The authors report a Q-factor analysis of 115 nations and 68 variables for the grouping of nations on the basis of their political characteristics. Five factors emerge to distinguish major political groupings of the con temporary world: polarchy, elitism, centrism, personalism, and tradi tionalism. Professor Banks is coauthor with R. B. Textor of A Cross-Polity Survey. Phillip M. Gregg is a doctoral candidate in Political Science at Indiana University.
Get full access to this article
View all access options for this article.
References
1.
Bruce M. Russett, "Delineating International Regions" (Carnegie-IDRC Joint Study Group on Measurement Problems, Paper Number G60, Indiana University , February, 1965). A revised version of this paper is to appear in J. David Singer, ed., Empirical Studies in International Relations (New York : Free Press, forthcom ng).
2.
Bruce M. Russett, Hayward R. Alker, Jr., Karl W. Deutsch, and Harold D. Lasswell , World Handbook of Political and Social Indicators (New Haven: Yale University Press , 1964). Despite the focus implied by the title, most of the data series in this collection are of limited political significance.
3.
Arthur S. Banks and Robert B. Textor , A Cross-Polity Survey ( Cam-bridge : The M.I.T. Press, 1963 ) A somewhat cruder analysis, based on a non-factorial design and embracing all CPS variables is included in Arthur S. Banks, "A Cross-Polity Survey: Preliminary Analysis," Paper delivered at the 1964 Annual Meeting of the American Political Science Association.
4.
The only independent nations not included are China, Gambia, Mali, Western Samoa, Zambia, and Zanzibar (now part of Tanzania).
5.
Harold E. Driver and Karl F. Schuessler, "Factor Analysis of Ethnographic Data," American ArthropologistLIX (August, 1957), 659
6.
. For an analysis of the Survey data based on the R-technique, see Philip M. Gregg and Arthur S. Banks, "Dimensions of Polit cal Systems: Factor Analysis of A Cross-Polity Survey, American Political Science ReviewLIX (September, 1965).
7.
The numbers in parentheses in Table 1 indicate the Suricy raw chaiacteristics (nominal and ordinal polychotomous variables) from which the present variables are derived. Thus Variables 10-12 of Table 1 are all derived from Survey Raw Characteristic 21 ("Former Colonial Ruler").
8.
Less than ten percent of the 115 nations received substantive scores on these variables.
9.
Rudolph J. Rummel , "The Dimensions of Conflict Behavior Within and Between Nations," General Systems YearbookVIII (1963), 1-50.
10.
The calculating formula employed was Pearson's product moment correlation coefficient. When used to intercorrelate ordinal and nominal variables, as in the present study, the formula is the equivalent of Spearman's rank-order correlation, rho, and the phi coefficient, respectively.
11.
See G. Udny Yule and M.G. Kendall, An Introduction to the Theory of Statistics (14th ed.; London: Charles Griffin, 1950), pp. 261, 271.
12.
Calculat ons were performed by the Indiana Research Computing Center's IBM 709. The MESA-3 program employed was developed by John B. Carroll at Harvard, coded by R A Sandsmark at Northwestern, and revised by Norman Swartz with the assistance of Gary Flint at Indiana.
13.
For a discussion of Hotelling's iterative technique, see Harry H. Harman, Modern Factor Analysis (Chicago: University of Chicago Press, 1960), p 154, ff.
14.
Henry F. Kaiser , "The Varimax Criterion for Analytic Rotation in Factor Analysis," PsychometrikaXXIII (September, 1958), 187-200.
15.
Henry F. Kaiser , "The Application of Electronic Computers to Factor Analysis," Educational and Psychological MeasurementXX (1960), 141-151
16.
; Raymond B. Cattell, Factor Analysis: An Introduction and Manual for the Psychologist and Social Scientist (New York: Harper & Brothers, 1952), p. 298
17.
Russett, op. cit , employs a somewhat different factorial procedure. Instead of using the Q-technique, he factors a "distance" matrix based on summed factor scoies foi each nation Professor Rummel has utilized this procedure at Yale in grouping 79 nations across a representative selection of CPS political variables w th iesults that exhibit a high degree of correspondence with four of the five groups reported on here. Our Group II does not appear in Rummel's analysis because virtually all of the African nations were excluded.
18.
Cf., Seymour M. Lipset, "Some Social Requisites of Democracy: Economic Development and Political Legitimacy," American Political Science ReviewLIII (March, 1959 ), 69-105
19.
and the remarks by James S. Coleman in Gabriel A. Almond and James S. Coleman, eds., The Politics of the Developing Areas (Princeton: Princeton University Press, 1960), p. 536, ff.
20.
The computer program employed, unlike many similar programs, was designed to cope with the problem of missing data. It does so by permitting only correlations between substantive codings to enter into the factor analytic calculations Where the missing data component is high (as is the case for a limited number of CPS nations), distortions can be introduced that are interpretable only by an examination of the original codings. In the case of South Africa, with its quite different policies toward the two major racial groups, many of the political variables were coded as "ambiguous" (a residual category) in the Survey. However, for certain variables, such as "Status of the [All-White] Legislature," relatively "polyarchic" substantive codmgs were entered.
21.
The term "parochial" is Gabriel Almond's. See Gabriel A. Almond and Sidney Verba, The Civic Culture (Princeton: Princeton University Press, 1963), p. 17, ff.
22.
In Rummel's factor analysis of ten CPS political variables, a traditional group emerged that included Afghanistan, Ethiopia, Iran, Joidan, Nepal, Sa'udi Arabia, Cambodia, Yemen, and Libya.
