¹ New York Times, January 7, 1962. See also Moon, H. L., Balance of Power: the Negro Vote (Garden City, N. Y., Doubleday, 1949), p. 7 and passim Google Scholar.

² 71 Stat. 635; 74 Stat. 86. Cf. U.S. Commission on Civil Rights, 1959 Report (Washington, 1959)Google Scholar; 1961 Report, Vol. I, “Voting” (Washington, 1961)Google Scholar.

³ New York Times, January 7, 1962; Lomax, Louis E., “The Kennedys Move in on Dixie,” Harpers Magazine, 05 1962, pp. 27–33 Google Scholar.

⁴ Wall Street Journal, November 6, 1961; New York Times, July 10, 1961.

⁵ Baltimore, Afro-American, 10 7, 1961 Google Scholar; New York Times, August 17, 1961.

⁶ The 1962 Atlanta, Georgia, national convention of the NAACP had the “Battle of the Ballots” as its theme. Raleigh (N.C.) News and Observer, June 24, 1962.

⁷ New York Times, March 29, 1962. Louis E. Lomax, op. cit.

⁸ For useful summaries of the literature see Lane, Robert E., Political Life (Glencoe, Ill.: The Free Press, 1959), ch. 16Google Scholar; and Lipset, Seymour M. et al., “The Psychology of Voting,” in Lindzey, G. (ed.), Handbook of Social Psychology (Cambridge, Mass.: Addison-Wesley Publishing Company, 1954), Vol. II, pp. 1126–1134 Google Scholar.

⁹ Hyman, Herbert H. and Sheatsley, Paul B., “Attitudes Toward Desegregation,” Scientific American, Vol. 195 (1956), pp. 35–39 CrossRefGoogle Scholar; Bettelheim, B. and Janowitz, M., The Dynamics of Prejudice (New York; 1950)Google Scholar; Tumin, Melvin M., Desegregation: Resistance and Readiness (Princeton University Press, 1958), p. 195 and passim.Google Scholar Zanden, James W. Vander, “Voting on Segregationist Referenda,” Public Opinion Quarterly, Vol. 25 (1961), pp. 92–105 CrossRefGoogle Scholar, finds the evidence in support of the relationship in voting on segregationist referenda in the South “inconsistent and even contradictory … this study seems to suggest that the socioeconomic factor may not play as simple or as critical a role as some of us doing research in this field have been prone to assign it” (p. 105).

¹⁰ In addition to the problem of the relative importance of political variables, we are postponing consideration of still another possibility—that variations in state systems (social, economic, and political) account for a significant proportion of the variation in Negro registration among southern counties.

¹¹ A complete list of sources used to obtain county frequencies for the independent variables used in this analysis would be too lengthy to reproduce here. A mimeographed list will be supplied by the authors upon request.

We are indebted to the following research assistants for their help in collecting these data: Lawton Bennett, Lewis Bowman, Barbara Bright, Jack Fleer, Donald Freeman, Douglas Gatlin, and Richard Sutton. All told, the collection and coding of these data took one man-year of work.

¹² Voter registration rates, by race, are presented in U.S. Commission on Civil Rights, 1959 Report and 1961 Report, Vol. I, “Voting.” The 1958 registration data, contained in the 1959 Report, are more complete and were used for all states except Tennessee. The 1960 figures, printed in the 1961 Report, are the only ones available for Tennessee.

¹³ There are 1136 counties in the 11 southern states, 1028 of which have populations containing at least 1 per cent Negroes.

¹⁴ For a good discussion of correlation analysis see Hagood, M. J. and Price, D. O., Statistics for Sociologists (New York, 1952), chs. 23 and 25Google Scholar.

¹⁵ All computations were made on the University of North Carolina's UNIVAC 1105 high-speed digital computer. The inaccuracy of some of the registration figures tends to reduce the magnitude of all correlations obtained by this analysis. The assumption of linearity underlying the computation of r also reduces the size of the correlations where the relationship between dependent and independent variables is, in fact, a curvilinear one. It is therefore safe to assume that the r's reported in this article err in the conservative direction.

¹⁶ Key, V. O. Jr., Southern Politics (New York, 1949)Google Scholar gives little attention to Negro voting since it was of little importance at the time he wrote (see, however, p. 518). His stress upon the over-riding importance of Negro concentration for all aspects of southern politics makes his study highly relevant, nonetheless. Other works specifically on Negro voting which stress the importance of Negro concentration include: Barnes, James F., Negro Voting in Mississippi, M.A. thesis, University of Mississippi, 1955 Google Scholar; Price, Margaret, The Negro and the Ballot in the South (Atlanta, Georgia: Southern Regional Council, 1959)Google Scholar; Price, H. D., The Negro and Southern Politics: A Chapter of Florida History (New York: New York University Press, 1957)Google Scholar; Strong, Donald, “The Future of the Negro Voter in the South,” Journal of Negro Education, Vol. 26 (Summer, 1957), pp. 400–407 CrossRefGoogle Scholar; United States Commission on Civil Rights, 1961 Report, Vol. I, “Voting.”

¹⁷ On this point see H. D. Price, op. cit., p. 41ff.

¹⁸ See H. D. Price, op. cit.

¹⁹ In view of the relatively high associations between Negro concentration and a wide variety of political phenomena (including Negro registration rates), it might be argued that Negro concentration is, in fact, a “political” rather than a “demographic” variable. But Negro concentration is as strongly associated with many social and economic characteristics of southern counties as it is with their political peculiarities. And while the correlations of Negro concentration with political characteristics are relatively large, they fall far short of a 1.0 correlation. As we shall demonstrate in a subsequent article, a number of political variables have an association with Negro registration that is independent of Negro concentration. Under these circumstances, to call Negro concentration a “political” variable would be distinctly misleading.

²⁰ See Lane, op. cit.; Lipset et al., op. cit.; Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E., The American Voter (New York, 1960), ch. 13Google Scholar; Key, V. O. Jr., Public Opinion and American Democracy (New York, 1961), ch. 6Google Scholar. For a study of these variables and political participation among southern Negroes, see Seasholes, Bradbury, “Negro Political Participation in Two North Carolina Cities,” Ph.D. dissertation, University of North Carolina, 1962 Google Scholar.

²¹ For an extreme statement of this position, see Frazier, E. Franklin, Black Bourgeoisie: The Rise of a New Middle Class in the United States (Glencoe, Ill.: The Free Press, 1957)Google Scholar. Less exaggerated statements to the same effect may be found in the literature cited in n. 16, above.

²² This is the basic argument of Frazier, op. cit. A more mundane explanation would be called for if counties from particular states were clustered at particular points on the curve in Figure 4, but examination of the same relationships for each state reveals no such state-by-state clustering.

²³ See especially, U.S. Commission on Civil Rights, 1961 Report, Vol. I, “Voting,” pp. 143–199.

²⁴ This and other measures of county-wide characteristics might better be considered separately for Negroes and whites, but they are not separately reported in the census.

²⁵ On Negro voting in urban settings see Farris, Charles D., “Effects of Negro Voting Upon the Politics of a Southern City: An Intensive Study, 1946–48,” Ph.D. dissertation, University of Chicago, 1953 Google Scholar; Hillery, George A., “The Presence of Community Among Urban Negroes: A Case Study of a Selected Area in New Orleans,” M.A. thesis, Louisiana State University, 1951 Google Scholar; Reissman, Leonard et al., “The New Orleans Voter: A Handbook of Political Description,” Tulane Studies in Political Science, Vol. II (1955), pp. 1–88 Google Scholar; Roberts, Cleo, “Some Correlates of Registration and Voting Among Negroes in the 1953 Municipal Election of Atlanta,” M.A. thesis, Atlanta University, 1954 Google Scholar; Walker, Harry J., “Changes in Race Accommodation in a Southern Community,” Ph.D. dissertation, University of Chicago, 1945 Google Scholar.

²⁶ The Bureau of the Census defines Standard Metropolitan Areas as a county or group of contiguous counties which contains at least one city of 50,000 inhabitants or more. The contiguous counties must be socially and economically integrated with the central city to be included in the SMA.

²⁷ See the literature cited in n. 9, above.

²⁸ Eleven of the 28 counties in which the average white adult has completed less than seven years of schooling are French-Catholic parishes in Louisiana. Even if those parishes are eliminated, the trend shown in Figure 6 remains the same. The partial correlation between white school years and Negro registration, controlling for per cent Roman Catholic, is —.25.

²⁹ Stouffer, Samuel A., Communism, Conformity, and Civil Liberties (New York, Doubleday, 1955)Google Scholar.

³⁰ Sindler, Allan P., Huey Long's Louisiana (Baltimore: The Johns Hopkins Press, 1956)Google Scholar; V. O. Key, Jr., op. cit., ch. 8; Fenton, John H. and Vines, Kenneth N., “Negro Registration in Louisiana,” this Review, Vol. 51 (1957), pp. 704–13Google Scholar.

³¹ The most recent attempt to compile county-by-county figures on church membership is reported in a census by the National Council of Churches of Christ, Churches and Church Membership in the U.S., Series C, 1956. Negro churches are not included in this census, and the figures reported for many white churches appear to be incomplete.

³² A simple Kendall tau rank order correlation of the two distributions of correlations in Tables II and IV is —.54.

³³ Urban counties in the South undoubtedly purge their registration lists with greater regularity than the more rural ones. How much effect this may have on these correlations cannot be ascertained.

³⁴ Indeed, it was on the basis of a roughly equal multiple correlation, based on survey data rather than aggregate county data, that an early voting behavior study concluded that “social character¬istics determine political preference.” Lazarsfeld, Paul F., Berelson, Bernard, and Gaudet, Hazel, The People's Choice (New York: Columbia University Press, 1948), p. 27 Google Scholar. This work reports a multiple correlation between voting preference and social factors of “approximately .5” (p. 25).

³⁵ For example, Martin Luther King's statement in a speech to the 1962 NAACP annual convention about southern Negroes being “able to elect at least five Negroes to Congress in the next few years” seems to underestimate wildly the social and economic barriers to Negro political participation. New York Times, July 6, 1962. See also the sanguine expectations of Lomax, op. cit.

³⁶ No trend data were available on religious affiliation. Median income figures, by race, were not available for 1940. In Figure 8, it is assumed that median income for both races increased at the same rate between 1940 and 1950 as between 1950 and 1960.

³⁷ If white school years completed continues to increase at the 1950–60 rate, the average southern white will have completed 11.4 years of schooling by 1970 and many southern counties will have average white school years completed of 12 years or more. Assuming that the relationship presented in Figure 6 continues to hold true, the effect of white education on Negro registration may gradually reverse.