¹ The theoretical discussion of political participation from the viewpoint of an individual act may be seen in Campbell, Anguset al., The American Voter (New York: John Wiley and Sons, 1960)Google Scholar; Dahl, Robert A., Who Governs? (New Haven: Yale University Press, 1961)Google Scholar; and in the useful summary of the literture in Milbrath, Lester, Political Participation (Chicago: Rand McNally and Company, 1965)Google Scholar. The emphasis upon individual motivation is illustrated by Dahl's insistence (page 279) that “Instead of seeking to explain why citizens are not interested, concerned, and active, the task is to explain why a few citizens are.” (His emphasis.) It is perhaps more fully illustrated by Milbrath's approach (e.g., page 6) where he states that “taking any political action generally requires two decisions: one must decide to act or not to act; and one must also decide the direction of his action.” Milbrath then discusses people “making up their minds,” making “choices,” their “understanding,” and other social-psychological aspects of political action. Given this orientation, level of information, degree of opinionatedness, intensity of feeling, rationality and quality of deliberation, consistency of ideological orientation, and other pysehological concomitants of participation become the primary focus of analysis.

² Further details of the study may be found in Alford, Robert R. and Scoble, Harry M., “Community Leadership, Education and Political Behavior,” American Sociological Review, 33 (April 1968), 259–272CrossRefGoogle Scholar.

³ See Wildavsky, Aaron, Leadership in a Small Town (Totowa: The Bedminster Press, 1964)Google Scholar.

⁴ To save space, the verbatim form of the items making up each index will not be presented here. (They may be obtained from the Wisconsin Survey Research Laboratory or from the authors.) Briefly, the index of local polıtical interest is based on five items asking: general interest in city politics; perception of major local problem(s); number of these; interest in decision(s) by city government or school board “in the past year;” and number of these. The index of local political information is based on four items asking: correct names of mayor, city clerk, school superintenddent, and Republican Party (or Democratic Party) county chairman. The index of attendance is based on three questions asking, “within the past two years,” if respondent had attended a public-issue discussion meeting, a political party meeting, and a city council (or school board) meeting. And the index of local electoral turnout is based on two items requesting reported frequency of voting in local general and in local primary elections “since you have qualified for voting … (and) wherever you have lived….”

⁵ Assumptions underlying the types are, for example, that an uninterested, uninformed voter is not a conspicuous local participant, whereas a non-attender who is nonetheless highly interested, informed, etc., is likely to leap quickly into action if he perceives an issue of concern.

⁶ For a general survey of the literature using, with some terminological differences, these same six categories, see Milbrath, op. cit. In the analysis sections which follow, for reasons of space we cannot fully cite the literature (especially where repetitively replicative); we will, however, attempt to cite literature of both a complementary and a contradictory sort on the lesser-researched independent variables, especially social mobility, community attachment, religion and ethnicity, and political environment.

⁷ See Coleman, James S., Introduction to Mathematical Sociology (New York: The Free Press of Glencoe, 1964), pp. 189–203Google Scholar. One reason Coleman advanced his alternative method may be simply stated: contingency tables, by the time one introduces three independent variables, become visually cumbersome, consume much space, and often elicit excessively subjective interpretation. An “effect parameter”—of the types denned and illustrated in our text—is a single, quantitative summary designed to avoid all three, interrelated problems.

⁸ It is only similar (not identical) because of the frequent need to change cutting points of the same indicator when moving from zero- to higher-order analysis.

⁹ Because the dependent variable was trichotomized (into “high,” “moderate,” and “low” local political involvement) rather than dichotomized, the use or arbitrary cutting points might unduly influence the relative association of independent variables with involvement. To check on precisely this, all multivariate analyses were run twice, using first “high” and then “low” involvement as the dependent variable. To save space in the text, effect parameters for “high” involvement only are reported—but in no case where these are “significant” (as defined) are they not also confirmed by effect parameters computed for “low” involvement.

¹⁰ The net effect of income actually increases slightly over gross effect, due to concentration of the population in low education and manual occupation categories (a sub-total of 840 cases) plus the relatively small positive influence of having a high income if the other two aspects of social status are both low.

¹¹ Net effect was here calculated on the assumption that farming is lower in status than the manual category; if one reverses the assumption the net effect parameter is +2.

¹² Unfortunately, this is not the first time that inquiry into social mobility has yielded somewhat unsatisfactory results. Since World War II, analysts have been interested in the political consequences of vertical social mobility. Research results have been highly variable and often conflicting, in part because of great variation in operational definitions of “mobility” and probably also because geographical mobility, often thought to accompany social mobility, has rarely been separated out of the inquiry. (Early conflicting results, including his own, may be found in Key, V. O. Jr., Public Opinion and American Democracy [New York: Alfred A. Knopf, 1961], pp. 145–148)Google Scholar. It is also relevant to note that early inquiries focused primarily on direction, or ideological content, of political opinion as the dependent variable, whereas we are here concerned with the more neutral amount of political activity and involvement.

On what might be termed a para-political subject similar to ours, research has produced further confusion if not contradiction. On the one hand, socially mobile persons were found to be not as socially active as stationaries in studies by Lenski, Gerhard E., “Social Participation and Status Crystallization,” American Sociological Review, 21 (Aug., 1956), 458–464CrossRefGoogle Scholar, and by Blau, Peter M., “Social Mobility and Interpersonal Relations,” American Sociological Review, 21 (June, 1956), 290–295CrossRefGoogle Scholar. On the other hand, a third inquiry found no such relationship, for which see Curtis, Richard, “Occupational Mobility and Membership in Voluntary Organizations: A Note on Research,” American Sociological Review, 24 (Dec., 1959), 846–848CrossRefGoogle Scholar.

More recently, we have found partial confirmation of our findings in the work of Jane H. Bayes, who made use of interview, mailed questionnaire, and aggregate electoral and census data on residents of Los Angeles County. (Bayes never explicitly specifies direction—including downward—of social mobility in her data; but because of the special demographic characteristics of the Los Angeles population, we assume that the greatest proportion of her cases of socially mobiles are upwardly mobile and thus that her findings lend support to that relevant portion of our data.) Having first established that socially active persons are also politically active, Bayes then demonstrated that: (1) in general, social mobility had no significant effect on organizational activity; (2) but when analysis was restricted to non-geographically-mobile residents, who had been raised in the county, then the social mobiles were significantly less active in organizations; (3) yet finally geographic mobility—alternatively defined by interstate movement, by intra-county residential mobility, and by length of residence—” is more destructive to previous social and civic obligations than is social or occupational movement within one geographic area.” (Page 164; emphasis added.) See Bayes, Jane H., “Political Participation and Geographic Mobility” (unpublished doctoral dissertation, Department of Political Science, UCLA, 1967)Google Scholar, especially chapters 6 and 7.

¹³ Early findings of general relevance may be seen in Sykes, Gresham, “The Differential Distribution of Community Knowledge,” Social Forces, 29 (May, 1951), 376–382CrossRefGoogle Scholar. The population was white, male wage-earners in Plainfield, New Jersey. The dependent variable was knowledge of local community affairs. Among the hypothesized independent variables having positive effect were home-ownership, geographic mobility, and length of residence. One of the earliest theoretical generalizations of such community-attachment findings to explicitly political behavior may be found in Chapter 30, “The Psychology of Voting,” by Lipset, Seymour M.et al, in Lindzey, Gardner (ed.), The Handbook of Social Psychology (Cambridge: Addison-Wesley Company, 1954)Google Scholar. And Bayes, op. cit., more recently demonstrated that geographic mobility itself, regardless of length of residence, was the most destructive for social and political participation of all factors she surveyed. (In thinking about the far smaller impact of inter-state geographic mobility in our data than in Bayes', we note that two-thirds of our respondents had lived all their lives in Wisconsin whereas more than 60 per cent of the Los Angeles population had been born outside California, suggesting that the local political system may be able to absorb without much change small amounts of geographic mobility. Future research of a comparative sort thus might usefully focus on whether there is a threshold effect or a “critical mass” factor involved.)

¹⁴ The index of subjective attachment was based on three items: “In general, this is a very good community to live in,” “I can hardly imagine myself moving out of this community at any time in the future,” and “I could be just about as satisfied with life in another community as I am here.” Agreements to the first two questions and disagreement to the third were each scored 1, producing an index ranging from 0 to 3 (high subjective attachment). The items were originally devised by Martin A. Trow for study of communication flows in Bennington, Vermont, and were based upon the discussion of “locals” and “cosmopolitans” in Merton, Robert K., Social Theory and Social Structure (Glencoe: The Free Press, 1957 edition)Google Scholar.

¹⁵ See two M.A. theses based on the same data reported here for more detailed analyses of the religious bases of political behavior in the four Wisconsin cities: Cousens, Patricia D., “Religion and Class and Predictors of Political Involvement” (unpublished Master's Thesis, Department of Sociology, University of Wisconsin, 1964)Google Scholar and Otto, Luther B., “Catholic and Lutheran Political Cultures in Medium Sized Wisconsin Cities” (unpublished Master's Thesis, Department of Sociology, University of Wisconsin, 1963)Google Scholar. The first study found that “with class controlled, there are no differences between Protestants and Catholics in levels of political involvement in the local community,” although distinct religious differences remained in party identification and voting. The second study produced a similar finding for Lutherans and Catholics.

¹⁶ See Campbell, op. cit., pp. 103–107 for motivational factors involved in the 1956 presidential election. The original scales are presented in Campbell, Angus, Gurin, Gerald, and Miller, Warren E. E., The Voter Decides (Evanston: Row, Peterson and Company, 1954), pp. 187–199Google Scholar, which we have adapted with specifically local referents.

¹⁷ The concept of political environment or milieu is essentially that of a subnational areal (e.g., “neighborhood”) political culture persisting through time and social space. Five representative political studies making use of the concept are: (1) Gosnell, Harold F., Machine Politıcs: Chicago Model (Chicago: University of Chicago Press, 1939), pp. 91–125Google Scholar, where the author used electoral and census data on Chicago subcommunities to determine types of voter-support for Franklin D. Roosevelt and also to compare 1928–1936 voting consistency; (2) a study similar to Gosnell's by Bartelmann, as reported in Tryon, R. C., Identification of Social Areas by Cluster Analysis; A General Method with an Application to the San Francisco Bay Area, vol. 8, No. 1 (University of California Publications in Psychology, 1955), pp. 1–100Google Scholar, in which the author compared the 1940 Roosevelt vote with the 1947 Congressional vote for Havenner (an FDR-Truman supporter), showing that the neighborhood distribution of the vote was practically the same in the two elections in San Francisco; (3) Kaufman, Walter C. and Greer, Scott, “Voting in a Metropolitan Community: An Application of Social Area Analysis,” Social Forces, 38 (Mar., 1960), 196–204CrossRefGoogle Scholar, comparing 1952 presidential vote with a 1954 local referendum (to establish a Metropolitan St. Louis Sewer District) in terms of both direction of voting and level of turnout and by reference to census tracts classified by Shevky's indices of social rank, urbanization, and minority segregation; (4) Eulau, Heinz, Zisk, Betty H., and Prewitt, Kenneth, “Latent Partisanship in Non-partisan Elections: Effects of Political Milieu and Mobilication,” in Jennings, M. Kent and Zeigler, L. Harmon (eds.), The Electoral Process (Englewood Cliffs: Prentice-Hall Inc., 1966), pp. 208–237Google Scholar, in which the researchers classified the city-wide partisan milieu—pre-dominantly Democratic, competitive, or pre-dominantly Republican, as denned by party registration data—of formally nonpartisan California cities and then demonstrated milieu effects on individual behavior (characteristics, perceptions, campaign activities, career expectations, etc., of incumbent councilmen); and (5) Bayes, op cit., chapter 8, in which Bayes arrayed Los Angeles County census tracts according to whether they were nationally-oriented in voting, state-local oriented, or mixed, and then demonstrated that the observable effect of political milieu was upon the behavior of the non-college educated respondents in her samples.

¹⁸ Our article cited in footnote 2 above treats the formal leadership of the four Wisconsin cities as such a political stratum.