, alternative images of the military have generally been drawn up with their domestic behavior in mind, the extrapolation of which to foreign affairs is somewhat ambiguous. The research does tend to con— firm the judgment of Gutteridge (1970) "there has been no firm ideolog— ical basis for action (among African military regimes)," but it also confirms my view that there are many similarities in orientation to foreign affairs which significantly distinguish the military from civilian regimes, and which make the military an important subset of African governments to be reckoned with in international relations. APPENDICES APPENDIX I AFRICA RESEARCH VARIABLES APPENDIX I AFRICA RESEARCH VARIABLES Description of Variable Variable # l Imports/GNP 2 Exports/GNP 3 Imports in millions of USA dollars 4 Exports in millions of USA dollars 5 Imports plus exports, rounded to nearest millions of USA 10 11 12 13 14 15 16 17 18 19 20 21 22 dollars Number of countries from which imports are received Number of countries to which exports are sent Number of countries from which imports are received in an amount equal to 10% or more of the total imports Number of countries to which exports are sent in an amount equal to 10% or more of total exports % of total imports coming from the one country sending the largest amount of imports % of total exports going to the one country which received the largest amount of exports Compared to variable 10, the second highest % of total imports coming from one other country Compared to variable 11, the second highest % of total exports going to one other country % of imports from all "Other Africa" (does not include UAR) % of exports to all "Other Africa" (does not inlcude UAR) % of imports from Communist Bloc % of exports to Communist Bloc Exports minus imports/GNP, with plus or minus indicated Number of USA diplomats assigned resident to the country Estimated GNP at market prices, in millions USA dollars Estimated GNP/capita at market prices, in USA dollars Rate of participation in roll—call votes in UN General Assembly and its Committees (1963 & 1964 votes combined, other years separated) 261 23 24 25 26 27 28 29 3O 31 32 33 34 35 36 37 38 39 4O 41 42 43 44 45 46 47 48 2621 Rate of voting against African majority in UN General Assembly & Committees Rate of voting against USA in UN General Assembly and Committees Rate of voting against USSR in UN General Assembly and Committees Difference: rate of voting against USA minus rate of voting against USSR Number of scholars and artists received in USA under Dept. of State grants USA educational & cultural expenses in the country, in USA dollars Number of countries to which diplomats are sent and are resident Number of countries from which diplomats are received, resident only Total number of memberships in IGO's other than UN and UN— related % "African only" IGO memberships of total IGO memberships (Var. 40/Var. 39) Total number of members in all IGO's the country is a member of Variable 33 divided by the number of IGO's the country is a member of Total number of African members in all IGO's the country a member of Variable 35 divided by the number of IGO's the country a member of , Total official financial aid (net) from all OECD countries, in millions USA $ Total official aid from the USA, in millions of USA dollars Total official aid from multilateral sources (UN, World Bank, etc.), millions $ Total communist bloc economic grants and credits, 1954 to date, millions USA $ Largest amount of aid from any one country, in millions of USA dollars Grand total of all aid (OECD plus Multilateral plus Communist) Grand total all aid/GNP (Variable 42/Variable 20) % of imports from the USA % of exports to the USA % of imports from USA minus % of imports from Communist Bloc (Variable 44 minus Variable l6) % of exports to USA minus % of exports to Communist Bloc (Variable 45 minus Variable 17) Total official aid from USA minus Total Communist Bloc aid (Variable 38 minus Variable 40) APPENDIX II SOURCES OF DATA BY VARIABLES 10. 11. 12. 13. 14. APPENDIX II SOURCES OF DATA BY VARIABLES Derived from variables 3 and 20. Derived from variables 4 and 20. Primary source: Direction of Trade, Annual volumes; cross-checked with data in the UN Statistical Yearbook, 1968, page 394—5. Same as variable 3. Derived from variables 3 and 4. Source: Direction of Trade, Annual volumes. The sum of all countries for which any of imports are shown. Source: Direction of Trade, Annual volumes. The sum of all countries for which any amount of exports are shown. Source: Direction of Trade, Annual volumes. The number of countries each of which represents at least 10% of the year's imports. In some cases where trade with selected countries is apparently very low, 291 reports regional subtotals; here, I have counted the region as a single country. Source: Direction of Trade, Annual volumes. The number of countries each of which represents at least 10% of the year's export trade. Regional subtotals where no country figures are presented are treated as a single country, as in variable 8. Computed from Direction of Trade data; the amount of imports from the country sending the greatest amount is transformed into a percentage of the total imports. Computed from Direction of Trade data; the amount of exports to the country receiving the greatest amount is transformed into a percentage of the total exports. Computed from Direction of Trade data as in variable 10, only the percentage is computed for the country sending the second highest amount of imports. Computed from Direction of Trade data as in variable 11, only the percentage is computed for the country receiving the second highest amount of exports. Percentage computed from Direction of Trade data. In its category of "Other Africa," QQT_includes the following countries: 15. l6. 17. 18. 19. 20. 21. 264 Algeria Gambia Mauritius Sao Tome Angola Ghana Morocco Senegal Burundi Guinea Portug. Mozambique Sierra Leone Cameroon Guinea Republic Niger Somali Republic Cape Verde Is. Guinea Spain Nigeria Somaliland Fr. Central Afr. R. Ivory Coast Reunion Sudan Chad Kenya Rhodesia Tanzania Congo Brazza. Liberia Malawi Togo Congo Dem. Rep. Libya Rwanda Tunisia Dahomey Madagascar Zambia Uganda Ethiopia Mali Zanzibar Upper Volta Gabon Mauritania Percentage computed from Direction of Trade data on exports, similar to computation made in variable 14 for imports. Percentage computed form Direction of Trade data on imports from what 293 classifies as "Soviet Areas," including the following countries: Albania Cuba Hungary Poland Bulgaria Czechoslovakia Korea North Rumania China Mainland Germany East Outer Mongolia USSR Viet Nam North Percentages computed from Direction of Trade data on exports to the "Soviet Areas," as defined in variable 16 computation. Computed as (variable #4 minus variable #3) divided by variable #20. Computed from yearly issues of the Foreign Service List, published by the U.S. Department of State. The 1966 figures were taken from the January 1967 issue, whereas the figures for other years were taken from the May issues of those years. Primary source: Yearbook of National Accounts Statistics, published by the United Nations. Cross—checked with data in other sources, including "Africa: This New Dialogue," Department of State Publication 8511 in the African Series #47 (May 1970) and Political Handbook and Atlas of the World, 1970, edited by Richard P. Stebbins and Alba Amoia, published for the Council on Foreign Relations by Simon and Schuster, New York, 1970. Gross national products are reported at factor cost rather than at market prices. GNP at factor cost is defined as the income of labor and property in the forms of wages, profits, interest, etc. earned from the productions of goods and services, with an allowance for capital depreciation. According to an explanatory note in the Yearbook of National Accounts Statistics, the factor cost calculation is used in order to minimize so far as possible any distorting effect arising out of differing national tax policies. Primary and secondary sources are as in variable #20, with the exception that Development Assistance, 1969 Review, published by the OECD, is an additional secondary source. GNP per capita is also calculated at factor cost rather than market cost. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 265 Calculated from tabulations of voting reported in the Official Records of the United Nations General Assembly and all Standing Committees of the General Assembly. Calculated from tabulations of roll—call voting reported in the Official Records of the United Nations. For each vote, the pre— ference of the majority of African states voting was ascertained. Then, for each African state in the study sample, the percentage of votes cast against the African majority, including abstentions, was calculated. Calculated from tabulations of roll—call voting reported in the Official Records of the United Nations. For each African state in the study sample, the percentage of votes cast against the vote of the United States was calculated, including abstentions. Calculated from tabulations of roll—call voting reported in the Official Records of the United Nations. For each African state in the study sample, the percentage of votes cast against the vote of the Soviet Union was calculated, including abstentions. Computed from variables 24 and 25 for each country in the sample. Source: annual publications of the U.S. Department of State: Educational and Cultural Diplomacy for 1964 and 1965, and International Exchange for 1967 and 1968. Source: same as for variable #27. Sources: annual editions of The Stateman's Yearbook and The Middle East and North Africa, Le Guid'Ougst Africain, West African Directory published by Thomas Skinner and Co. of Great Britain, and the West Africa Annual. Data for 1964 were cross—checked also with the data reported by Steven J. Brams in his "Transaction Flows in the International System," American Political Science Review, LX (December 1966). Same sources as for variable #29. Tabulated from membership lists published in annual editions of the Handbook of International Organizations. Tabulated and computed from membership lists published in annual editions of the Handbook of International Organizations, where "African Only" international governmental organizations are defined as IGO's which include only African states as members. Computed as the sum of the number of state members in all the international governmental organizations of which the particular country is a member. Computed for each country in the sample as variable #33 divided by variable #31. Tabulated from memberships lists published in annual editions of the Handbook of International Organizations, where the number of African members is summed for all IGO's of which the particular country is a member. Computed for each country in the sample as its score on variable #35 divided by its score on variable #31. Source: The Flow of Financial Resources to the Underdeveloped Countries, published annually by the Organization for Economic Cooperation and Development, Paris. Source: same as for variable #37. Source: same as for variable #37. Source: annual U.S. State Department estimates. 41. 42. 43. 44. 45. 46. 47. 48. 266 Taken from the lists of aid donors provided in the sources for variables #37 and #40. Sum of aid to a country from sources listed from variables #37, #39, and #40. Calculated for each country as its value on variable #42 divided by its value on variable #20. Source: Direction of Trade, annual volumes; percentage computed from raw data provided on imports from all countries. Source: Direction of Trade, annual volumes; percentage computed from raw data provided on exports to all countries. Computed from values for each country on variables #44 and #16. Computed from values for each country on variables #45 and #17. Computed from values for each country on variables #38 and #40. APPENDIX III SIZE, SOCIAL, & ECONOMIC CHARACTERISTICS OF AFRICAN STATES mmummnumpg H H H o 12 APPENDIX III SIZE, SOCIAL, & ECONOMIC CHARACTERISTICS OF AFRICAN STATES Population x 1000 56,400 (Nigeria) 22,200 Ethiopia 15,300 (Congo K.) 13,180 Sudan 12,959 Morocco 10,975 (Algeria) 10,325 Tanzania 9,104 Kenya 7,537 (Ghana) 7,367 Uganda 6,180 Malagasy R. 5,103 Cameroon 4,750 (Upper V.) 4,565 Tunisia 4,485 Mali 3,900 Malawi 3,750 Ivory C. 3,600 Zambia 3,420 Guinea 3,400 Senegal 3,300 Chad 3,237 Niger 3,018 Rwanda 2,780 (Burundi) 2,420 Somalia 2,300 (Dahomey) 2,240 (Sierra L.) 1,603 (Togo) 1,559 Libya 1.320 (CAR) 1,041 Liberia 900 Mauritania 826 Congo B. 459 Gabon Countries Ranked, Tied Scores Bracketed, Coup States in Parentheses Annual rate of Population growth it 3.7 Libya 3.5 Somalia .3 Ivory Coast E:;;Mger 3.1 Malagasy R. L:;:;handa 2.9 (Dahomey) 2.9 Kenya 2.9 Zambia Guinea 2.8 Malawi Morocco 2.7 (Ghana) 2.5 (Burundi) 2.5 Uganda 2.5 (Upper Volta) 2.3 Mali £:;:;enegal FETE‘Ybe> 3;2_Mauritania 2.1 Cameroon 2.1 (Congo K.) 2.1 (sierra L.) 2.0 (Nigeria) £:;:;unisia 1.9 Tanzania 1.8 Ethiopia 1.6 Congo Brazz. L;;;;abon 1.5 Chad 1.4 Liberia 0.9 (Algeria) 2ES7 Population density, per sq. mi. 126 Rwanda 120 (Burundi) 67 (Nigeria) 35 Malawi 34 (Ghana) 34 (Sierra L.) 34 Tunisia 32 Morocco 31 Tanzania 28 (Togo) 22 (Dahomey) 19 Ethiopia L;:;enegal 18 (Upper Volta) 17 Kenya 15 Guinea 13 Uganda 12 Cameroon Lg_lyory Coast 11 Malagasy R. 10 Liberia 7 (Congo K.) 6 Sudan 5 (Algeria) 5 Zambia 4 Mali L:;bmalia 3 Chad 3 Congo Brazz. 3 Niger 2 (CAR) [:gabon 1 Libya N 1 Mauritania Urbanization % 24.2 Morocco 19.0 Senegal 18.3 Libya 18.2 Tunisia 16.8 Zambia 15.4 Congo B. 14.1 (Algeria) 11.6 (Ghana) 11.4 (Nigeria) 11.0 Gabon 9.1 (Congo K.) 8.0 Liberia Ealagasy 7.2 (Sierra L.) 6.8 Ivory C. 6.2 Cameroon 5.9 Kenya 5.8 Somalia 5.5 (Dahomey) 5.0 Malawi 4.5 Sudan L:;:;Togo) 3.9 (CAR) SIS—Ethiopia 3.8 Tanzania 3.7 Guinea 2.3 (Upper V.) 2.1 (Burundi) 2.0 Niger 1.8 Mali 1.2 Uganda 1.0 Chad 0.4 Rwanda 0.0 Mauritania APPENDIX IV 1964 INTENSITY VARIABLES 268 SIZE, SOCIAL, & ECONOMIC CHARACTERISTICS OF AFRICAN STATES Countries Ranked, Tied Scores Bracketed, Land Area, Agricultural Gross GNP sq. miles and, acres National per Rank x 1000 per capita Product* capita** l 967 Sudan 39 Chad 4603 (Nigeria) 280 Gabon 2 920 (Algeria) 27 Somalia 3040 (Algeria) 230 (Algeria) 3 906 (Congo K.) 20 Mali 2545 Morocco 230 (Ghana) 4 679 Libya 20 Zambia 2492 (Ghana) 210 Libya 5 496 Chad l7 Libya 1720 (Congo K.) 200 Ivory C. 6 489 Niger 15 Malagasy R. 1561 Libya E80:fiberia fi 7 465 Mali 11 (CAR) 1482 Ethiopia 180 Tunisia if 8 457 Ethiopia L1_Niger 1400 Sudan 170 Morocco 9 419 Mauritania 10 (Algeria) 1114 Kenya 170 Senegal 10 386 Tanzania 10 Ethiopia 1014 Zambia 160 Zambia 11 357 (Nigeria) 10 Ivory Coast 1005 Ivory Coast 140 Congo 8. 12 288 Zambia 10 Tanzania 940 Tunisia 140 Mauritania 13 246 Somalia 9 Cameroon 830 Tanzania 120 (Sierra L.) 14 238 (CAR) 8 (Congo K.) 811 Senegal 110 Cameroon 15 230 Malagasy Rep. 6 (Sierra Leone) 734 Cameroon 100 (Nigeria) 16 225 Kenya L:;hdan 709 Uganda 95 Malagasy 17 183 Cameroon 5 Liberia 707 Malagasy R. L§_§udan 18 171 Morocco 4 Senegal 361 (Sierra L.) 90 (CAR) 19 132 Congo Brazza. 5_Ma1awi 358 Mali L;:;enya 20 125 Ivory Coast 3 Mauritania 295 Niger 85 (Togo) 21 106 (Upper Volta) 3 Morocco 289 Guinea 80 Uganda 22 103 Gabon 3 (Togo) 251 Chad 75 Niger 23 95 Guinea 3 Tunisia 245 (Upper Volta) 76_Chad 24 93 Uganda 3 Uganda 229 Liberia 70 (Congo K.) 25 92 (Ghana) §_lppper Volta) 209 (Togo) 70 (Dahomey) 26 76 Senegal P_Congo Brazza. 203 Malawi 70 Guinea 27 63 Tunisia 2 (Dahomey) 193 (Dahomey) Zgyganzania 28 46 Malawi 2 (Ghana) 187 (CAR) 65 Mali 29 45 (Dahomey) 2 Kenya 184 Gabon 50_?Burundi) 30 43 Liberia 3_§wanda 162 Congo Brazz. 50 Ethiopia 31 28 (Sierra Leone) 1 (Burundi) 154 (Burundi) 50 Rwanda 32 22 (Togo) 1 Gabon 143 Mauritania Momalia 33 ll (Burundi) 1 (Nigeria) 129 Somalia 45 (Upper v.) 34 10 Rwanda ? Guinea 128 Rwanda 4O Malawi *Gross National Product in millions of U.S. dollars, 1966. Coup States in Parentheses **GNP per capita in U.S. dollars, 1964. SIZE, SOCIAL, & ECONOMIC CHARACTERISTICS OF AFRICAN STATES Countries Ranked, Tied Scores Bracketed, Coup States in Parentheses Electric Power Inhabitants Students as % Rank per capita* per physician Literacy of 5-19 group 1 868 Zambia 76,230 (Upper V.) 50 (Congo K.) 49 Congo B. 2 339 Liberia 72,440 Chad 35 Malagasy Rep. 47 Zambia 3 322 (Ghana) 65,460 Niger Eunisia 45 Tunisia 4 254 Libya 62,430 Ethiopia 30 Libya 43 Cameroon 5 182 Cameroon 62,380 Rwanda 25 (Ghana) 2 Gabon 6 175 Gabon 56,320 (Burundi) 25 Kenya 2 Libya 7 167 (Congo K.) 49,200 Mali 25 Uganda 41 Kenya 8 130 Tunisia 46,900 Malawi 20 (Algeria) 4O Malawi 9 111 (Algeria) 44,620 (Nigeria) 20 Congo Braz. 39 (Congo K.) 10 110 Morocco 38,850 (CAR) 20 Ivory Coast 31 (Ghana) 11 107 Ivory Coast [30,000 Mauritania 20 (Nigeria) Eganda 12 77 Uganda 30,000 Somalia 14 Morocco 30 Malagasy . 13 73 Senegal 26,690 (Congo K.) mameroon Elligeria) 14 70 Congo Brazz. . 26,680 Cameroon 10 (CAR) 28 Ivory C. 15 64 (Sierra Leone) 24,590 Sudan 10 Gabon 27 Morocco 16 63 Kenya 22,110 (Togo) 10 Guinea 26 (Algeria) 17 53 Guinea 20,500 Guinea 10 Liberia 25 (Togo) 18 44 Mauritania 17,980 Ivory C. 10 Malawi 19 (Burundi) 19 38 (Togo) 17,000 Tanzania 10 Rwanda 19 (CAR) 20 35 Sudan 16,730 Senegal 10 Senegal 19 Rwanda 21 29 Malagasy Rep. 16,440 (Sierra L.) 10 (Sierra L.) 15 (Dahomey) 22 Eanzania 14,110 (Ghana) 10 Sudan 15 Liberia 23 27 (CAR) 12,120 Morocco 10 Tanzania 15 Senegal 24 Ealawi 11,720 Zambia 10 (T090) 15 Tanzania 25 22 (Nigeria) 11.640 Congo Braz. 1.0—Zambia 13 Guinea 26 18 Ethiopia 11,150 Liberia 5 Chad 12 Chad 27 12 (Dahomey) 10,600 Kenya 5 (Dahomey) 12 (Sierra L.) 28 11 Chad 10,370 Uganda 5 Ethiopia 12 Sudan 29 9 Niger 8,780 Tunisia 5 Mali . 6 Mali 30 7 Mali 8,550 (Algeria) 5 Mauritania 5 Mauritania . 31 7 Rwanda 5,860 Gabon 5 Niger 5 Somalia 32 7 Somalia 3,160 Libya 5 Somalia 5 (Upper V.) 33 5 (Burundi) ?? (Dahomey) 5 (Upper Volta) 4 Ethiopia 34 EUpper Volta) ?? Malagasy R. ? (Burundi) Niger *Kilowatt hours per year. APPENDIX IV 1964 Intensity Variables Data are Ranks #39 #2 #22 #29 #33 #42 #43 Total Algeria 33 24 34 30 20.5 34 31 206.5 Burundi 10 23 6 9.5 2.5 2 13 66 Cameroon 25 18 26.5 21 28 17 8 143.5 CAR 7 17 16 14 18 15 30 117 Chad 19 7 17 12 20.5 10 18.5 104 Congo Brazza 24 26 11.5 20 9 25 34 149.5 Congo Kinshasa 32 19.5 11.5 14 31 33 14 155 Dahomey 15 4 9 17.5 15 13 23 96.5 Ethiopia 21 6 22.5 31 8 12 2 102.5 Gabon 12 32 5 6.5 24 5 22 106.5 Ghana 34 16 31.5 34 27 29 9 180.5 Guinea 6 8 22.5 23.5 13 20 27 120 Ivory Coast 26 29 26.5 22 14 22 10 149.5 Kenya 3 12 2 2.5 5.5 32 28 85 Liberia 5 31 28.5 28.5 22 11 20 146 Libya 2 33 19 23.5 19 9 4 109.5 Malagasy Rep. 31 11 25 6.5 25 24 18.5 141 Malawi 4 22 2 4.5 2.5 18 32 85 Mali 28 2 31.5 25.5 17 19 24 147 Mauritania 13 30 22.5 9.5 26 4 25.5 130.5 Morocco 17 15 19 33 34 31 7 156 Niger 11 5 19 17.5 23 6 11 92.5 Nigeria 30 13 30 25.5 29 26 1 154.5 Rwanda 8 l 4 9.5 2.5 l 12 38 Senegal 27 14 22.5 28.5 30 27 21 170 Sierra Leone 9 27 13.5 17.5 11 3 6 87 Somalia 18 21 7 14.0 10 23 33 126 Sudan 29 10 15 27.0 32 16 3 132 Tanzania 23 25 13.5 4.5 12 30 29 137 Togo 14 19.5 28.5 9.5 7 7 25.5 111 271 #39 #2 #22 #29 #33 #42 #43 Total Tunisia 22 9 33 32 33 gg_ 17 174 Uganda 16 28 8 2.5 5.5 21 16 97 UAR -— -— —— -— -— —- —- —— Upper Volta 20 3 10 17.5 16 8 15 89.5 Zambia 1 34 2 1.0 2.5 14 5 59.5 Low Rank indicates low intensity. APPENDIX V 1967 INTENSITY VARIABLES APPENDIX V 1967 Intensity Variables Data are Ranks #39 #2 #22 #29 #33 #42 #43 Total Algeria C 1 26 29 31.5 29 32 11 159.5 Burundi C 14 19 3 3.5 11 4 19.5 74 Cameroon 31 20 25.5 17.5 28 23 15 160 CAR C 18 16 7 3.5 13.5 13 30 101 Chad l9 8 9.5 16 15 14 23 104.5 Congo Brazza 29 27 4 3.5 12 15 32 122.5 Congo Kinshasa C 9.5 29 18 29 25 31 18 159.5 Dahomey C 22 6 16 11.5 13.5 12 27 108 Ethiopia 25 4 32.5 30 2 l8 2 113.5 Gabon 12 32 1.5 3.5 21 5 19.5 94.5 Ghana C 3 1 32.5 34 20 28 10 128.5 Guinea 8 9 12.5 28 3 3 8 71.5 Ivory Coast 26 28 9.5 22 31 20 9 145.5 Kenya 33 13 32.5 9 23 27 13 150.5 Liberia 11 34 25.5 23.5 4 24 34 156 Libya 9.5 33 12.5 23.5 5 1 1 85,5 Malagasy Rep. 30 12 19 3.5 30 25 16 135.5 Malawi - 6 23 1.5 7 10 19 33 99.5 Mali 28 3 5 21 8 10 14 89 Mauritania 4 30 32.5 14 16 8.5 26 131 Morocco 27 18 21 31.5 33 29 6 165.5 Niger 23 7 21 17.5 22 17 21 128.5 Nigeria C 34 10.5 29 25 32 34 4.5 169 Rwanda 13 2 6 3.5 6 6 28 64.5 Senegal 32 21 16 33 24 26 17 169 Sierra Leone C 7 22 23.5 9 18.5 2 3 85 Somalia 17 14 8 19 l 8.5 31 98.5 Sudan 24 10.5 29 26.5 27 21 4.5 142.5 Tanzania 21 25 21 20 17 22 12 138 Togo C 16 17 16 9 9 7 22 96 273 #39 #2 #22 #29 #33 #42 #43 Total Tunisia 20 15 26.5 26.5 34 33 227 184 Uganda 5 24 12.5 14 18.5 16 7 97 UAR -- -- -— —- -- —— -- -- Upper Volta C 15 5 23.5 14 26 11 24.5 119 Zambia 2 31 12.5 11.5 7 30 24.5 118.5 APPENDIX VI 1964 EXTENSITY VARIABLES APPENDIX VI 1964 Extensity Variables Data are Ranks #7 #10 #11 #29 #34 #41 Total Algeria 30.5 2 3 30 31.5 2 99 Burundi 3 20 2 9.5 2.5 32 69 Cameroon 19.5 5 7 21 23.5 17 93 CAR 10.5 8 13 14 23.5 27 96 Chad 5.5 13 8 12 31.5 21.5 91.5 Congo Brazza 14.5 10 3O 20 11 13 98.5 Congo Kinshasa 24 27 17 14 34 3 119 Dahomey 8 4 4 17.5 18 19.5 71 Ethiopia 23 34 12 31 9 27 136 Gabon 14.5 6 11 6.5 30 30 98 Ghana 27 29 29 34 16 15 150 Guinea 17.5 21 32 23.5 8 13 115 Ivory Coast 25 3 21 22 17 ll 99 Kenya 33 25 31 2.5 5.5 5 102 Liberia 3 15 15 28.5 21 19.5 102 Libya 21 31.5 23 23.5 15 24 134 Malagasy Reg. 19.5 1 9 6.5 22 9 67 Malawi 14.5 17 15 4.5 2.5 10 63.5 Mali 12 26 28 25.5 13 29 133.5 Mauritania 8 12 27 9.5 26 25 107.5 Morocco 32 18 19 33 25 4 131 Niger 8 9 6 17.5 28 21.5 90 Nigeria 34 24 20 25.5 20 13 136.5 Rwanda 1 35 (1) 34 (l) 9.5 2.5 35 50 Senegal 22 7 1 28.5 33 8 99.5 Sierra Leone 10.5 19 5 17.5 10 33 95 Somalia 5.5 30 15 14 14 31 109.5 Sudan 30.5 31.5 33 27 27 27 176 Tanzania 28.5 23 25 4.5 7 6.5 94.5 Togo 14.5 28 18 9.5 12 34 116 275 #7 #10 #11 #29 #34 #41 Total Tunisia 26 14 10 32 29 6.5 117.5 Uganda 28.5 22 26 2.5 5.5 18 102.5 UAR -- -~ -— -- -— —— -- Upper Volta 3 11 24 17.5 19 23 97.5 Zambia 17.5 16 22 1.0 2.5 16 75 Low rank indicates low extensity. APPENDIX VII 1967 EXTENSITY VARIABLES APPENDIX VII 1967 Extensity Variables Data are Ranks #7 #10 #11 #29 #34 #41 Total trade trade trade diplo diplo aid Algeria 26 l 6 31.5 22 1 87.5 Burundi l 31 1 3.5 31 29 96.5 Cameroon 19.5 4 15 17.5 l4 14 84 CAR 11.5 6 12 3.5 6.5 15 54.5 Chad 7.5 13 7 16 5 20 68.5 Congo Brazza l4 9 31 3.5 2 22 81.5 Congo Kinshasa 22.5 32 10 29 25 2 120.5 Dahomey 7.5 11 33.5 11.5 6.5 23 93 Ethiopia 25 34 11 3O 26 18 144 Gabon 16.5 5 18 3.5 21 27 91 Ghana 28 21 26 34 9 7 125 Guinea 16.5 18 24 28 18.5 28 133 Ivory Coast 19.5 8 16 22 20 11 96.5 Kenya 33.5 19 27 9 ll 13 112.5 Liberia l4 2 21 23.5 34 5 99.5 Libya 24 26 29 23.5 1 34 137.5 Malagasy Rep. 18 3 ' 17 3.5 33 10 84.5 Malawi 7.5 27.5 8 7.0 12 12 74 Mali 7.5 22 20 21 10 24 104.5 Mauritania 10 14 30 l4 17 25 110 Morocco 32 16 13 31.5 23.5 8 124 Niger 2.5 12 4 17.5 13 19 68 Nigeria 33.5 24.5 23 '25 29 6 141 Rwanda 2.5 29 33.5 3.5 18.5 26 113 Senegal 22.5 10 3 33 16 9 93.5 Sierra Leone 11.5 27.5 5 9 27 33 113 Somalia 4 15 2 19 32 30 102 Sudan 30.5 33 32 26.5 30 32 184 Tanzania 30.5 24.5 22 20 4 21 122 T090 14 23 14 9 3 31 94 277 #7 #10 #11 #29 #34 #41 Total trade trade trade diplo diplo aid Tunisia 27 20 25 26.5 28 3 129.5 Uganda 29 17 28 14 8 16 112 UAR -- -— -- —— -- -- —- Ugger Volta 5 7 9 14 15 17 67 Zambia 21 3O 19 11.5 23.5 4 109 Low rank indicates low extensity. APPENDIX,VIII 1964 ALIGNMENT VARIABLES APPENDIX VIII 1964 Alignment Variables Data are Ran s Africa USA-USSR Africa USA-USSR Total #14 #15 #23 #26 #32 #35 #46 #47 #48 Africa USA— USSR Algeria 15 12 l 33 12 ll 15 9 34 51 91 Burundi 3O 3 9 27.5 2.5 2.5 23 34 2.5 47 87 Cameroon 22 24 27 9 24 32 20 27 17 129 73 19 5 13 11 26.5 25.5 10 31 17 89 68 Chad 23 31 32 21.5 21 23 12 2.5 7 130 43 Congo Brazza 21 16 28.5 16.5 32.5 12 6 12 26 110 60.5 Congo Kinshasa 17 9.5 22 10 15 31 33 21 27 94.5 91 Dahomey 28 25 30 19 26.5 17 9 8 7 126.5 43 Ethiopia 5.5 22 6 25 32.5 8 27 33 30.5 74 115.5 Gabon 16 18 19 7 15 27 28 24 13 95 72 Ghana 24 7.5 14.5 30 11 14 24 22 33 71 109 Guinea 11 30 16.5 27.5 30 15.5 26 18 32 103 103.5 Ivory Coast 20 26 24 8 26.5 18 25 25 20 114.5 78 Kenya . 8 28 33.5 2.5 9.5 5.5 16 19 28 84.5 65.5 Liberia 1 1.5 3 6 18.5 21 5 34 30 21 45.5 91 ' 7 9.5 14.5 19 5.5 9 32 14 19 45.5 84 Malagasy Rep. 9 29 5 2.5 18.5 30 13 29 7 91.5 51 5 Malawi 33 4 33 5 2.5 2.5 2.5 2 15 13 75.5 32.5 Mali 27 33 4 34 29 20 17 32 30.5 113 113.5 Mauritania 5.5 14.5 21 16.5 18.5 28.5 31 5 2.5 88 55 Morocco 14 19 10 32 7 34 30 23 24 84 109 Niger 29 32 26 13 22.5 25.5 11 2.5 7 135 33.5 Nigeria 4 17 8 21.5 13 24 21 17 25 66 84.5 Rwanda 34 1.5 20 5 2.5 2.5 2 2.5 2.5 60.5 12 Senegal 18 23 16.5 12 18.5 28.5 19 ll 13 104. 55 Sierra Leone 10 7.5 18 14 15 10 8 26 17 60.5 65 . Somalia 26 6 23 23 32.5 13 29 13 29 100.5 94 13 11 7 24 8 21.5 18 20 22 60.5 84 Tanzania 2 14.5 31 31 22.5 19 14 10 28 89 83 T090 25 13 28.5 15 32.5 7 7 7 13 106 42 Tunisia 12 27 11 26 5.5 33 22 16 13 88 5 7; Uganda 3 20 25 29 9.5 5.5 5 28 23 63 8 92139: Volta 31 34 12 19 26.5 15.5 2 2.5 7 5 1:9: 5 :g-5 Zambia 32 21 33.5 2.5 2.5 2.5 4 6 2. Low rank indicates low alignment. 278 APPENDIX IX 1967 ALIGNMENT VARIABLES 1967 APPENDIX IX Alignment Variables Data are Ranks 279 Africa USA—USSR Total #14 #15 #23 #32 #35 #26 #46 #47 #48 Africa USA- trade trade diplo diplo diplo diplo trade trade aid USSR Algeria C 12 14 20 11 18 31 11 13 34 75 89 Burundi C 31 2 28 17.5 7 33 31 34 4.5 85.5 102.5 Cameroon 21 17 34 22.5 33 17 6 28 15 127.5 66 CAR 20 8.5 9 32.5 20 19 4 32 14 90 69 23 30 25 28.5 21 20 12 19 8.5 127.5 59.5 Congo Brazza 22 22 35 32.5 19 30 21 11 23 130.5 85 Congo Kinshasa C 17 7 33 l9 16 18 32 20 22 92 92 Dahomey C 27 2 15 32.5 25 9 18 2.5 4.5 101.5 34 Ethiopia 4.5 18 18.5 1 5 5 12 27 33 31 63.5 103 Gabon 19 21 3.5 22 5 22 7 20 24 4.5 88 55.5 Ghana C 13 11 2 7 . 5 10 8 22 16 33 43 . 5 79 Guinea 26 30.5 20 4 28.5 24 18 32 86.5 102.5 Ivory Coast 24 24.5 21 34 5 16 26 1.5 109.5 48.5 enya 7.5 28 18.5 13 24 15 15 17 27 91 74 Liberia 5.5 3.5 9 3 4 33 31 24 22 92 Libya 11 4 30.5 1 1 28.5 26 12 1.5 47.5 68 Malagasy Rep. 18 31 8 16 26 6 9 30 4.5 99 49.5 Malawi 32 23 1 10 10 l 7 15 12 76 35 29 34 24 26.5 9 34 34 25 30 122.5 123 Mauritania 10 15 29 26 5 23 23.5 19 7 12 103.5 61.5 Morocco 14 16 21 6 29 32 13 23 26 86 94 Niger 30 32 10 28.5 28 10 23 2 5 8.5 128.5 44 Nigeria C 4.5 12 12.5 4 32 13.5 14 14 19 65 60.5 Rwanda 33 13 16 3O 8 11 l 2.5 8.5 100 23 Senegal 28 29 32 24 27 16 10 5 16 140 47 Sierra Leone C 15. 8.5 7 12 13 3 3 27 20 55.5 53 Somalia l6 2 22 14.5 2 21 29 10 29 56.5 89 Sudan 7.5 5.5 17 4 12 26 25 21 25 46 97 ‘ 1 Tanzania 2 27 14 14 5 15 25 5 6 28 72.5 64 ‘Togo c 25 20 26 32.5 14 27 17 a 8.5 117.5 60.5 'hmisia 9 24.5 12.5 2 31 22 28 22 17 79 89 Uganda 3 19 5 7.5 17 2 8 29 21 51.5 60 UAR —— —— —— -— -- —— —— —— -— -— —_ UJ’per Volta C 34 33 27 25 30 13.5 2 2.5 12 149 30 5 Zambia 26 10 11 4 6 23 . 5 6 9 18 57 56. low rank indicates low alignment. BIBLIOGRAPHY BIBLIOGRAPHY Adelman, Irma, and Cynthia Taft Morris. 1967. Society, Politics, and Economic Development. Baltimore: The John Hopkins Press. Adelphi Paper 27. London: Institute for Strategic Studies. Alba, Victor. 1962. "The Stages of Militarism in Latin America," in John J. Johnson, ed., The Role of the Military in Under— developed Countries. Princeton: Princeton University Press. Alker, Hayward. 1964. "Regionalism vs. Universalism," in Bruce Russett et a1. World Handbook of Political and Social Indicators. New Haven: Yale University Press. Alker, Hayward, and Bruce Russett. 1965. World Politics in The General Assembly. New Haven: Yale University Press. Allison, Graham T. 1971. Essence of Decision: Explaining the Cuban. Boston: Little, Brown and Company. Almond, Gabriel A. 1950. The American People and Foreign Policy. New York: Harcourt Brace and World. Almond, Gabriel A., and Sidney Verba. 1963. The Civic Culture. Princeton: Princeton University Press. Apter, David E. 1965. The Politics of Modernization. Chicago: University of Chicago Press. Banks, Arthur 5., and Robert B. Textor. 1963. A Cross—Polity Survey. Cambridge: The M.I.T. Press. Beck, Carl, and James Malloy. 1971. Political Elites: A Mode of Analysis. Pittsburgh: University Center for International Studies, University of Pittsburgh. Bendix, Reinhard. 1949. Higher Civil Servants in American Society. Boulder: University of Colorado Press. 280 281 Bienen, Henry, ed. 1968. The Military Intervenes: Case Studies in Political Development. New York: Russell Sage Foundation. Black, Cyril E. 1966. Dynamics of Modernization: A Study in Compara— tive History. New York: Harper and Row. Blalock, Hubert M., Jr. 1972. Social Statistics (Second Edition). New York: McGraw-Hill Book Company. Brecher, Michael; Blema Steinberg; and Janice Stein. 1969. “A Framework for Research on Foreign Policy Behavior,“ Journal of Conflict Resolution, Vol. 13, pp. 75-101. Bretton, Henry L. 1973. Power and Politics in Africa. Chicago: Aldine Publishing Company. Brice, Belmont. 1966. "The Nature and Role of the Military in Sub-Saharan Africa," African Forum, Vol. (Summer), pp. 57—67. Campbell, Donald T., and Julian C. Stanley. 1963. Experimental and Quasi-Experimental Designs for Research. Chicago: Rand McNally and Company. Collins, John Norman. 1971. Foreign Conflict Behavior and Domestic Disorder in Africa. Syracuse: Eastern Africa Studies IV, Maxwell School of Citizenship and Public Affairs, Syracuse University. Cowan, L. Gray. 1966. "Political Determinants," in Vernon McKay, ed. African Diplomacy: Studies in the Determinants of Foreign Policy. New York: Frederick A. Praeger, Publishers. Cowan, L. Gray. 1968. The Dilemmas of African Independence (revised edition). New York: Walker and Company. Deutsch, Karl W. 1966. "External Influences on the Internal Behavior of States," in R. Barry Farrell, ed. Approaches to Comparative and International Politics. Evanston: Northwestern University Press. Direction of Trade. Various Annual Volumes. Dodge, Dorothy. 1967. "African Voting Cohesion in the United Nations," Africa Report (October), pp. 58—61. Easton, David. 1953. The Political System. New York: Knopf. Easton, David. 1957. "An Approach to the Analysis of Political Systems," in World Politics, Vol. 9, No. 3 (April), pp. 383—400. Eckstein, Harry, and David Apter. 1963. Comparative Politics: A Reader. New York: The Free Press of Glencoe. 282 Farrell, R. Barry, ed. 1966. Approaches to Comparative and Inter— national Politics. Evanston: Northwestern University Press. Feit, Edward. 1968. "Military Coups and Political Development: Some Lessons from Ghana and Nigeria,“ World Politics, Vol. 20, No. 2 (January), pp. 179-93. Finer, Samuel Edward. 1962. The Man on Horseback: The Role of the Military in Politics. New-York: Frederick A. Praeger, Publishers. Fisher, Roger. 1969. International Conflict for Beginners. New York: Harper and Row. Foltz, William J. 1966. “Military Influences," in Vernon McKay, ed. African Diplomacy: Studies in the Determinants of Foreign Policy. New York: Frederick A. Praeger, Publishers. Foltz, William. 1966. Review of Edward Luttwak's Coup d'Etat: 5 Practical Handbook, Africa Report (November), pp. 34-35. Frankel, Joseph. 1963. The Making of Foreign Policy: An Analysis of Decision-Making. London: Oxford University Press. Geertz, Clifford, ed. 1963. Old Societies and New States. New York: The Free Press of Glencoe. Glickman, Harvey. 1966. "The Military in African Politics: A Bib- liographic Essay," African Forum (Summer), pp. 68—75. Good, Robert C. 1962. "State—Building as a Determinant of Foreign Policy in the New States," in Laurence W. Martin, ed. Neutralism and Nonaligment. New York: Frederick A. Praeger; Greene, Fred. 1966. "Toward Understanding Military Coups," Africa Report (February), pp. 10-14. - Grundy, Kenneth W. 1968. "The Negative Image of Africa‘s Military," The Review of Politics, Vol. 30, No. 4 (October), pp. 428-39. Gutteridge, William F. 1965. "Education of Military Leadership in Emergent States,“ in James Coleman, ed. Education and Political Development. Princeton: Princeton University Press. Gutteridge, William F. 1969. The Military in African Politics. London: Metheun. Gutteridge, William F. 1970. "Why Does an African Army Take Power?“ Africa Report (October), pp. 18-21. 283 Hanrieder, Wolfram. 1967. “Compatibility and Consensus: A Proposal for the Conceptual Linkage of External and Internal Dimensions of Foreign Policy," American Political Science Review, Vol. LXI, No. 4 (December), pp. 971-82. Hermann, Charles F. 1969. “International Crisis as a Situational Variable," in James N. Rosenau, ed. International Politics and Foreign Policy: A Reader in Research and Theory (revised edition). New York: The Free Press. Hoffmann, Stanley. 1960. Contemporary Theory in International Relations. Englewood Cliffs, New Jersey: Prentice-Hall. Horowitz, Irving Louis. 1966. Three Worlds of Development: The Theory and Practice of International Stratification. New York: Oxford University Press. Hunter, Floyd. 1953. Community Power Structure. Chapel Hill: University of North Carolina Press. Hunter, Floyd. 1959. Top Leadership, U.S.A. Chapel Hill: University of North Carolina Press. Huntington, Samuel P. 1957. The Soldier and the States: The Theory of Politics of Civil-Military Relations. Cambridge: Belknap Press of Harvard University Press. Jacobson, Harold, and Zimmerman, Williams. 1969. The Shaping of Foreign Policy. New York: Atherton Press. Janowitz, Morris. 1956. "Social Stratification and the Comparative Analysis of Elites," Social Forces, Vol. 35, No. 1 (October), pp. 81—95. Janowitz, Morris. 1964. The Military in the Political Development of New Nations. Chicago: University of Chicago Press. Jervis, Robert. 1968. "Hypotheses on Misperception," World Politics, Vol. XX, pp. 454—79. Johnson, John J. 1964. The Military and Society in Latin America. Stanford: Stanford University Press. Kautsky, John H. 1962. Political Change in Underdeveloped Countries. New York: John Wiley and Son. Kornhauser, William. 1959. The Politics of Mass Society. Glencoe: The Free Press. Lasswell, Harold; David Lerner; and C. Easton Rothwell. 1952. The Comparative Study of Elites. Stanford: Stanford University Press. r—wmi -- 284 pg Guid'Ouest Africain. Paris: Havas International Regis. Lee, J. M. 1969. African Armies and Civil Order. New York: Frederick A. Praeger, Publishers. London, Kurt. 1965. The Making of Foreign Policy, East and West. Philadelphia: Lippincott. Luckham, Robin. 1971a. "A Comparative Typology of Civil—Military Relations,“ Government and Opposition (Winter), pp. 5-35. Luckham, Robin. 1971b. The Nigerian Military: A Sociological Analysis of Authoripy and Revolt, 1960-67. Cambridge: University Press. Matthews, Donald R. 1954. The Social Background of Political Decision Makers. Garden City: Doubleday. Matthews, Donald R. 1960. U.S. Senators and Their World. Chapel Hill: University of North Carolina Press. Mazrui, Ali. 1969. "Anti-Militarism and Political Militancy in Tanzania," in Mazrui. Violence and Thought. London: Harlow Longmans. McGowan, Patrick. 1967. "Factors and Correlations in African Foreign Policy." Paper delivered at 1967 Annual Meeting of the African Studies Association of the United States, Hilton Hotel, New York, November 1-4. McGowan, Patrick J. 1968. "Africa and Nonalignment: A Comparative Study of Foreign Policy," International Studies Quarterly, Vol. 12, No. 3 (September), pp. 262—95. McKay, Vernon. 1966. African Diplomacy: Studies in the Determinants of Foreign Policy. New York: Frederick A. Praeger, Publishers. Mills, C. Wright. 1945. "The American Business Elite: A Collective Portrait," Journal of Economic History, Supplement 5 (December), pp. 20-44. Mills, C. Wright. 1956. The Power Elite. New York: Oxford University Press. Morgenthau, Hans J. 1972. Politics Among Nations (Fifth Edition). New York: Alfred A. Knopf. Morrison, Donald G., and Hugh M. Stevenson. "Integration and Insta— bility: Patterns of African Political Development," American Political Science Review, Vol. 66, No. 3 (September), pp. 902—27. 285 Morrison, Donald G., and Hugh M. Stevenson. 1971. "Political Instability in Independent Black Africa: More Dimensions of Conflict Behavior Within Nations," Journal of Conflict Resolution, Vol. XV (September), pp. 347-68. Nordlinger, Eric. 1970. "Soldiers in Mufti," American Political Science Review, Vol. LXIV, No. 4 (December), pp. 1131—48. Organization for Economic Cooperation and Development, Development Assistance, 1969 Review. Paris: OECD. Organization for Economic Cooperation and Development. Geographical Distribution of Financial Flows to the Less Developed Countries. Annual editions. Paris: OECD. Organski, A. F. K. 1965. The Stages of Political Development. New York: Alfred Knopf. Pareto, Vilfredo. 1963. The Mind and Society, translated by Andrew Bongiorno and Arthur Livingston, edited by Arthur Livingston. New York: Dover Publications, Inc. Potholm, Christian P. 1970. Four African Political Systems. Englewood Cliffs, New Jersey: Prentice-Hall, Inc. Przeworski, Adam, and Henry Teune. 1970. The Logic of Comparative Social Inguiry. New York: Wiley-Interscience, A Division of John Wiley and Son. Pye, Lucian. 1962. “Armies in the Process of Political Modernization," in John J. Johnson, ed. The Role of the Military in Under- developed Countries. Princeton: New Jersey's Princeton University Press. Rapoport, D. 1963. "A Comparative Theory of Military and Political Types," in Samuel P. Huntington, ed. Changing Patterns g: Military Politics. New York. Riggs, Fred W. 1964. Administration in Developing Countries: The Theory of Prismatic Society. Boston: Houghton Mifflin. Rosenau, James N. 1966. "Pretheories and Theories in R. Barry Farrell, ed. Approaches to Comparative and International Politics. Evanston: Northwestern University Press. Rosenau, James N. 1967. "Compatibility, Consensus, and an Emerging Political Science of Adaptation," American Political Science Review, Vol. LXI, No. 4 (December), pp. 983—88. Rosenau, James N. 1970. "The Adaptation of National Societies: A Theory of Political System Behavior and Transformation." New York: McCaleb-Seiler Publishing Company. 286 Rostow, Walt Whitman. 1960. The Process of Economic Growth. Oxford (England): Clarendon Press. Rubin, Leslie, and Brian Weinstein. 1974. Introduction to African Politics: A Continental Approach. New York and Washington: Praeger Publishers. Rummel, Rudolph. 1966a. "Dimensions of Conflict Behavior Within Nations, 1946-59,“ The Journal of Conflict Resolution, Vol. X (March), PP. 65-73. Rummel, Rudolph J. 1966b. "The Dimensionality of Nation Project," in Richard L. Merritt and Stein Rokkan, eds. Comparing Nations. New Haven: Yale University Press. Rummel, Rudolph J. 1969. "Indicators of Cross—National and Inter— national Patterns," American Political Science Review, Vol. (March), pp. 127—149. Russett, Bruce, et a1. 1964. The World Handbook of Political and Social Indicators. New Haven: Yale University Press. Russett, Bruce. 1967. International Regions and the International System: A Study in Political Ecology. Chicago: Rand McNally. Scipio (pseud.). 1965. Emergent Africa. Boston: Houghton Mifflin. Shils, Edward. 1962. "The Military in the Political Development of the New States," in John J. Johnson, ed. The Role of the Military in Underdeveloped Countries. Princeton: Princeton University Press. Siegel, Sidney. 1956. Non—Parametric Statistics for the Behavioral Sciences. New York: McGraw—Hill Book Company. Singer, J. David. 1961. "The Level-of—Analysis Problem in Inter— national Relations," in Klaus Knorr and Sidney Verba, eds. The International System: Theoretical Essays. Princeton: Princeton University Press. Singer, J. David. 1972. "The Scientific Study of Politics: An Approach to Foreign Policy Analysis.“ General Learning Press, Morriston, New Jersey. Slessor, John. 1962. Forward to Gutteridge, William F. Armed Forces in New States. Oxford: Oxford University Press. Snyder, Richard C.; H. W. Bruck; and Burton Sapin. 1962. Foreign Policy Decision—Making: An Approach to the Study of Inter— national Politics. New York: The Free Press. 287 Snyder, Richard C, and Edgar Furniss. 1954. American Foreign Policy. New York: Rinehart. Spiro, Herbert J. 1966. Africa: The Primacy of Politics. New York: Random House. Stebbing, Richard P., and Alba Amoia, eds. 1970. Political Handbook and Atlas of the World. New York: Simon and Schuster, for the Council on Foreign Relations. Stinchcombe, Arthur. 1968. Constructing Social Theories. New York: Harcourt Brace and World. Stolper, Wolfgang F. 1966. Planning Without Facts: Lessons in Resource Allocation from Nigeria's Development. Cambridge: Harvard University Press. Tanter, Raymond. 1966. "Dimensions of Conflict Behavior, 1958—1960," The Journal of Conflict Resolution, Vol. X (March), pp. 41-64. Taylor, Charles Lewis, and Michael C. Hudson. 1972. World Handbook of Political and Social Indicators (Second Edition). New Haven and London: Yale University Press. The Middle East and North Africa. Various annual editions. London: Europa Publications, Ltd. The Statesman‘s Yearbook. Various annual editions. London: Macmillan London, Ltd. Thorne, C. T., Jr. 1966. "External Political Pressure, in Vernon McKay, ed. African Diplomacy: Studies in the Determinants of Foreign Policy. New York: Frederick A. Praeger, Publishers. United Nations. Official Records. Various annual editions. New York. United Nations. United Nations Statistical Yearbook. Various annual editions. New York. United Nations. United Nations Yearbook of National Accounts Statis— tics. Various annual editions. New York. United States Agency for International Development. 1971. Selected Economic Data for the Less Developed Countries. Washing- ton, D.C.: Office of Statistics and Reports, U.S.A.I.D. United States Arms Control and Disarmament Agency. 1969. World Military Expenditures. Washington, D.C. United States Department of State. 1970. Africa: This New Dialogue. Department of State Publications 8511 in the African Series 47. Washington, D.C. 288 United States Department of State. 1964 and 1965. Educational And Cultural Diplomacy. Annual editions. Washington, D.C. United States Department of State. Various annual editions. Foreign Service List. Washington, D.C. United States Department of State. 1967 and 1968. International Exchange. Annual editions. Washington, D.C. Van den Berghe, Pierre. 1965. "The Role of the Army in Contemporary Africa," Africa Report (March), pp. 12-18. Van den Berghe, Pierre. 1970. “The Military and Political Change in Africa," in Claude Welch, ed. Soldier and State in Africa. Evanston: Northwestern University Press. Waltz, Kenneth. 1959. Man, The State, and War. New York: Columbia University Press. Welch, Claude E., Jr., ed. 1970. Soldier and State in Africa: A Comparative Analysis of Military Intervention and Political Change. Evanston: Northwestern University Press. Welch, Claude E., Jr. 1973. "Radical and Conservative Military Regimes: A Typology and Analysis of Post-Coup Government in Tropical Africa." Paper delivered at Annual Meeting of the American Political Science Association, Jung Hotel, New Orleans, Louisiana, September 4—8, 1973. Welch, Claude E., Jr., and Arthur K. Smith. 1973. Military Role and Rule. North Scituate: Duxbury Press. West Africa Annual. Various annual editions. Lagos: John West Publications, Ltd. West African Directory. Various annual editions. London: Thomas Skinner and Co. of Great Britain. Wright, Quincy. 1955. The Study of International Relations. New York: Appleton-Century—Crofts. Yearbook of International Organizations. Various annual editions. Brussels: Union of International Associations. Zartman, I. William. 1966a. "National Interest and Ideology," in Vernon McKay, ed. African Diplomacy. New York: Frederick A. Praeger, Publishers. Zartman, I. William. 1966b. International Relations in the New Africa. Englewood Cliffs: Prentice—Hall, Inc. 289 Zolberg, Aristide. 1966. Creating Political Order: The Party—States of West Africa. Chicago: Rand McNally. Zolberg, Aristide. 1968. "Military Intervention in the New States of Tropical Africa: Elements of Comparative Analysis," in Henry Bienen, ed. The Military Intervenes. New York: Russell Sage Foundation. "HHMMLIN