¹ For useful examinations of various aspects of these changes, see Campbell, Bruce A., “Patterns of Change in the Partisan Loyalties of Native Southerners: 1952–1972,” paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, May 1–3, 1975 Google Scholar; Havard, William C., ed., The Changing Politics of the South (Baton Rouge: Louisiana State University Press, 1972)Google Scholar; Ladd, Everett Carll Jr., with Hadley, Charles D., Transformations of the American Party System (New York: W. W. Norton and Company, 1975), pp. 129–177 Google Scholar; Schreiber, E. M., “‘Where the Ducks Are’: Southern Strategy versus Fourth Party,” Public Opinion Quarterly, 35 (Summer, 1971), 157–167 CrossRefGoogle Scholar; Strong, Donald S., “Further Reflections on Southern Politics,” The Journal of Politics, 33 (May, 1971), 239–256 CrossRefGoogle Scholar; and Wolfinger, Raymond E. and Arseneau, Robert S., “Partisan Change in the South,” paper presented at the Annual Meeting of the American Political Science Association, Chicago, August 29–September 2, 1974 Google Scholar.

² The data used in this study come from national surveys conducted by the Survey Research Center and, more recently, the Center for Political Studies (hereafter SRC/CPS) of the University of Michigan in each presidential election year since 1952. For purposes of this study, the South is defined as the eleven former Confederate states minus Tennessee. (While it would have been highly desirable to add Tennessee to the analysis sample, the original sample design's specification of that state as a border state” does not permit it. Interviews conducted in Tennessee were meant to be representative of border states in general and not the state itself. Thus, the addition of Tennessee to the other states of the Confederacy does not yield an unbiased sample of those eleven states.) Sample estimates are very similar to the census statistics, where available, on the relative sizes of the various groups analyzed in this study. The sample does tend, however, to exaggerate somewhat the number of southerners who moved from outside the South.

³ Democratic presidential candidates won the following percentages of the two-party vote in the ten southern states in the sample: 66% in 1948, 52% in 1952, 48% in 1956, 51% in 1960, 51% in 1964, 48% in 1968, and 29% in 1972. In 1948 and 1968, when third-party totals were substantial, Democratic candidates polled 50% and 31%, respectively, of the total vote.

⁴ Both registration figures apply to the eleven states of the old Confederacy. See Campbell, David and Feagin, Joe R., “Black Politics in the South: A Descriptive Analysis,” The Journal of Politics, 37 (February, 1975), 133–134 CrossRefGoogle Scholar.

⁵ These figures on black elected officials cover the eleven-state South and come from The New York Times, March 4, 1974, and March 6, 1975.

⁶ The proportions of this monopoly are staggering. Democrats occupied 88 per cent of the seats in the lower houses and 92 per cent of the seats in the upper houses in the ten-state South in 1975 – both totals slightly above those in the previous year. Only in Florida (72%) and Virginia (66%) did the Democratic percentage drop below 86 per cent in lower houses. Only in Florida (68%) did it drop below 88 per cent in the upper houses. The Democratic advantage has grown even larger with the 1976 election.

⁷ See Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E., The American Voter (New York: John Wiley and Sons, 1960), pp. 120–145 Google Scholar.

⁸ For purposes of this analysis, the original SRC/CPS party identification measure was collapsed into four categories. Democratic and Republican identifiers include both weak and strong identifiers from the initial coding, but not the Independent leaners. Independents are all those who responded “Independent” to the first party identification question and include pure Independents as well as leaners. The Other category contains minor party identifiers and those classified as apolitical. Nonresponses have been eliminated as missing data.

⁹ The classic illustration of defection without partisan change comes from the two presidential elections in which Eisenhower was elected President while the overwhelmingly Democratic loyalties of the electorate were maintained. See Campbell, Angus, “A Classification of the Presidential Elections,” in Elections and the Political Order, ed. Campbell, Angus, Converse, Philip E., Miller, Warren E. and Stokes, Donald E. (New York: John Wiley and Sons, 1966), pp. 69–74 Google Scholar.

¹⁰ The strongest proponent of this thesis is Phillips, Kevin P., The Emerging Republican Majority (New Rochelle, N.Y.: Arlington House, 1969), pp. 187–209 Google Scholar. See also Burnham, Walter Dean, Critical Elections and the Mainsprings of American Politics (New York: W. W. Norton and Company, 1970), pp. 135–174 Google Scholar; and E. M. Schieiber, “‘Where the Ducks Are’…”

¹¹ See Strong, Donald S., Urban Republicanism in the South (University, Alabama: University of Alabama Bureau of Public Administration, 1960)Google Scholar.

¹² See Converse, Philip E., “On the Possibility of a Major Political Realignment in the South,” in Elections and the Political Order, pp. 212–242 Google Scholar. For a more recent assessment, see Converse, , “Change in the American Electorate,” in The Human Meaning of Social Change, ed. Campbell, Angus and Converse, Philip E. (New York: Russell Sage Foundation, 1972), pp. 307–317 Google Scholar.

¹³ See Inglehart, Ronald and Hochstein, Avram, “Alignment and Dealignment in France and the United States,” Comparative Political Studies, 5 (October, 1972), 343–372 CrossRefGoogle Scholar.

¹⁴ Interregion migrants are defined as individuals who lived in a different region or nation (since some foreign-born, mostly Cubans and Mexicans, are involved) from the one in which they had grown up. Because there is no other way to measure migrations from the SRC/CPS survey, each year's group of migrants contains those who had migrated through the previous presidential year plus those who had migrated since. There is no way to determine the actual date when the migration took place. Black migrants have been counted as migrants and excluded from the “native” black category. Thus, the three subgroups contained in Table 2 plus the subgroup of native whites to be analyzed later are mutually exclusive; they are also exhaustive of the entire southern subsample except where region of origin was not ascertained or where race was “other.”

¹⁵ See Permaloff, Anne, “Partisan Change in the American Electorate,” paper presented at the Annual Meeting of the Southwestern Political Science Association, Dallas, March 28–30, 1974, p. 20 Google Scholar.

¹⁶ It should be reiterated that the SRC/CPS sample seems to overestimate the number of southerners who are immigrants. The United States Census for 1960 and 1970 shows fewer immigrants. Part of this difference, though, surely lies in the different definitions of immigrants in the sample and the Census. In the sample, immigrants are southerners who had grown up outside of the ten-state region. In the census, immigrants had to be defined as southerners born outside of the Census South (15 states plus the District of Columbia) because no other data were readily available.

¹⁷ See Salamon, Lester M. and Evera, Stephen Van, “Fear, Apathy, and Discrimination: A Test of Three Explanations of Political Participation,” American Political Science Review, 67 (December, 1973), 1288–1306 CrossRefGoogle Scholar.

¹⁸ In 1960, most blacks had no opinion on school integration. Of those who did, about half perceived no party differences on the issue. Of those who saw party differences, four times as many believed the Republicans to be less anti-integration than the Democrats. By 1964, these perceptions had undergone dramatic change. Three-fifths of all blacks perceived the Democratic party as more favorable to school integration. The remainder either perceived no difference or reported no opinion. Similar figures appear in 1968, when three-quarters of all blacks perceived the Democrats as more favorable to school integration. The question on perceived party differences was not asked in 1972 and was not asked in comparable form in 1956.

¹⁹ The base for these operations is the partisan distribution of the southern electorate in 1972. The impact of emigrants was estimated by simply adding them to this base. The result is a new distribution of partisanship in the South, assuming that no one had left the region. The impact of immigrants was estimated by removing them from the 1972 electorate. The effect of each interregional migration, then, is the difference between each new distribution and the base. The impact of black mobilization was determined by rearranging the 1972 distribution so that 30.4% of blacks fell into the “other” category (as was the case in 1952) with the remainder distributed proportionally as in 1972, and comparing the result to the base. This counts only movement of blacks out of the “other” class and ignores the changes in black identification between 1952 and 1972.

²⁰ Ladd and Hadley concur with this contention. They show convincingly that the South steadfastly supported New Deal policies through the war. Indeed, southerners were found to support the New Deal more fully than residents of any other region. Postwar changes, they argue, were the result of changing Democratic party positions on the racial issue and, more recently, the increasing general conservatism brought about by the modernization of the region. See Ladd, with Hadley, , Transformations of the American Party System, pp. 130–151 Google Scholar.

²¹ See Beck, Paul Allen, “A Socialization Theory of Partisan Realignment,” in The Politics of Future Citizens, ed. Niemi, Richard G. (San Francisco: Jossey-Bass Publishers, 1974), pp. 199–219 Google Scholar.

²² See Beck, “A Socialization Theory,” for a full statement of the theory, as well as empirical support for it on a nation-wide scale.

²³ Analysis of the full cohort table is hampered by the large sampling errors which accompany most of the cell estimates. A more detailed analysis of the native white electorate – to be initiated shortly – necessitates some combination of cohorts. For the theoretical and empirical reasons I have cited here, the combination I employ seems to be the most reasonable. To make the two generations used in this study consistent with my earlier work, however, I have divided the sample into those who became 21 years old before 1946 and those who reached this age in 1946 or after. Thus, the 1945–1948 age cohort in Table 5 is broken up in Table 6. This makes no difference in subsequent analysis.

²⁴ For example, only a quarter of the Solid South Generation Democrats (of native white southern voters) supported McGovern in 1972. About half supported Humphrey in 1968.

²⁵ In 1956 and 1960 respondents were asked to agree or disagree with the statement: “The government in Washington should stay out of the question of whether white and colored children should go to the same school.” From 1964 through 1972, the question was: “Do you think the government (in Washington) should see to it that white and Negro children go to the same schools or stay out of this area as it is none of its business?” Segregationists were defined as those selecting the “stay out” alternative in each case. A school integration question was not asked in the 1952 survey. The focus on federal government intervention probably accounts for both lower support for school integration and much lesser growth in support than are reported by Greeley and Sheatsley for whites in the South and border states in 1963 and 1970. See Greeley, Andrew M. and Sheatsley, Paul B., “Attitudes toward Racial Integration,” Scientific American, 225 (December, 1971), 13–19 CrossRefGoogle ScholarPubMed.

²⁶ The question asked in 1960 was: “Which party do you think is more likely to stay out of the question of whether (the government sees to it that) white and colored children go to the same school; the Democrats or the Republicans, or wouldn't there be any difference?” In 1964 and 1968, the question was: “Which party do you think is more likely to want the government to see to it that the white and Negro children go to the same schools? The Democrats, and Republicans, or wouldn't there be any difference between them on this?” No party image question on school integration was asked in 1972, and the party image question asked in 1956 is not strictly comparable with the others.

²⁷ See Brody, Richard A. and Page, Benjamin I., “Comment: The Assessment of Policy Voting,” American Political Science Review, 66 (June, 1972), 455–457 CrossRefGoogle Scholar, for a clear statement of the criteria for issue voting. The criteria I have used represent a modification of those suggested by Brody and Page.

²⁸ Analysis may be restricted to segregationists – those opposing governmental intervention to promote school integration (1960-1968) or to aid blacks and other minorities (1972) – because of the nature of the hypothesized impact of racial attitudes. The principal change in southern partisanship has been the decline in Democratic loyalties at a time when that party was emerging as the champion of the civil rights cause. Thus, it is anticivil-rights attitudes which should be looked to as the possible source of change.

²⁹ Because of changes in question wording between 1960 and 1964, the perceived party position categories are different. Respondents in 1960 were asked to choose the more anti-integrationist party, while in 1964 and 1968 they were asked to identify the more integrationist party. Hence, in Figure 1, “con” positions are entered for 1960 and “pro” positions for the later years. Because of the possibility that no party differences were perceived, “con” positions in 1960 are not the reciprocal of “pro” positions.

³⁰ The scale contains seven opinion categories, ranging from “government should help minority groups” at one extreme to “minority groups should help themselves” at the other. For the purposes of this analysis, responses were trichotomized into neutral (category four), liberal or pro-aid, and conservative or anti-aid. The distribution of respondent opinion on the minority aid question approximates that for school integration in 1972, with government intervention a bit less popular for minority aid than for school integration. Attitudes towards government intervention in minority aid and school integration exhibit a gamma correlation of .34. While the items clearly share a domain of content, this correlation suggests also some empirical distinctiveness between the items – making it impossible to determine whether the relationships to be discussed below are unique to the minority aid question or would appear for school integration too in 1972. Had perceptions of party positions on school integration been elicited in 1972, of course, I would have been able to use both items in analysis. A focus on school integration would have provided continuity with earlier years; a focus on minority aid would have contributed the full picture of party images which the school integration question did not.

³¹ Virtualiy a majority of the respondents had no perceptions of either party's position on this question. Well over half of these nonresponses are due to the absence of opinions on the initial minority aid issue. Whatever the reason for their appearance, nonresponses detract from the extent of issue partisanship.

³² The task of identifying segregationist issue partisans is complicated by the virtual absence of any trend toward Republican affiliations. The GOP may not yet have been perceived by many southerners as a viable alternative for long-run commitments of loyalty. In the initial coding of issue partisans, this possibility will be considered, and only perceptions of Democratic party positions will be utilized.

³³ Additional evidence of a decline in the importance of racial issues in the South comes from analysis of the questions in the SRC/CPS survey concerning likes and dislikes of the two parties. See Black, Merle and Rabinowitz, George, “An Overview of American Electoral Change: 1952–1972,” paper presented at the Annual Meeting of the Southern Political Science Association, New Orleans, November 8, 1974 Google Scholar. Also see Trilling, Richard J., Party Image and Electoral Behavior (New York: Wiley, 1976), pp. 149–173 Google Scholar.

³⁴ Partisan dealignment is a nationwide phenomenon, although it is less pronounced outside the South. See Inglehart and Hochstein, “Alignment and Dealignment…,” for regional comparisons. See Beck, A Socialization Theory,” for the socialization explanation of the dealignment process.

³⁵ See Key, V. O., Southern Politics in State and Nation (New York: Alfred A. Knopf, 1949) pp. 298–311 Google Scholar; and Schattschneider, E. E., Party Government (New York: Holt, Rinehart, and Winston, 1942)Google Scholar.

³⁶ It is quite possible, of course, that students of southern politics have overemphasized the importance of the racial question in the past. This would not be the first time that elite images of mass publics fell wide of the mark. Yet traditional southern politics seems to have been designed so as to maximize the impact of even a small segregationist issue public. The cost of segregationist rhetoric to politicians was low because of the disenfranchisement of the target group. Traditionally low rates of voting participation in the region's elections also increased the potential impact of an issue public. Finally, the absence of structured competition in the region's only competitive elections (the primaries), if only because of the uncertainty about outcome it imposed, made intense issue publics very valuable and, hence, strongly courted as potential members of an electoral coalition.