1 LeDuc, Lawrence, “Canada: The Politics of Stable Dealignment” in Dalton, Russell J., Flanigan, Scott C. and Beck, Paul Allen (eds.), Electoral Change in Advanced Industrial Democracies (Princeton: Princeton University Press, 1984), 403.Google Scholar

2 Sniderman, Paul M., Forbes, H. D. and Melzer, Ian, “Party Loyalty and Electoral Volatility: A Study of the Canadian Party System,” this JOURNAL 7 (1974), 268–88.Google Scholar

3 LeDuc, “Canada: The Politics of Stable Dealignment.”

4 Jenson, Jane, “Party Loyalty in Canada: The Question of Party Identification,” this JOURNAL 8 (1975), 543–55Google Scholar; Clarke, Harold D., Jenson, Jane, LeDuc, Lawrence, and Pammett, Jon H., Political Choice in Canada (Toronto: McGraw-Hill, 1979)Google Scholar, chap. 5, and, by the same authors, Absent Mandate: The Politics of Discontent in Canada (Toronto: Gage, 1984), 56–62, 73;Google ScholarLeDuc, L., Clarke, H. D., Jenson, J., Pammett, J. H., “Partisan Instability in Canada: Evidence from a New Panel Study,” American Political Science Review 78 (1984), 470–84.CrossRefGoogle Scholar

5 Jenson, Jane, “Comment: The Filling of Wine Bottles is Not Easy,” this JOURNAL 11 (1978), 437”46Google Scholar, and Clarke, et al., Absent Mandate, 69.Google Scholar

6 See the summary of contending theories of partisan dealignment in the editors' introduction to Dalton, et al., Electoral Change, 2–23.Google Scholar

7 An excellent summary of this argument is Inglehart, Ronald, “The Changing Structure of Political Cleavages in Western Society,” in Dalton, et al., Electoral Change, 25–69.Google Scholar

8 See the emphasis on this issue in Borre, Ole, “Critical Electoral Change in Scandinavia,” in Dalton, et al., Electoral Change, 355–61.Google Scholar

9 See the concise statement of the nature of the brokerage party system in Clarke, et al., Absent Mandate, 10–16.Google Scholar

10 Sniderman et al., “Party Loyalty and Electoral Volatility,” explicitly attack the disintegrative consequences of the brokerage party system and its failures to produce policy innovation in ways which closely parallel the critique a decade later by Clarke et al. (Absent Mandate).

11 The central proposition in the early research on party identification has been expressed as follows: “The more electors are attached by enduring psychological links to political parties…the more the polity is insured against ‘flash parties’ and sudden demagogic incursions.” See Budge, I., Crewe, I., and Fairlie, D. (eds.), Party Identification and Beyond (Toronto: Wiley, 1976), 3.Google Scholar In these terms, the United States was interpreted as a model of democratic stability, and countries like France, where long-term political and economic changes had not yet produced such stable conditions, were seen to be converging towards such conditions in the postwar era. See Converse, Philip, “Of Time and Partisan Stability,” Comparative Political Studies 2 (1969), 139–71.CrossRefGoogle Scholar The dramatic decline in the incidence and intensity of party identification among Americans after 1964 has, however, prompted a substantial critical revision of this perspective. In the United States itself, parties have been seen less as a model for other democratic polities than as a symptom of domestic political troubles, or as “one component of the malperforming engine of governance in the United States.” See Ladd, E. C., Where Have All the Voters Gone? (New York: Norton, 1978), xvi.Google Scholar In other advanced capitalist countries, the intensity of party identification has never been as strong as in the heyday of American party democracy, and it has similarly declined. According to P. A. Beck, “party loyalties are more instrumental elsewhere and tend to be less distinguishable from vote choice at any particular time. The rules of the electoral game in these systems—for example, multipartyism, parliamentarism, short and infrequent ballots, even proportional representation and party lists—reduce the need for a separate party loyalty that can remain intact in the face of vote defection. More primordial group loyalties (e.g. to a social class or a religious grouping) instead anchor the voter in the electoral arena, making it difficult for the parties themselves to become objects of psychological identification” (Introduction to Part Four in Dalton, et al., Electoral Change, 234Google Scholar). In typically hybrid fashion, neither the psychological nor group loyalties are strong in Canada.

12 Elkins, David J., “Party Identification: A Conceptual Analysis,” this JOURNAL 10 (1978), 423.Google Scholar

13 Elkins' argument is in fact quite different. He sees party identification as constituting a long-term, stabilizing effect upon the vote, with short-term factors like campaign and leadership characteristics influencing the residual vote switching, and a lagged adjustment of partisanship to vote. Jenson properly points out that this formulation evades “the logical requirement, the necessity of stable partisan ties” in the original Michigan theory, and that in this form it tends to be circular and unfalsifiable. It holds where voting stability is observed (long-term forces outweigh short-term) and where voting instability is observed (short-term forces outweigh long-term). According to Jenson, “if the argument is to be theoretically viable it is necessary to provide an independent demonstration, rather than merely an assertion, that short-term forces are the reason for the difference” (“Comment,” 441).

14 Jenson, “Party Loyalty in Canada,” 544, and “Comment,” 443.

15 Documentation of the greater influence of issues on the voting choices of unstable partisans is given in Clarke, et al., Absent Mandate, 94, 131–33Google Scholar, and in Clarke, et al., Political Choice in Canada, 346–47.Google Scholar

16 According to Sniderman et al., “if Canadian party loyalty is not weakly rooted and fickle in the way commonly assumed, major parties can advance innovative economic and social policies without courting electoral disaster or the disruption of the party system. At a minimum, we see no evidence of irresistible pressures on the major parties to pursue similarly centrist policies. In short, the textbook theory may be misleading not only about what is the case, but also about what can and what ought to be the case” (“Party Loyalty and Electoral Volatility,” 288). This conclusion rests, of course, upon acceptance of the opening conditional clause.

17 Jenson, “Party Loyalty,” 553.

18 See the historical detail in Brodie, M. Janine and Jenson, JaneCrisis, Challenge and Change: Party and Class in Canada (Toronto: Methuen, 1980)Google Scholar, and the historical context of recent partisan instability given in Clarke et al., Absent Mandate, chap. 1.

19 See, for example, the discussion of the Toronto-based reform movement which worked in the 1960s to reorient the Liberal party under the leadership of Pearson and the policies of Gordon in McCall-Newman, Christina, Grits (Toronto: Macmillan, 1982)Google Scholar, and the discussion of ideologically motivated reform groups within the Conservative party leading up to the selection of Mulroney as party leader in Martin, PatrickGregg, Allan and Perlin, George, Contenders: The Tory Quest for Power (Scarborough: Prentice-Hall, 1983)Google Scholar. Evidence for the ideological distinctiveness of the two major federal parties, and for the influence of ideology on internal party dynamics is contained in Stevenson, H. M. “Ideology and Canadian Party Leadership Selection: No Relationship or Type I Error” (Toronto: York University, Institute for Social Research, January 1986)Google Scholar. For a general history of this topic see Christian, William and Campbell, Colin, Political Parties and Ideology in Canada (2nd ed.; Toronto: McGraw-Hill Ryerson, 1983).Google Scholar

20 The 1984 National Election Study has, however, incorporated a great deal more information on class-related questions of ideology and political attitude.

21 Documentation on these surveys is contained in Social Change in Canada: Technical Documentation (Toronto: York University Institute for Behavioural Research, June 1984).Google Scholar

22 See the titles and substance of the successive annual reports of the Economic Council of Canada for 1979 to 1982: Two Cheers for the 1980's, A Climate of Uncertainty, Room for Manoeuvre, and Lean Times (Ottawa: Minister of Supply and Services Canada).Google Scholar

23 These changes are the subject of a growing library of recent political books by Canadian journalists. They include Gwyn, Richard, The Northern Magus (Toronto: McClelland and Stewart, 1980)Google Scholar; Simpson, Jeffrey, Discipline of Power (Toronto: Personal Library, 1980)Google Scholar; McCall-Newman, Grits; Sheppard, Robert and Valpy, Michael, The National Deal (Toronto: Fleet Books, 1982)Google Scholar; Martin et al., Contenders; Fraser, Graham, P.Q.: René Lévesque and the Parti Québécois in Power (Toronto: Macmillan, 1984)Google Scholar; Snider, Norman, The Changing of the Guard (Toronto: Lester and Orpen Dennys, 1985)Google Scholar; Gwyn, Richard, The 49th Paradox (Toronto: McClelland and Stewart, 1985)Google Scholar; and Graham, Ron, One-Eyed Kings (Toronto: Collins, 1986)Google Scholar. This list excludes the less serious, if seriously intended, anthologies of short pieces by such journalists such as Alan Fotheringham and Charles Lynch.

24 Our results are confirmed by the spread in vote intentions reported by major polling firms in the spring of 1981, which gave a 15 per cent advantage to the Liberals. In the summer of 1981, however, the advantage shifted so that a year later the Conservatives enjoyed a 10 per cent lead, which grew to more than 20 per cent shortly after Brian Mulroney's election as leader in the summer of 1983. Liberal support then increased until the party was again even with the Conservatives in the spring of 1984. See summary poll data in the Globe and Mail, August 24, 1984.

25 See the discussion and references to alternative conceptualizations of class in Ornstein, Michael D., Stevenson, H. Michael and Williams, A. Paul, “Region, Class and Political Culture in Canada,” this JOURNAL 13 (1980), 227–72Google Scholar. For the importance of distinctions between established and non-established labour forces, see R. Cox, “Employment, Labour and Future Political Structures,” in Byers, R. and Reford, R.W. (eds.), Canada Challenged (Toronto: Canadian Institute of International Affairs, 1979), 262–92.Google Scholar For the coding details on the component occupational groups in our class categories, see Peter C. Pineo, “Revision of the Pineo-Porter-McRoberts Socio-Economic Classification of Occupations for the 1981 Census” (QSEP Research Report #125, McMaster University, Program for Quantitative Studies in Economics and Population, February 1985).

26 See Cairns, Alan C., “The Electoral System and the Party System in Canada, 1921-1965,” this JOURNAL 1 (1968), 55–80Google Scholar, and Smith, D. E., The Regional Decline of a National Party: Liberals on the Prairies (Toronto: University of Toronto Press, 1981).Google Scholar

27 The impact of new voters is difficult to estimate. See the discussion in Clarke, et al., Absent Mandate, 153–57Google Scholar, where it is argued (without very convincing documentation) that new voters tended disproportionately to support the Liberals, but that this tendency became less noticeable in the 1974, 1979 and 1980 elections.

28 The panel of 1,665 respondents is slightly biased by attrition from the original 1977 national sample toward more women, youth, and well-educated respondents and fewer unmarried and Quebec respondents. The panel slightly overrepresents identification with each of the three major parties, and underrepresents those with no or unknown partisanship. This suggests that the panel slightly overrepresents the “attentive public,” and that it should give reliable estimates of partisan shifts and their correlates. More detailed information on the panel sample is available from the author.

29 These differences may also reflect an increase in partisan instability due to the worsening economic crisis of 1981, and the inability of the government to follow through on its 1980 campaign promises of economic recovery. Further, our estimates may be inflated slightly because of the greater length of time between our surveys and actual elections. The overall figure of partisan instability in the national election panel is taken from Clarke, et al., Absent Mandate, Table 3.3, 67Google Scholar, the figures for changes between the 1974and 1979 elections and the 1979 and 1980 elections from Table 3.1, 62. These estimates are apparently based on calculations that exclude missing data on party identification, although there is no clear indication of how missing data were treated apart from the different N's reported in different tables for the three-wave panel. If we had excluded respondents with “don't know” or missing data on the party identification item from our operationalization of partisan instability, our estimate would have been very close to that of the National Election Study. As shown in Table 4 below, elimination of missing data in the 1977 wave of the panel reduces our estimate from 56 per cent to 50 per cent, and elimination of additional missing data from other waves would bring our estimate closer to 45 per cent. We felt it conceptually more appropriate, however, to include a non-response to the question of partisan identity in the measure of partisan instability.

30 See references and data in LeDuc, “Canada: The Politics of Stable Dealignment,” and LeDuc et al., “Partisan Instability in Canada.”

31 This does not mean that new politics issues are of less interest to younger as opposed to older Canadians, or that younger Canadians are less supportive of new social movements, self-actualization goals, and so forth.

32 The size of the group who confine their changes of partisanship to the Liberal and New Democratic parties is remarkable. The group includes those who shift between the two parties and those who shift to only one of them from “independent” or “don't know” categories. Clarke, et al., Absent Mandate, Table 3.3Google Scholar, indicate that a maximum of 24.8 per cent of their 1974–1980 panel would be in this group if it included all “homing” Liberals and NDP identifiers, all “movers” to the Liberal and New Democratic parties, all non-identifiers, and all third-wave changers. Their estimate would be lower adjusting for the exclusion of conversions to the Conservatives from the last two groups. Our estimates would not be substantially different, however, if we adjusted for the differences in operationalization discussed in footnote 29.

33 Key, V. O. Jr., The Responsible Electorate (New York: Random House, 1968), 58–59.Google Scholar

34 We would expect, that is, to find a much larger group of all-party switchers after the 1984 election and to find, as in our data for the earlier period, a disproportionate number of relatively left-leaning voters in this group. Evidence from the 1984 National Election Study confirms this expectation: “If there was a pattern in the stated attitudes of the two groups of voters, it was that those who abandoned Turner and the Liberal Party for the Conservatives were more ‘libertarian’ in their outlook than were loyal Liberals” (Kay, Barry J., Brown, Steven D., Curtis, James E., Lambert, Ronald D., and Wilson, John M., “Some Aspects of Electoral Change in 1984,” in Fox, Paul W. and White, Graham [eds.], Politics: Canada [6th ed.; Toronto: McGraw-Hill Ryerson Limited, 1987], 434).Google Scholar

35 Gwyn, The 49th Paradox, and Graham, One-Eyed Kings.