¹ Dahl, Robert A., Modern Political Analysis, 2nd ed. (Englewood Cliffs, 1970)Google Scholar, chap. 3. See also Harsanyi, John C., “Measurement of Social Power, Opportunity Costs, and the Theory of Two-Person Bargaining Games,” Behavioral Science, 7 (January 1962), 67–80.CrossRefGoogle ScholarPubMed

² As a limiting case, “pure” nonpartisanship would mean the absence not only of parties which contest either provincial or national elections but also of all slate-group activity. Unpublished findings on the vote-clustering of candidates in previous elections, in studies carried out by James G. Allen, indicate that voters must have perceived a common association among candidates sponsored by the same group.

³ The reasons usually cited for this indifference are the absence of compelling issues and sharp controversy; the attitude of “this is administration not politics”; the scarcity of adequate information defining the issues and linking them with candidates; and the institutions, particularly at-large elections of council members and the absence of competing parties. For amplification, see Banfield, Edward C. and Wilson, James Q., City Politics (New York, 1968)Google Scholar, chaps. 2 and 16, and Hawley, Willis D., “The Partisan Bias of Nonpartisanship,” PHD thesis, University of California at Berkeley, 1970.Google Scholar

⁴ Studies pointing to the difference between municipal and higher-level voting turnout and analysing municipal voting rates are Alford, Robert R. and Lee, Eugene C., “Voting Turnout in American Cities,” American Political Science Review, 62 (September 1968), 796–813CrossRefGoogle Scholar; and Goldsmith, M. J., “Political Structure and Voter Turnout in Local Elections in Ontario,” unpublished paper presented to the Annual Meeting of the Canadian Political Science Association, Montreal, June 1972.Google Scholar

⁵ For studies noting this tendency, often ascribed to the impact of nonpartisanship itself, see Charles Adrian, R., “Some General Characteristics of Nonpartisan Elections,”. American Political Science Review, 46 (1952), 766–76CrossRefGoogle Scholar; and the evidence in Gilbert, Charles E. and Clague, Christopher, “Electoral Competition and Electoral Systems in Large Cities,” Journal of Politics, 24 (1962), 323–49CrossRefGoogle Scholar, which suggests that at-large electoral systems may be more fundamental to this pattern than nonpartisanship itself. Contrary evidence is provided by Lee, Eugene C., The Politics of Nonpartisanship (Berkeley and Los Angeles, 1960), 66.Google Scholar

⁶ When Bodie first announced the slate, in early September 1968, there were only seven aldermen on the ticket. Later they gained another one, bringing their complement to eight.

⁷ Edmonton Journal, 6 September 1968, 3.

⁸ Ibid., 10 October 1968, 5.

⁹ When asked whether they favoured or opposed having political parties in local elections in Edmonton, only 13 per cent of the respondents in my sample endorsed the idea, while 59 per cent opposed it, and another 29 per cent had no opinion. A postelection survey of the 1971 Alberta provincial election, directed by Professor Richard E. Baird, found that only 25 per cent of the province-wide sample favoured “having political party type organizations run candidates in municipal elections.” For a spirited argument in favour of partisan politics in Canadian cities, see Lightbody, James, “The Rise of Party Politics in Canadian Local Elections,” Journal of Canadian Studies, 6 (February 1971), 39–44CrossRefGoogle Scholar; and for the United States, see Hawley, “The Partisan Bias.”

¹⁰ Though its endorsement both of Bodie and the entire slate of UCAP candidates for city council was pretty lukewarm, the paper did recommend them (and no other candidates) to the voters the day before the election. See the editorial of 15 October 1968.

¹¹ Average votes received by aldermanic candidates were as follows: 1 / all incumbents (N = 7), 42,527; 2 / UCAP incumbents and former office-holders (N = 5), 41,866; 3 / all incumbents and former officeholders (N = 10), 40,381; 4 / all UCAP candidates (N = 8), 36,546; 5 / UCAP newcomers (N = 3), 27,679; 6 / all candidates (N = 32), 25,695; 7 / other candidates not included in 1–5 (N = 19), 17,695. As we can see, incumbency as such was a major ingredient in vote-getting; but while incumbency was a key factor in the relative success of UCAP candidates, sponsorship by the group also contributed something, as comparisons between 5 and 7, and 2 and 3 show.

¹² “Political Participation,” International Encyclopedia of the Social Sciences, ed. Sills, David L. (New York, 1968), vol. 12, 261.Google Scholar

¹³ “Communications and Public Opinion,” Communications in Modern Society, ed. Schramm, Wilbur (Urbana, 1948), 178.Google Scholar

¹⁴ Campbell, Angus et al., Elections and the Political Order (New York, 1966)Google Scholar, “Information Flow and the Stability of Partisan Attitudes,” 136–57; “Attitudes and Non-Attitudes: Continuation of a Dialogue,” The Quantitative Analysis of Social Problems, ed. Tufte, Edward R. (Reading, 1970), 168–89Google Scholar; and “The Nature of Belief Systems in Mass Publics,” Ideology and Discontent, ed. Apter, David E. (New York, 1964), 206–61.Google Scholar

¹⁵ “The Effect of Communication on Voting Behavior,” The Science of Human Communication: New Directions and New Findings in Communication Research, ed. Schramm, Wilbur (New York, 1963), 136.Google Scholar

¹⁶ Many other factors besides these bear on this, of course. For a general statement on mediating conditions, see Klapper, Joseph T., “Mass Communication, Attitude Stability, and Change,” Attitude, Ego-Involvement, and Change, ed. Sherif, Carolyn W. and Sherif, Muzafer (New York, 1967), 297–310Google Scholar, which summarizes much of Klapper's, larger work, The Effects of Mass Communication (Glencoe, 1960)Google Scholar. Personality factors affecting “persuasibility” or susceptibility to influence are discussed and some pertinent findings presented in Linton, Harriet and Graham, Elaine, “Personality Correlates of Persuasibility,” Personality and Persuasibility, eds. Hovland, Carl I. and Janis, Irving L. (New Haven, 1959)Google Scholar; and McGuire, William J., “Personality and Susceptibility to Social Influence,” Handbook of Personality Theory and Research, ed. Borgatta, Edgar F. and Lambert, William W. (Chicago, 1968).Google Scholar

¹⁷ Negroes and the New Southern Politics (New York, 1966), 258.

¹⁸ “Television and Voting Turnout,” Public Opinion Quarterly, 29 (Spring 1965), 77.

¹⁹ See especially “Information Flow and the Stability of Partisan Attitudes,” in which he merely relates stability of voting according to party identification and media usage. Dreyer, Edward C., “Media Usage and Electoral Choices: Some Political Consequences of Information Exposure,” Public Opinion Quarterly, 35 (Winter 1971–2), 544–53CrossRefGoogle Scholar, refutes some of Converse's findings, but, like Converse, he does not attempt to use a more appropriate design for testing the original formulation.

²⁰ Who Governs? (New Haven, 1961). In a survey of voting in an election on a new city charter, he and his associates found that “among those… who knew of the Register's opposition, twice as many voted against the charter as among those who did not know how the paper stood” (pp. 266–7). He does speculate about the effects of issue salience and the anchoring of voting preferences “in well-established attitudes” (p. 266), as well as that of the ambivalent position of the Democratic party. But none of these assumptions are tested, and he does not attempt to assess the differential impact of the paper on voters of varying orientations to politics.

²¹ “Attitudes and Non-Attitudes,” 184 (emphasis in original).

²² The survey was carried out shortly after the 16 October 1968 municipal election. Interviewing, carried out by trained interviewers provided by Verne Hardman and Associates, started about two weeks after the election and was completed about three weeks later. The sample was drawn in the following manner. First, the 276 polls of the city were stratified according to socio-economic and ethnic characteristics of the census tract in which they fell, and then from each stratum a random sample of polls was chosen. Then from each of these polls a random sample of voters was drawn from the enumeration list. If originally designated respondents refused to be interviewed or could not be reached after several call-backs, substitutes chosen randomly from the enumeration list of the poll were provided. The eventual sample totaled 255.

²³ In earlier versions of the paper, this variable was trichotomized. While dichotomization has its obvious disadvantage, it has the advantage of facilitating discussion and presentation of findings in terms of percentages, and it affects the findings below only slightly.

²⁴ For example, the average number of votes won by such candidates on the UCAP ticket (N = 5) was 41,866, compared to 38,896 for similar types not sponsored by the group (N = 5).

²⁵ There were several matters on which plebiscites were being held: whether the city should be divided into wards (but with no specific plan being presented); whether city council should regulate store hours; and whether a cost study should be made of a convention and sports centre. Except for the regulation of store hours, none of the plebiscite issues was particularly controversial at that time; and the store hours issue does not seem to have been a divisive one among the candidates. The Edmonton Journal favoured doing away with the regulation of store hours and the “omniplex” study and opposed the institution of a ward system. The voters endorsed the position that store hours should not be regulated, that the cost study should be undertaken, and that a ward system should be instituted.

²⁶ Some of the correlates (gamma) of level of interest and cognitive competence (using the full range of values for these variables) with indicators of media exposure, broader political interest, and other attitudes are:

²⁷ Only about 39 per cent of eligible voters actually voted in the 1968 election, but 62 per cent of the respondents reported having voted. The sample is therefore unrepresentative in this sense, or respondents were giving socially acceptable responses to questions about voting participation. Whether this biases the results here is a moot question, but I do not think it does.

²⁸ Other factors were associated with voting in the election and for UCAP candidates. Respondents who said they knew candidates personally were more likely to have voted in the election, and also to have cast higher proportions of their votes for UCAP candidates. UCAP-voting was higher among voters who approved of slate groups as compared to those who did not. And, needless to add, there is a definite correlation between voting for Bodie for mayor and UCAP-voting. But, though all of these additional factors were associated with voting preferences, none of them seems to invalidate or override the findings reported below.

²⁹ Media effects on information level of voters parallel those reported for all respondents in Table I, though the correlations tend to be somewhat lower in strength. The weaker correlations are the result of having eliminated many of the less attentive and cognitively less competent respondents from the analysis; these respondents are, of course, the ones assumed to be most subject to influence from the media. But including data on voters only does not add much to the table, so, in the interest of economy, these correlations have been left out.

³⁰ Nor did attitudes toward the introduction of a ward system in Edmonton, opposed by the Journal, seem to be related to the frequency of reading of editorials. Respondents were also asked whether they generally agreed or disagreed with the position of the paper in its local editorials, but the impact of editorial-reading was no greater among those who agreed with them than among those who disagreed with them.

³¹ While not necessarily to be taken as proof of their having been. manipulated, there is a suggestion that these voters could not discern and follow up their interests or else that they held attitudes that were not at all consistent with the pattern of their voting. Except for the most undiscerning voter, it would have been hard to avoid perceiving the basic business orientation of the UCAP.

Yet, while for all other subgroups voting for UCAP candidates was inversely correlated with the belief that business had too much influence in politics in Edmonton, for the less interested and cognitively less competent voters there was actually a slight positive association between the two (gamma = .22, NS).