¹ In this analysis, the term “modernization” is used merely as a convenient equivalent of the term “development.” I avoid the term “economic development” because it implies too much in the way of causation. As Robert Heilbroner puts it, “Economic development is not primarily an economic, but a political and social process.” The Great Ascent (New York: Harper & Row, 1963), 24. The following items, selected from a voluminous literature, are especially suggestive for the coneerns of this study: Lipset, Seymour M., “Economic Development and Democracy,”— Chapter 2 of Political Man (Garden City, N.Y.: Doubieday, 1963)Google Scholar; Lerner, Daniel, The Passing of Traditional Society (New York: Free Press of Glencoe, 1958)Google Scholar; Ward, Robert E. and Rustow, Dankwart (eds.), Political Modernization in Japan and Turkey (Princeton: Princeton University Press, 1964)CrossRefGoogle Scholar; Rostow, W. W., The Stages of Economic Growth (Cambridge, England, 1960)Google Scholar; Karl W. Deutsch, “Social Mobilization and Political Development,” this Review, 55 (September, 1961), 493–514; McClelland, David C., The Achieving Society (Princeton: Van Nostrand, 1961).CrossRefGoogle Scholar Development has been a natural concern of recent attempts to apply survey research techniques to differences among nation states. See Lipset, op. cit.; Banks, Arthur S. and Textor, Robert B., A Cross-Polity Survey (Cambridge, Massachusetts: M.I.T. Press, 1963)Google Scholar; Russett, Bruceet al., World Handbook of Political and Social Indicators (New Haven: Yale University Press, 1964)Google Scholar; Almond, Gabriel and Verba, Sidney, The Civic Culture (Boston: Little Brown, 1965).Google Scholar Important works appearing since the basic structure of this paper was established include Pye, Lucian W., Aspects of Political Development (Boston: Little Brown, 1966)Google Scholar; and Elazar, Daniel, American Federalism: a View from the States (New York: Crowell, 1966).Google Scholar

² Cited in Lerner, op. cit., p. 43. Perhaps the most influential framework for the study of modernizing political systems is contained in Almond, Gabriel and Coleman, James, The Politics of Developing Areas (Princeton, N.J.: Princeton University Press, 1960).Google Scholar For comparisons among American states, see Jacob, Herbert and Vines, Kenneth N. (eds.), Politics in the American States (Boston: Little Brown, 1965).Google Scholar

³ McClelland, op. cit., 1–9; Lerner, op. cit., 49.

⁴ Pye, op. cit.; Pye, Lucian W. and Verba, Sidney, Political Culture and Political Development (Princeton: Princeton University Press 1965).CrossRefGoogle Scholar

⁵ Jacob and Vines, op. cit., 52, 69, 114, 152, 155, 199, 230, 400, 441.

⁶ “… Differences in the level of human reproductivity more sharply differentiate more-developed from less-developed areas than any other criterion, including per capita income, urbanization, literacy, industrialization, etc.”: Hauser, Philip M., “World Population Problems” (Headline Series No. 174, Foreign Policy Association, 12, 1965), p. 10.Google Scholar

⁷ A preliminary analysis suggested that most of these indicators are highly stable in the short run. Little would be gained by adding variables from other years unless a specific time dimension were incorporated in the research design. The nature of the indicators is for the most part obvious either from the title or the discussion below. Benchmark variable titles are shorthand for median family income, median school years completed by persons twenty-five years old and over, and children ever born per ten thousand women ever married. Variable 10, Commuting, is based on the percent of workers who neither worked at home nor walked to work. Ethnic Diversity, variable 18, was obtained by multiplying the percentage of foreign born combined with the percentage native of foreign or mixed parentage, total quantity by the percentage non-White. Voting, variable 19, is taken from data reported at p. 40 of Jacob and Vines, op. cit., and consists of average turnout in gubernatorial and Senatorial elections in non-presidential years 1952–1960. Variables 20 through 23 merge general revenue or expenditure patterns of state and local units. Variable 21 is amount of general revenue from own sources per $1000 of personal income. Government Employment, variable 23, refers to the number of state and local government employees per 10,000 population. Executive Salary Score, variable 24, was based on combined salaries of Governor, Attorney General, Justice of State Supreme Court and Auditor. Variables 25, 26 and 27 refer to state government percent of state and local tax revenue, state government percent of state and local direct general expenditure, and percent of state and local employees employed by state government. Power of Governor, variable 28, is an index developed by Joseph A. Schlesinger which can be found at p. 229 of Jacob and Vines, op. cit., based upon budget powers, appointive powers, tenure potential and veto powers. Variable 32, Party Competition, is depicted by deviations from fifty percent of grand averages of electoral victories by the Democratic party in the period 1946–1963, and represents a processing of data at p. 65 of Jacob and Vines, op. cit. Variable 33, Democratic Influence, depicts the same data in raw form and therefore amounts simply to the proportion of Democratic party victories in the period 1946–1963.

⁸ Almond and Verba, op. cit., p. 3; Pye, op. cit., pp. 46–47.

⁹ Agger, Robert, Goldrich, Daniel, and Swanson, Bert, The Rulers and Ruled (New York: Wiley, 1964), pp. 6–14.Google Scholar

¹⁰ Ibid., p. 7. “Scope of Government” has interesting affinities with “Political Development”; see Pye, op. cit., chapter 2.

¹¹ One difficulty is how to treat federal grants. Indicators 20 and 21 exclude federal grants, while 22 and 23 include them directly or indirectly. When these pairs of indicators are intercorrelated, the r's are all found to be +.49 or higher, and two of them are higher than +.80. The correlations may seem high in view of the absolute dollar magnitude of federal grants in recent years. However, a large quantity of federal grants need not appreciably change the orders of the states in their levels of spending. The extent to which federal grants “equalize” can be easily exaggerated. A separate analysis suggested that federal grants operate to somewhat reduce the range of differences among states, but that they have little effect on the ordering of states within a given range.

¹² Op. cit., pp. 6–7. See also the interesting finding of Russett et al. that the central government's role in the economy increases with development, but may diminish at very high levels. Their data show the United States to be much lower than most highly developed countries in the ratio of central government revenue to G.N.P., op. cit., pp. 308–309.

¹³ Benson, G. C. S., The New Centralization (New York; Farrar and Rinehart, 1941), p. 116Google Scholar; and Kaufman, Herbert, Politics and Policies in State and Local Governments (Englewood Cliffs, N.J.: Prentice-Hall, 1963), p. 12.Google Scholar Kaufman's comment is relevant: “No one has yet devised a method of measuring organizational centraliza tion, so it is mostly by means of intuitive application of rough clues that we form our impressions about this trait of organization.”

¹⁴ The essence of the distinction, as suggested above, is that Party Competition measures the extent to which the parties divide the electoral stakes equally, regardless of the direction in which the deviation from fifty percent occurs. A state in which Democrats won 48% of the victories is scored as being just as competitive as a state where Democrats won 52% of the victories. Democratic Influence uses the same data to simply indicate the ordering of the states on the percentage of electoral victories won by Democrats. I thus clarify, and present separately, the two elements contained in such classificatory designations as “One-Party Democratic,” “Modified One-Party Republican,” etc. Austin Ranney's scheme, presented at p. 65 of Jacob and Vines, op. cit., sets cutting points at comparable intervals, but orders twenty five states in the “Two-Party” category on Democratic Influence only, and not (directly) on party competition.

¹⁵ The author is grateful to James B. Wigle of the Institute for Social Research, University of Michigan, for the use of his Factor Analysis Program for the IBM 7090, and to the Technical Services Branch of the Inter-university Consortium for Political Research for performing the computations. Neither Mr. Wigle nor the Consortium is in any way responsible for flaws in the method, approach, or conclusions of the study. Standard treatments of factor analysis are described in Fruchter, Benjamin, Introduction to Factor Analysis (New York: Van Nostrana, 1954)Google Scholar; and Harman, Harry H., Modern Factor Analysis (Chicago: University of Chicago Press, 1960).Google Scholar For an interesting example of factor analysis which may be compared to that used in this study see Philip M. Gregg and Arthur S. Banks, “Dimensions of Political Systems: Factor Analysis of A Cross-Polity Survey,” this Review, 59 (September, 1965), 602–615.

¹⁶ Harman, op, cit., pp. 307–308. Data were also submitted to Oblimin rotation with results that await further analysis.

¹⁷ Systematic comparison of our findings with results of cross-nation studies is not attempted in this paper. However, the possibilities are tantalizing. For example, in the world contest too, message exchange may suggest consolidation rather than instability. Thus Russett et al. report significantly high negative correlations between Deaths from Domestic Group Violence and nine communications indicators: op. cit., p. 272.

¹⁸ In-Migration is the proportion living in a given state in 1960 which resided out of state in 1955. Population Increase compares 1950 with 1960.

¹⁹ Jacob and Vines, op. cit., p. 228.

²⁰ Factor scores for the Varimax factor structure were computed according to the method of complete estimation (regression) from a matrix of standard score coefficients. The program employed produces a mean of 50.0 and a standard deviation of 10.0 for each factor on each observation. Factor V scores are omitted from Table 4 for reasons stated above.

²¹ This is especially true of the notion of pervasive competition. One difficulty is that competition among states is usually assumed to center around tax rates. If this is true, and if the eventual result is more tax rate equality, the developmental differences among states should lead to service inequalities with the more affluent states possessing a higher level of services. It is a nice question whether inequalities in service levels would in turn set competitive forces in motion that would reduce the inequalities.

²² Data are not reported in this and the following illustrations which should be regarded as tentative findings only. See Dawson and Robinson's analysis in Jacob and Vines, op. cit., Chapter 10; and Richard Hofferbert, “The Relation Between Public Policy and Some Structural and Environmental Variables in the American States,” this Review, 60 (March, 1966), 73–83.

²³ Cf. Quinney, Richard, “Suicide, Homicide, and Economic Development,” Social Forces, 43 (03, 1965), 401–406.CrossRefGoogle Scholar On the basis of a unidimensional notion of development, this author simply asserts that highly developed countries have relatively low homicide rates, but relatively high suicide rates.

²⁴ Op. cit., Chapter 4.

²⁵ Op. cit., p. 109.