1 Party identification may also affect political attitudes and issue positions. The evidence on these points, however, is less extensive, so I shall focus almost wholly on its effect on voting.

2 Budge, Ian, Crewe, Ivor, and Farlie, Dennis (eds.), Party Identification and Beyond (London: John Wiley and Sons, 1976)Google Scholar, especially chaps. 1, 2, 7, 14, and 20, provide a comprehensive review of many important issues. See also, Brody, Richard, “Change and Stability in the Components of Partisan Identification,” DEA News, no. 13 (1977), 13–18Google Scholar.

3 Kelley, Stanley Jr., and Mirer, Thad W., “The Simple Act of Voting,” American Political Science Review 68 (1974), 572–91CrossRefGoogle Scholar; and Jenson, Jane, Clarke, Harold, LeDuc, Lawrence, and Pammett, Jon, “Patterns of Partisanship in Canada: Split Identification and Cross-Time Variation,” paper delivered to the American Political Science Association meetings,San Francisco, September 1975Google Scholar.

4 Jenson, Jane, “Party Loyalty in Canada: The Question of Party Identification,” this Journal 8 (1975), 543–53Google Scholar.

5 Nie, Norman, et al. , The Changing American Voter (Cambridge: Harvard University Press, 1976)Google Scholar, Appendix 4. especially 391.

6 Jenson, et al., “Patterns of Partisanship in Canada,” 10, 18.

7 Kelley and Mirer show this for the United States; and Jenson, et al., do so for Canada, although using a somewhat different approach.

8 Jenson, “Party Loyalty in Canada,” 548–49.

9 Ibid., 553. It is perhaps unfair to single out one person, but space does not permit the multiplication of examples.

10 Major sources for the theory of party identification are Campbell, Angus, et al. , The Voter Decides (Evanston: Row, Peterson, 1954)Google Scholar; Campbell, , et al. , The American Voter (New York: Wiley, 1960)Google Scholar; Campbell, , et al. , Elections and the Political Order (New York: Wiley, 1966)Google Scholar; and Converse, Philip E., The Dynamics of Party Support: Cohort-Analyzing Party Identification (Beverly Hills: Sage, 1976)Google Scholar.

11 The original statement on reference groups was Hyman, Herbert, “The Psychology of Status,” Psychological Monographs 53, No. 269 (1941)Google Scholar. Classic works have been reprinted in Hyman, H. and Singer, E. (eds.), Readings in Reference Group Theory and Research (New York: Free Press, 1968)Google Scholar.

12 It is an open question as to the basis of affective ties to a party: childhood or adult socialization, rational choice, information economizing, “style” and personality of party leadership, and others. I shall offer a few speculations on this topic in the conclusion.

13 But note that there seems to be little difference, despite anecdotes to the contrary, between the amount of volatility in the Canadian and American party systems according to the recent comparative work of Sniderman, Paul, Forbes, Donald, and Melzer, Ian, “Party Loyalty and Electoral Volatility: A Study of the Canadian Party System,” this Journal 7 (1974), 268–88Google Scholar.

14 Dobson, Douglas and Angelo, Douglas St., “Party Identification and the Floating Vote: Some Dynamics,” American Political Science Review 69 (1975)CrossRefGoogle Scholar. Tables 2, 3, 4, present data on switching toward and away from the respondents' party of current identification for American presidential and congressional elections in 1956, 1958, and 1960.

15 Besides the data summarized here, see also Converse, Philip, “Of Time and Partisan Stability,” Comparative Political Studies 2 (1969), 139–71CrossRefGoogle Scholar; Kubota, Akira and Eldersveld, Samuel, “Party Identification in India and Japan—In the Context of Western Theory and Research,” paper delivered at the Canadian Political Science Association meetings,Montreal, August 1973Google Scholar; and Butler, David and Stokes, Donald, Political Change in Britain (New York: St. Martin's, 1969)Google Scholar.

16 Jenson, “Party Loyalty in Canada,” 546–48, presents a careful analysis which also concludes that partisans “clearly differentiate their current vote from some longerstanding support for a particular political party.”

17 See, for example, McPhee, William and Glaser, William (eds.), Public Opinion and Congressional Elections (New York: Free Press, 1962)Google Scholar; and Philip E. Converse, “Information Flow and Partisan Stability,” in Campbell, et al., Elections and the Political Order.

18 An additional source of these absolute differences concerns the fact that, except for the 1974 Canadian federal election study, we do not know whether respondents are reporting switching at only the federal level or between federal and provincial levels. The substantial switching between levels and different proportions of identification at each level may inflate the appearance of instability in Canada. See Jenson, et al., “Patterns of Partisanship in Canada.”

19 For recent data on the United States, see Nie, et al., The Changing American Voter; and Converse, The Dynamics of Party Support.

20 Jenson, Jane, “Party Strategy and Party Identification: Some Patterns of Partisan Allegiance,” this Journal 9 (1976), 27–48Google Scholar; and LeDuc, Lawrence et al. , “Partisanship, Political Interest, and Electoral Campaigns in Canada,” paper delivered at the International Political Science Association meetings,Edinburgh, Scotland, August 1976Google Scholar. Jenson apparently does not consider party strategy or the charismatic qualities of John Diefenbaker to be part of the “short-term” forces influencing voters to defect. She says: “This realignment occurred without a great deal of fanfare as a result, not of important national events, but of strategic calculations by the leadership of one particular party” (31). Yet she then (32–33) outlines the massive changes (called a “cataclysm”) Diefenbaker instituted, including altering voters' expectations, calling into question “traditional Tory ideals by advocating policies which appealed to a new group of potential supporters,” because of which “a full-scale battle was fought between the old and the new guard.…” Surely, such intentions and actions constitute major forces on voters, along the lines I have just outlined.

21 For a general discussion, see Converse, “Of Time and Partisan Stability,” and Kubota and Eldersveld, “Party Identification in India and Japan”; for Canada, see Irvine, William, “Explaining the Brittleness of Partisanship in Canada,” paper delivered at the Canadian Political Science Association meeting,Edmonton, June 1975Google Scholar.

22 Meisel, John, Working Papers on Canadian Politics (Montreal: McGill-Queens Press, 1973), 67CrossRefGoogle Scholar, argues as follows: “The concept of Party Identification, as used by scholars associated with the Michigan Survey Research Center…may be almost inapplicable in Canada. On the basis of admittedly limited evidence…we have found that party identification seems to be as volatile in Canada as the vote itself.”

23 Jenson presents the most explicit version: “When the relationship between party identification and vote is examined, it is found to be strong. In fact, its very strength is suspicious.” (“Party Loyalty in Canada,” 545.) See also Jenson, et al., “Patterns of Partisanship in Canada,” 48: “It (party identification) has been found not to embody some of the requisite characteristics, in particular stability over time, and to be too closely related to behaviour, particularly the vote” (emphasis added).

24 See the discussion in Dobson and St. Angelo, “Party Identification and the Floating Vote.” 487–88, and Table 6.

25 Sniderman, “Party Loyalty and Electoral Volatility”; and LeDuc, et at., “Partisanship.”

26 This is contrary to what David Robertson states in “Surrogates for Party Identification in the Rational Choice Framework,” in Budge, et al. (eds.), Party Identification and Beyond: “… party identification explains most votes at any one election … while ill-defined “forces” are then brought in to explain those who cannot be explained by this method” (369). In fact, these “forces” are well-defined, and they include issues, candidates, government performance, and the like. Jenson, “Party Strategy and Party Identification,” focusses primarily on the de-stabilizing forces of the Diefenbaker period in Canada. Her article is a good example of how attention to specific, plausible, well documented forces can explain the higher than normal levels of vote defection and changes of party identification. For the period 1972–74, see LeDuc, et al., “Partisanship.”

27 When large numbers of electors change their traditional allegiance to a party, the election or series of elections is called “critical.” V. O. Key first stated this view (“A Theory of Critical Elections,” Journal of Politics 17 [1955], 3–18)CrossRefGoogle Scholar; Burnham, Walter Dean, in Critical Elections and the Mainsprings of American Politics (New York: Norton, 1970)Google Scholar, has refined the concept. By definition, most elections are not critical; hence a focus on change nevertheless underscores the more frequent stability.

28 Subtle differences in the cues available to voters can plausibly account for some variation in the stabilizing effects of party identification. For example, countries, provinces (or states), and periods differ in the complexity of ballots and in whether candidates' party labels appear on the ballot.

29 When there are no countervailing forces, the vote will equal the division of party identifications. This is the definition of a “normal vote”; see Philip Converse, “The Concept of a Normal Vote,” in Campbell, et al., Elections and the Political Order, chap. 2.

30 Jenson, “Party Strategy and Party Identification.”

31 Sniderman, et al., “Party Loyalty and Electoral Volatility,” 285.

32 Jenson, “Party Loyalty in Canada,” 551 (emphasis added); this point bears emphasis because she alleges in the paragraph before the one quoted that partisan transmission is “no better than chance for all parties taken together.…”

33 For the justification for this approach, consult any standard statistics text, such as Wonnacott, T. H. and Wonnacott, R. J., Introductory Statistics for Business and Economics (New York: Wiley, 1972), 66–70Google Scholar. 131–32, 174–76.

34 In the examples given, I have used Democrats in the United States and Liberals in Canada, but the Z-scores are highly significant for all parties in both countries, and for mother's party identification as well where that has been ascertained.

35 Pammett, Jon H., “The Development of Political Orientations in Canadian School Children,” this Journal 4 (1971), 132–41Google Scholar; Komberg, Allan, et al. , “Some Differences in the Political Socialization Patterns of Canadian and American Party Officials: A Preliminary Report,” this Journal 2 (1969), 64–88Google Scholar; and Smith, Joel, et al. , “Patterns of Early Political Socialization and Adult Party Affiliation,” Canadian Review of Sociology and Anthropology 5 (1968), 123–55CrossRefGoogle Scholar.

36 Jenson, et al., “Patterns of Partisanship in Canada;” LeDuc, et al., “Partisanship.”

37 Converse, The Dynamics of Parly Support; and Dobson and St. Angelo, “Party Identification and the Floating Vote,” document this process for the United States.

38 In Budge, et al., Party Identification and Beyond, one may find examples of these and other orientations.

39 For the distinction between these types of variables, see MacCorquodale, Kenneth and Meehl, Paul E., “On a Distinction Between Hypothetical Constructs and Intervening Variables,” Psychological Review 55 (1948), 95–107CrossRefGoogle ScholarPubMed.

40 Warren Miller, “The Cross-National Use of Party Identification as a Stimulus to Political Inquiry,” in Budge, et al. (eds.). Party Identification and Beyond, 28.

41 Ibid., 24.