1 Harold D., ClarkeJenson, JaneLeDuc, Lawrence and Pammett, John H.Political Choice in Canada (abridged edition; Toronto:McGraw-Hill Ryerson 1980), 97, Table 5.1.Google Scholar

2 Ibid 107.

3 Ibid202.

4 Ibid 206.

5 Ibid 208-09, 212.

6 The data on British Columbia were obtained by a multi-stage, stratified cluster sample survey conducted following the 1979 provincial election and the federal elections of 1979 and 1980 (N = 1,051) by Canadian Facts Ltd., on behalf of the principal investigators:Elkins, David J.Blake, Donald E. and Johnston, Richard. Foxr details of the survey see Donald E. Blake, David J. Elkins, and Richard Johnston, “Sources of Change in the B.C. Party System, ” B.C. Studies No. 50 (1980), Appendix.Google Scholar

7 Data from the 1974 study were made available by the University of British Columbia Data Library. Neither the Data Library nor the principal investigators are responsible for the use made of the data in this analysis.

8 “A Multidimensional Concept of Party Identification, ” Political Behavior 2 (1980), 33-60. Other significant contributions are by Katz, Richard “The Dimensionality of Partisanship: Cross-National Perspectives, ” Comparative Politics 11 (1979), 147–63;Google Scholar and Valentine, David C. and Van Wingen, John R. “Partisanship, Independence, and the Partisan Identification Question, ” American Politics Quarterly 2 (1980), 165–86.CrossRefGoogle Scholar

9 Weisberg, “A Multidimensional Concept of Party Identification, ” 36, 39.

10 Former provincial Liberal and Conservative identifiers shifting to Social Credit are much more likely than those shifting to the NDP to cite negative features of the NDP as a reason for their current party identification. Moreover, “negative reasons” were cited with greater frequency than any other category of reason. Blake, SeeJohnston, and Elkins, “Sources of Change in the B.C. Party System, ” 12–14.Google Scholar and Maggiotto, Michael A. and Piereson, James E. “Partisan Identification and Electoral Choice: The Hostility Hypothesis, ” American Journal of Political Science 21 (1977), 745–67.CrossRefGoogle Scholar

11 The following discussion is based on the work of Harold Guetzkow, Multiple Loyalties: Theoretical Approach to a Problem in International Organization (Princeton: Center for Research on World Political Institutions, Princeton University, 1955). Other studies of cross pressures are available, but Guetzkow's work is closer in spirit to dualism based on the federal-provincial divisions in Canada. I am indebted to Peter Busch for drawing it to my attention. The four strategies described below are based on Guetzkow's distinctions between re-definition, compartmenting, narrowing of function, and nonperception of conflict.

12 IbidThe following discussion is based on the work of Harold Guetzkow, Multiple Loyalties: Theoretical Approach to a Problem in International Organization (Princeton: Center for Research on World Political Institutions, Princeton University, 1955). Other studies of cross pressures are available, but Guetzkow's work is closer in spirit to dualism based on the federal-provincial divisions in Canada. I am indebted to Peter Busch for drawing it to my attention. The four strategies described below are based on Guetzkow's distinctions between re-definition, compartmenting, narrowing of function, and nonperception of conflict,Ibid 43.

13 IbidGuetzkow calls this “compartmenting, ” where “the contradictions in behavior demanded by two loyalties are neither severe nor pervasive... ”Ibid.

14 When queried about issues, respondents in the British Columbia survey were able to make assignments of responsibility to different levels of government in certain issue areas, noted areas of shared responsibility, and were able to place the parties in the different issue spaces. For a discussion, see Elkins, David J.“What Is at Issue in Political Issues? ” a paper presented to the annual meeting of the Canadian Political Science Association, Halifax 1981.Google Scholar Also see Clarke, et al., Political Choice in Canada, 118, for a discussion of the “ability of Canadians to distinguish between and relate to each of the two levels of government. ” The unabridged edition of their work contains the observation that “the structural potential exists for the maintenance of two parallel reference groups in the voter's mind” (138). However the authors do not pursue the implication that behaviour in one arena may have little to do with behaviour in the other.Google Scholar

15 Ibid When queried about issues, respondents in the British Columbia survey were able to make assignments of responsibility to different levels of government in certain issue areas, noted areas of shared responsibility, and were able to place the parties in the different issue spaces. For a discussion, see Elkins, David J.“What Is at Issue in Political Issues? ” a paper presented to the annual meeting of the Canadian Political Science Association, Halifax 1981.Google Scholar Also see Clarke, et al., Political Choice in Canada, 118Google Scholar, for a discussion of the “ability of Canadians to distinguish between and relate to each of the two levels of government. ” The unabridged edition of their work contains the observation that “the structural potential exists for the maintenance of two parallel reference groups in the voter's mind” (138). However the authors do not pursue the implication that behaviour in one arena may have little to do with behaviour in the other, Ibid108.

16 Conclusions about the relative importance and statistical significance of the variables used in this study do not differ whether ordinary least squares regression or a technically more appropriate maximum likelihood method such as probit is used.

17 Length of residence is a dummy variable which is “1” for those who have lived in British Columbia for 3 years or less—that is, those arriving since an election was held at either level. Approval of conflict over fundamental ideological differences was also represented by a dummy variable.

18 While differences in the “liking” scores were used in this case, use of “disliking” scores produces similar results. The availability of both liking and disliking scores allows for the investigation of ambivalence and indifference, but that is beyond the scope of the present study. See Jean A. Laponce, “Measuring Party Preference: The Problem of Ambivalence, ” this JOURNAL I1 (1978), 139-52, for a discussion of the theoretical issues.

19 Acceptance of the rhetoric of class conflict is a dummy variable scored “1” for those who believe that elections in BC are contests between “free enterprise” and “socialism. ” This was used in preference to a measure based on “approval” of this kind of conflict on the grounds that it serves as a better indicator of psychological involvement for those likely to be identified with the federal Liberals ands Conservatives. In fact, the “factual” and “normative” judgments about this conflict are highly correlated and produce similar results in the regression.

20 This is a dummy variable scored “1” for those who assign themselves to a social class and who feel that “class is important to how they think of” themselves.

21 A dummy variable scored “1” for those who are willing to vote for a party other than their preferred party if sincere voting might result in a victory by a party they disliked at the riding level, or result in the disliked party forming the government. For an analysis of strategic voting using British Columbia data see Richard Johnston, David Elkins, and Donald Blake, “Strategic Voting: Individual Reasoning and Collective Consequences, ” a paper presented at the annual meeting of the American Political Science Association, Washington, DC, 1980. One of the anonymous reviewers of this article suggested that length of residence might play a role in the adoption of split partisanship on the grounds that some voters might retain their federal partisanship until they became familiar with the provincial arena. This possibility was tested but proved not to be a factor.

22 Their argument is perhaps best illustrated by a quotation from Katz: “The intense partisan is one who thinks about and reacts to politics in terms of parties. This way of relating to political life need not change when the voter's preference changes” (“The Dimensionality of Party Identification, ” 158).

23 Blake, Johnston, and Elkins, “Sources of Change in the B.C. Party System ”.Google Scholar

24 Federal Liberal identifiers have the highest average income in the British Columbia sample. Only 32 per cent of Liberal identifiers willing to accept a class label consider themselves working class. The figures for the Conservatives and NDP are 39 percent and 62 per cent, respectively.

25 Clarke, et al. Political Choice in Canada 107.Google Scholar

26 A cross-tabulation of this variable with data from the 1974 study on those who switched their votes between 1972 and 1974 also undermines its utility. Of those who voted in both elections and claim to have always voted the same way, 15.6 per cent actually report voting differently in the two federal elections.

27 IbidA cross-tabulation of this variable with data from the 1974 study on those who switched their votes between 1972 and 1974 also undermines its utility. Of those who voted in both elections and claim to have always voted the same way, 15.6 per cent actually report voting differently in the two federal elections IbidTable 5.5, 107.

28 IbidA cross-tabulation of this variable with data from the 1974 study on those who switched their votes between 1972 and 1974 also undermines its utility. Of those who voted in both elections and claim to have always voted the same way, 15.6 per cent actually report voting differently in the two federal elections Ibid 99, emphasis added.

29 IbidA cross-tabulation of this variable with data from the 1974 study on those who switched their votes between 1972 and 1974 also undermines its utility. Of those who voted in both elections and claim to have always voted the same way, 15.6 per cent actually report voting differently in the two federal elections.Ibid.

29 Kent Jennings, M. and Niemi, Richard G. “Party Identification at Multiple Levels of Government, ” American Journal of Sociology 72 (1966-67), 94–95.Google Scholar

31 Ibid 87