¹ The literature on political participation is voluminous. Seymour Martin Lipset summarizes much of the empirical work linking nonparticipation to low income and education in “Who Votes and Who Doesn't?” in Lipset, , Political Man (Garden City, New York: Anchor Books, 1963), pp. 183–229 Google Scholar; and summarizes, Robert Lane the theories purporting to explain this link in Political Life (Glencoe, Ill.: The Free Press, 1959), pp. 220–234 Google Scholar. See also: Berelson, Bernard, Lazarsfeld, Paul F. and McPhee, William N., Voting (Chicago: University of Chicago Press, 1954)Google Scholar; Hastings, Philip K., “The Voter and the Non-Voter,” American Journal of Sociology, 62 (November, 1956), 303–04CrossRefGoogle Scholar; Milbrath, Lester, Political Participation (Chicago: Rand-McNally Co., 1965)Google Scholar.

² The closest Lane comes to acknowledging a role for fear is in his discussion of the low sense of political efficacy and limited self-confidence among the poor. Lane, , Political Life, pp. 222–34Google Scholar.

³ Lipset, “Who Votes?” in Political Man; and Milbrath, , Political Participation, particularly pp. 48–141 Google Scholar.

⁴ Matthews, Donald R. and Prothro, James W., Negroes and the New Southern Politics (New York: Harcourt, Brace & World, Inc., 1966), p. 308 (emphasis added)Google Scholar.

⁵ Watters, Pat and Cleghorn, Reese, Climbing Jacob's Ladder: The Arrival of Negroes in Southern Politics (New York: Harcourt, Brace and World, 1967), pp. 124–25 (emphasis added)Google Scholar.

For other evidence from participant observers about the role of fear in southern black politics, see: Belfrage, Sally, Freedom Summer (New York: Viking Press, 1965)Google Scholar; Moody, Anne, Coming of Age in Mississippi (New York: The Dial Press, 1968)Google Scholar; Greenberg, Polly, The Devil Has Slippery Shoes: A Biased Biography of the Child Development Group of Mississippi (London: The Macmillan Company, 1969)Google Scholar; McCord, William, Mississippi: The Long Hot Summer (New York: W. W. Norton and Co., 1965)Google Scholar; Silver, James, Mississippi: The Closed Society (New York: Harcourt, Brace and World, 1963)Google Scholar.

⁶ Tanter, for instance, found that domestic violence is greatest in countries where income inequality is greatest. Tanter, Raymond, “Toward a Theory of Conflict Behavior in Latin America,” paper prepared for delivery at the Annual Meeting of the International Political Science Association, 1967 Google Scholar.

This discrimination factor corresponds very closely to the concept of “relative deprivation” formulated by students of domestic violence. In its more refined forms, however, the “relative deprivation” notion focuses not simply on the actual distribution of resources between the lower classes and their “betters,” but also on the perceptions of the lower classes about what they have a right to expect and on the relationship between these expectations and the possibility of fulfilling them. Most empirical tests of this notion to date, however, have had to measure these aspirations indirectly. See: Davies, James, “Toward a Theory of Revolution,” American Sociological Review, 27 (February, 1962), 5–19 CrossRefGoogle Scholar; Eckstein, Harry, “On the Etiology of Internal Wars,” History and Theory, 4 (1965), 133–63CrossRefGoogle Scholar; Gurr, Ted with Ruttenburg, Charles, The Conditions of Civil Strife: First Tests of a Causal Model (Princeton, N.J.: Center of International Studies, April 1967)Google Scholar; Smelser, Neil J., Theory of Collective Behavior (New York: The Free Press, 1963)CrossRefGoogle Scholar. For an opposite view, albeit on a slightly different point, see Britton, Crane, The Anatomy of Revolution, rev. ed. (New York: Vintage Books, 1952), especially pp. 264–65Google Scholar.

⁷ The Voting Rights Act of 1965 was the culmination of extended efforts by civil rights forces to secure federal assistance in their efforts to eliminate the barriers to black political participation in the South.

The new law outlawed the literacy tests and other discriminatory voter registration procedures, authorized the Attorney General to appoint voting registrars in counties where less than half of the voting age blacks were registered, and created a procedure to review any changes in election law in five southern states, including Mississippi. Voting Rights Act of 1965, 79 Stat. 437, 42 U.S.C., 1973. The Act was extended in 1970 for another five years.

⁸ Quoted in Sundquist, James. Politics and Policy: The Eisenhower, Kennedy and Johnson Years (Washington, D.C.: The Brookings Institution, 1968), p. 274 Google Scholar.

⁹ Voter Education Project, Black Elected Officials in the Southern States, mimeo (1969). p. ii Google Scholar; V.E.P. News, Vol. 3, No. 7, July 1969. p. 1; V.E.P. News, Vol. 4, Nos. 1 & 2, January-February 1970, p. 3 Google Scholar.

¹⁰ V.E.P. News, Vol. 4, Nos. 1 & 2, January-February 1970, p. 3. In Mississippi, the black registration gain was more substantial: 242,500 black registrants were added compared to 147,000 whites.

¹¹ This information comes from the Voter Education Project, Atlanta, Georgia. The four counties with black sheriffs as of July 1972 were Greene, Lowndes, Macon, and Bullock—all in Alabama. The four counties with black majorities on their governing boards were Greene in Alabama, Hancock in Georgia, and Surrey and Charles City in Virginia. In addition, at this writing, blacks have a majority on the Election Commission in Holmes County, Mississippi.

¹² Millsaps College, The Institute of Politics in Mississippi, “General Election 1971: Some Notes on Elections Involving Black Candidates.” mimeographed. For an analysis of this election, see: Salamon, Lester M., “Mississippi Post-Mortem: The 1971 Elections,” New South, 27 (Winter 1972), 43–47 Google Scholar.

¹³ Demographic data is from U.S. Department of Commerce, Bureau of the Census, 1960 Census of Population: General Social and Economic Characteristics, PC (1)-26C (Washington, D.C.: U.S. Government Printing Office, 1960)Google ScholarPubMed.

Voting statistics are from Abney, F. Glenn, Mississippi Election Statistics, 1900–1967 (University, Mississippi: The University of Mississippi Bureau of Governmental Research, 1968), p. 28 Google Scholar.

¹⁴ The classic statement of this problem can be found in Robinson, W. S., “Ecological Correlations and the Behavior of Individuals,” American Sociological Review, 15 (June, 1950), 351–57CrossRefGoogle Scholar. See also, Ranney, Austin, “The Utility and Limitations of Aggregate Data in the Study of Electoral Behavior,” in Essays on the Behavioral Study of Politics, ed. Ranney, Austin (Urbana, Ill.: University of Illinois Press, 1962), pp. 91–102 Google Scholar.

₁₅ The literature on “ecological inference” has come virtually full circle since Robinson's statement of the problem in 1950. In a recent volume on ecological analysis, for example, Erik Allardt argues that pre-occupation with the question of faulty inference has distracted attention from the more important question of the fruitfulness of various hypotheses. And in Allardt's view, the ecological approach may “facilitate fruitful causal interpretations better than the corresponding individual data, if such are available.” Allardt, Erik, “Aggregate Analysis: The Problem of its Informative Value,” in Quantitative Ecological Analysis in the Social Sciences, ed. Dogan, Mattei and Rokkan, Stein (Cambridge, Massachusetts: The M.I.T. Press, 1969), p. 8 Google Scholar.

Even more important, several scholars have identified ways to minimize the dangers of ecological inference without sacrificing the entire approach. These methodological clues have been incorporated in the research reported here. In a 1967 article in this Review, for example, W. Phillips Shively pointed out that the dangers of “ecological inference” can be minimized so long as “individuals have been grouped in such a way that their scores on the dependent variable are unrelated to the aggregation in which they fall, except indirectly through their scores on the independent variable.” As will become clear below, the explanatory models used here meet this condition. See Shively, W. Phillips, “‘Ecological’ Inference: The Use of Aggregate Data to Study Individuals,” American Political Science Review, 63 (December 1969), 1186 Google Scholar; and Blalock, Hubert M. Jr., Causal Inferences in Non-experimental Research (Chapel Hill: University of North Carolina Press, 1965), pp. 95–126 Google Scholar.

The dangers of “ecological inference” have been reduced even further here by adopting the suggestion made recently by Dogan and Rokkan to “develop statistical models for the exploration of alternative linkages between individual distributions and higher unit characteristics” as a way to avoid the “ecological fallacy.” In this article, three “alternative linkages” are explored. See Dogan, Mattel and Rokkan, Stein, “Introduction,” Quantitative Ecological Analysis, ed. Dogan, and Rokkan, , p. 8 (emphasis added)Google Scholar.

¹⁶ The county was chosen as the unit of analysis because it is the smallest unit for which political, economic, and social statistics are readily available.

Only the counties with black majorities were used because they are, by and large, the only ones where black candidates have run for countywide office and where civil rights activity has occurred, both of which are important in measuring the dependent variable, as will become clear below. Only the black majority counties in one state are used to eliminate the effects of different state voting laws and political histories. Mississippi was chosen because it is the state with which we are most familiar and because, except for Georgia, it is the only one with a large enough number of black majority counties to permit a statistically significant test.

The “explanations” developed here were tested using correlation and multiple regression analysis, two standard statistical measures of relationship. The data analysis was performed on the Vanderbilt University Sigma-7 computer using standard Product Moment Correlation and Multiple Regression programs. We are indebted to Laird Heal of the Vanderbilt University Computer Center for assistance with these programs.

¹⁷ See, for example, Matthews, Donald R. and Prothro, James W., “Social and Economic Factors and Negro Voter Registration in the South,” American Political Science Review, 57 (March, 1963), 24–44, especially p. 26CrossRefGoogle Scholar.

¹⁸ Countywide offices were at stake in two Mississippi elections during this period—in 1967 and 1968. The posts for which candidates ran county-wide in 1967 were: Sheriff. Chancery Clerk, State Senator, and School Superintendent. In 1968, they were: Election Commission (five members per county) and School Board (five members per county).

¹⁹ Four of the 29 counties analyzed here had only one black candidate for countywide office, and nine had none.

²⁰ This was the case, for example, with regard to the white candidate for Chancery Clerk in Wilkinson County, and the white candidate for State. Senator in Holmes County.

²¹ In mathematical terms, this participation index was computed as follows:

²² See note 5, supra.

²³ Matthews, and Prothro, , Negroes and the New Southern Politics, pp. 166 and 303 Google Scholar.

²⁴ Matthews and Prothro, p. 308.

²⁵ U.S. Commission on Civil Rights, Political Participation: A Study of the Participation by Negroes in the Electoral and Political Processes in Ten Southern States Since the Passage of the Voting Rights Act of 1965 (Washington: U.S. Government Printing Office, 1968), p. 127 Google Scholar.

²⁶ For a revealing discussion of the role of violence and intimidation of blacks in the southern caste system, see: Davis, Allison, Gardner, Burleigh B., and Gardner, Mary R., Deep South: A Social Anthropological Study of Caste and Class (Chicago: University of Chicago Press, 1941), pp. 392–400 Google Scholar.

²⁷ Watters and Cleghorn, Climbing Jacob's Ladder. Watters and Cleghorn report (p. 129) that Mississippi bankers used to tell black loan applicants in the 1960s that: “If you can afford to vote, you don't need a loan.”

²⁸ U.S. Commission on Civil Rights, Hearings, Jackson, , Mississippi, February 16–20, 1965 (Washington. D.C.: U.S. Government Printing Office, 1965), I, 209 Google Scholar. Interestingly, this study was done by James Prothro, yet Prothro neglected to mention it in his book on Negro politics in the South published two years later.

²⁹ U.S. Commission on Civil Rights, Political Paticipation, p. 127 Google Scholar.

³⁰ The most systematic evidence on the point is the four-county survey of black Mississippi school teachers conducted for the U.S. Civil Rights Commission. Commissioner Rankin summarized this study well when he asked the primary investigator if low teacher voter participation could not be attributed to “the advice of the superintendents and principals that ‘you be careful or you might lose your job. …’” The answer was an unequivocal yes. U.S. Commission on Civil Rights, Mississippi Hearings, 1965, I, pp. 207–215, especially p. 214Google Scholar.

³¹ It is important to note here that Matthews and Prothro, in their study of Southern black political participation, found little relationship between black occupational level and black political participation. How-ever, Matthews and Prothro were content to divide occupations simply into the standard white-collar vs. blue-collar categories commonly used by sociologists and political scientists. In doing so, however, they were implicitly accepting a conceptual framework we have argued is inadequate to bring the key features of black politics in the South into focus. Matthews, and Prothro, , Negroes and the New Southern Politics, pp. 12–13 Google Scholar.

Dividing the black labor force into “most vulnerable” and “least vulnerable” categories, however, is not a simple task. The Census classifications of occupations do not correspond very precisely to the distinction of interest to us here. For example, the occupations grouped together in the Census classification “Operatives and Kindred Workers” include highly vulnerable plantation tractor drivers and far less vulnerable railroad employees. The Census classification “Farmers and Farm Managers” includes both farm-owners and sharecroppers. To compile Table 2, therefore, it was necessary to make a number of difficult decisions, reflected in the Table notes. These decisions were based on the existing sociological literature on southern society, and on numerous interviews with civil rights activists in Mississippi.

³² Homeownership figures came from the U.S. Census of Housing, 1960, Vol. I: States and Small Areas, Part 26, Mississippi (Washington, D.C.: U.S. Government Printing Office, 1963), pp. 38–44 Google Scholar. Significantly, these two measures of black economic dependence were closely related. The correlation between the percentage of a county's black electorate in the Most Vulnerable occupations, and the percentage of blacks in owner-occupied homes in a county was —0.863.

³³ The probability that this result might have happened purely by chance was a mere 0.0029, or only 3 in 1000. Using a 5 per cent level of significance, this result is therefore clearly significant.

Strictly speaking, the significance level or probability measure is not really applicable here because these counties are not perceived as a random sample from a larger population. Nevertheless, the significance level is still interesting as a benchmark.

³⁴ Sue and Henry Lorenzi, “Managing Fear by a Community—Holmes County, Mississippi,” mimeo. The Lorenzi paper suggests that blacks in the rural South never lose their fear. Rather, they learn to manage it by participating in group activity that threatens the two-caste system. For a fuller analysis of the dynamics of black organizational development in three Mississippi counties in the 1960s, see Salamon, Lester M., Backwardness and Change in the American South: Mississippi as a Developing Society (Bloomington: Indiana University Press, forthcoming 1974), chapter 5Google Scholar.

³⁵ Crotty, William J., “Party Effort and Its Impact on the Vote,” American Political Science Review, 65 (June 1971), 439–50CrossRefGoogle Scholar; Blydenburgh, John C., “A Controlled Experiment to Measure the Effects of Personal Contact Campaigning,” Midwest Journal of Political Science, 15 (May 1971), 365–81CrossRefGoogle Scholar; Price, David E. and Lupter, Michael, “Volunteers for Gore: The Impact of a Precinct-Level Canvass in Three Tennessee Cities,” The Journal of Politics, 35 (May 1973), 410–38CrossRefGoogle Scholar.

³⁶ COFO was the coalition of civil rights groups formed in 1963 to sponsor voter education projects throughout Mississippi and to coordinate the black political battle in the state. At the height of COFO activity, in the historic summer of 1964, some 500 out-side workers, mostly white students from the North, were in the state. “COFO files,” Delta Ministry Office, Mt. Beulah, Edwards, Mississippi. See Watters, and Cleghorn, , Climbing Jacob's Ladder, pp. 63–68 Google Scholar; Belfrage, Freedom Summer.

³⁷ Until his murder in 1963, the NAACP Field Director in Mississippi was Medgar Evers. Soon after this, Medgar Evers's brother, Charles Evers, took over this job and served until 1969, when he was elected Mayor of Fayette, Mississippi.

³⁸ The eight interviewees were: the Reverend Harry Bowie, originally a SNCC volunteer, who works now for the Delta Ministry in McComb, Mississippi; Lawrence Guyot, former state chairman of the Freedom Democratic Party; Aaron Henry, President of the Mississippi NAACP and co-director of COFO; Jan Hillegas, Editor of the Freedom Information Service Newsletter, a Movement publication; Charles Horowitz, Delta Ministry staff member; the Reverend Edwin King, former chaplain at Tougaloo College and Movement candidate for Lieutenant Governor in 1963; William F. Minor, Jackson correspondent for the New Orleans Times-Picayune.

³⁹ The Goldwater vote was computed as a percentage of the two-party vote, instead of as a percentage of the potential white electorate, because, in 1964, no real threat to white political dominance existed and whites therefore felt little need to turn out for the election. A measure of white turnout, therefore, would have been meaningless, whereas a measure of county preference for Goldwater over Johnson had social relevance in view of Johnson's support for social welfare and civil rights legislation.

By 1968, the presence of a real black challenge at the polls put a premium on white turnout. Not surprisingly, therefore, white turnout increased sharply. In 1964, both presidential candidates together received 409,146 votes; in 1968, Nixon and Wallace alone received 502,146 while Humphrey, drawing mainly on black voters, received an additional 149,419. In the 1968 election, therefore, the best measure of white hostility to black rights was the turnout of voters for Wallace as a percentage of the potential white electorate, for this measure permits us to determine the extent to which the white community in a particular county mobilized itself to express its conservative opinion.

⁴⁰ Matthews, and Prothro, , Negroes and the New Southern Politics, p. 308 Google Scholar.

⁴¹ The fact that we discover a positive, if slight, correlation between white turnout and black mobilization suggests that what we are really witnessing here is a white response to the threat of black mobilization. The more blacks become mobilized, the more whites turn out to vote. This pattern is consistent with findings dating back to Key's, V. O. Southern Politics in State and Nation, particularly pp. 513–516 Google Scholar. It also conforms to the finding of Matthews and Prothro that there is a slight positive correlation between the incidence of violence and the level of black participation ( Matthews, and Prothro, , Negroes and the New Southern Politics, p. 308)Google Scholar. In both cases, the expression of white hostility is a response to black mobilization, not a cause of it.

⁴² Addition of the income discrimination variable to the “expanded fear model” boosts the explained variation by only 1.5 per cent, and the likelihood that any chance factor could have produced the same result was an unacceptable 30 per cent. Addition of the education discrimination factor contributes only 2.3 per cent to the explained variation, but with a probability of 20 per cent that the same result could have been produced by chance.

⁴³ The nonwhite SES Index for Mississippi counties was taken from: Mississippi State College, “Socio-Economic Status Indexes for Mississippi Counties,” byBryant, Ellen S., Bulletin 724 (State College, Mississippi: Mississippi State University Agricultural Experiment Station, April 1966)Google Scholar.

⁴⁴ As Lane points out: “Perhaps, for a simple conventional act such as voting, income is more important, while more complex forms of participation are more dependent upon qualities associated with education.” Lane, Robert, Political Life, p. 222 Google Scholar.

⁴⁵ The correlation between nonwhite median education and the percentage of a county's labor force in Vulnerable Occupations is —.773, an extremely high figure.

⁴⁶ Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E., The American Voter: An Abridgment (New York: John Wiley and Sons, Inc., 1964), p. 251 Google Scholar.

⁴⁷ U. S. Commission on Civil Rights, Hearings, Jackson, , Mississippi, 1965, I, 211–12Google Scholar.

⁴⁸ The term “closed society” was used by historian James Silver to characterize the lack of political liberty in Mississippi in the late 1950s and early 1960s. See: Silver, James W., Mississippi: The Closed Society (New York: Harcourt, Brace, and World, Inc., 1963), especially p. xvi Google Scholar.

⁴⁹ For an example of how these pressures worked, see: Salamon, , “Mississippi Post-Mortem,” pp. 45–7Google Scholar.

⁵⁰ For further discussion of this point, see: Salamon, Lester M., “Family Assistance: The Stakes in the Rural South,” The New Republic (February 20, 1971), pp. 17–18 Google Scholar; and Van Evera, Stephen, “Welfare Reform: Bonanza for Dixie's Blacks,” Progressive, June 1972, p. 27 Google Scholar. Recent versions of this idea, by eliminating the provision for national administration, threaten to destroy the potential liberating impact of the plan.

⁵¹ Tucker, Robert C., “Russia, the West, and World Order,” in Tucker, Robert, The Soviet Political Mind (New York: Praeger Publishers, Inc., 1963), pp. 181–82Google Scholar.

⁵² Hartz, Louis, The Liberal Tradition in America, A Harvest Book (New York: Harcourt, Brace and World, Inc., 1955), p. 306 Google Scholar.