1 For a sophisticated empirical study of this issue see Laponce, Jean A., Left and Right: The Topography of Political Perceptions (Toronto: University of Toronto Press, 1981), 183–186Google Scholar.

2 Inglehart, Ronald, “New Perspectives on Value Change,” Comparative Political Studies 17 (1985), 487CrossRefGoogle Scholar.

3 Laponce, Left and Right, 135, 206, and Nevitte, Neil, Bakvis, Herman and Gibbins, Roger, “The Ideological Contours of ‘New Politics’ in Canada: Policy, Mobilization and Partisan Support,” this Journal 22 (1989), 502Google Scholar.

4 Inglehart, Ronald and Klingemann, Hans D., “Party Identification, Ideological Preference, and the Left/Right Dimension Among Western Mass Publics,” in Budge, Ian, Crewe, Ivor and Farlie, Dennis, eds., Party Identification and Beyond (London: John Wiley, 1976), 259Google Scholar; Bürktin, Wilhelm P., “The Split Between the Established and the Non-established Left in Germany,” European Journal of Political Research 13 (1985), 289Google Scholar; and Arian, Asher and Shamir, Michal, “The Primarily Political Functions of the Left-Right Continuum,” Comparative Politics 15 (1983), 151CrossRefGoogle Scholar.

5 Lambert, Ronald D., Curtis, James E., Brown, Steven D. and Kay, Barry J., “In Search of Left/Right Beliefs in the Canadian Electorate,” this Journal 19 (1986), 541–563Google Scholar; Gibbins, Roger and Nevitte, Neil, “Canadian Political Ideology: A Comparative Analysis,” this Journal 18 (1985), 577–598CrossRefGoogle Scholar; Johnston, Richard, “The Ideological Structure of Opinion on Policy,” in Perlin, George, ed., Party Democracy in Canada: The Politics of National Party Conventions (Scarborough: Prentice-Hall, 1988)Google Scholar; and Nevitte, Bakvis and Gibbins, “The Ideological Contours of ‘New Politics’ in Canada.”

6 Lambert et al., “Left/Right Beliefs,” 561.

7 A useful contrast of exploratory and confirmatory factor analysis can be found in Bollen, Kenneth A., Structural Equations With Latent Variables (New York: John Wiley, 1989), 226–232Google Scholar.

8 See Converse, Philip E., “The Nature of Belief Systems in Mass Publics,” in Apter, David E., ed., Ideology and Discontent (New York: The Free Press, 1964)Google Scholar; Bishop, George F., “The Effect of Education on Ideological Consistency,” Public Opinion Quarterly 40 (1976)CrossRefGoogle Scholar; and Luskin, Robert C., “Measuring Political Sophistication,” American Journal of Political Science 31 (1987), 856–899CrossRefGoogle Scholar.

9 See, for example, Parkin, Frank, Class Inequality and Political Order: Social Stratification in Capitalist and Communist Societies (London: MacGibbon and Kee, 1971), 81–92Google Scholar.

10 Laponce, Left and Right, 115–37.

11 See Lau, Richard R. and Sears, David O., “Social Cognition and Political Cognition: The Past, the Present, and the Future,” in Lau, Richard R. and Sears, David O., eds., Political Cognition (Hillsdale, New Jersey: Lawrence Erlbaum Associates, 1986), 347–366Google Scholar; and Fiske, Susan T. and Taylor, Shelley E., Social Cognition (New York: Random House, 1984)Google Scholar.

12 Fiske, Susan T., “Schema-Based Versus Piecemeal Politics: A Patchwork Quilt, But not a Blanket, of Evidence,” in Lau, and Sears, , eds., Political Cognition, 44Google Scholar.

13 Ibid., 46.

14 Conover, Pamela Johnston and Feldman, Stanley, “How People Organize the Political World: A Schematic Model,” American Journal of Political Science 28 (1984), 95–126CrossRefGoogle Scholar.

15 This is one view of the relationship between cognition and affect. See Epstein, Seymour, “Controversial Issues in Emotion Theory,” in Shaver, Philip, ed., Review of Personality and Social Psychology, Vol. 5 (Beverly Hills: Sage, 1984)Google Scholar.

16 Luskin, “Measuring Political Sophistication,” 857–58.

17 For examples of these types of schemata see Conover and Feldman, “How People Organize the Political World”; Allen, Richard L., Dawson, Michael C. and Brown, Ronald E., “A Schema-Based Approach to Modeling an African-American Racial Belief System,” American Political Science Review 83 (1989), 421–441CrossRefGoogle Scholar; and Hamill, Ruth, Lodge, Milton and Blake, Frederick, “The Breadth, Depth, and Utility of Class, Partisan, and Ideological Schemata,” American Journal of Political Science 29 (1985), 850–870CrossRefGoogle Scholar.

18 For example, see Peffley, Mark A. and Hurwitz, Jon, “A Hierarchical Model of Attitude Constraint,” American Journal of Political Science 29 (1985), 871–890CrossRefGoogle Scholar.

19 Compare Johnston, “The Ideological Structure of Opinion,” 69, footnote 15.

20 See Sniderman, Paul M. and Tetlock, Philip E., “Interrelationship of Political Ideology and Public Opinion,” in Hermann, Margaret G., ed., Political Psychology (San Francisco: Jossey Bass, 1986), 62–96Google Scholar.

21 Converse, “The Nature of Belief Systems,” 209–14.

22 Ibid., 212–13.

23 Marcus, George E., Tabb, David and Sullivan, John L., “The Application of Individual Differences Scaling to the Measurement of Political Ideologies,” American Journal of Political Science 18 (1974), 406–407CrossRefGoogle Scholar.

24 Gibbins and Nevitte, “Canadian Political Ideology,” 581–82; Lambert et al., “Left/Right Beliefs,” 556; and Nevitte et al., “The Ideological Contours of ‘New Politics’ in Canada,” 482.

25 Inglehart and Klingemann, “Party Identification,” 265, and Burklin, “The Split,” 286.

26 See, for example, Gramsci, Antonio, Selections from the Prison Notebooks (New York: International Publishers, 1971), 326–327Google Scholar; Parkin, Class Inequality and Political Order, 81–92; and Mann, Michael, “The Social Cohesion of Liberal Democracy,” American Sociological Review 35 (1970), 423–439CrossRefGoogle Scholar.

27 The notion of interpretive packages is taken from Gamson, William A. and Modigliani, Andre, “Media Discourse and Public Opinion on Nuclear Power: A Constructionist Approach,” American Journal of Sociology 95 (1989), 1–9CrossRefGoogle Scholar.

28 Lambert, R. D., Brown, S. D., Curtis, J. E., Kay, B. J. and Wilson, J. M., 1984 Canadian National Election Study Codebook (Waterloo: Department of Sociology, University of Waterloo, 1986)Google Scholar. The 1984 National Election Study was funded by the Social Sciences and Humanities Research Council of Canada. The original collectors of the data and SSHRCC bear no responsibility for the analyses and interpretations presented here.

29 With 41 per cent of the national sample excluded from this study, the possibility of selection bias must be entertained. This issue is explored in the Conclusions.

30 Green, Donald Philip, “On the Dimensionality of Public Sentiment Toward Partisan and Ideological Groups,” American Journal of Political Science 32 (1988), 763–765CrossRefGoogle Scholar.

31 My approach draws upon the earlier work of Erik Olin Wright. See, for example, Wright, Erik Olin, Class Structure and Income Determination (New York: Academic Press, 1979), chap. 2Google Scholar.

32 Covariances between indicators were calculated using pairwise deletion of missing values. Twenty-six per cent of the 1,982 cases were missing on the feelings towards left-wingers item, and 24 per cent were missing on the feelings towards right-wingers item. In contrast, for the 32 political attitude indicators missing cases averaged only 6.8 per cent. These included no opinion, refused and neither agree nor disagree responses.

33 Most of the observed variables included in the analyses are measured on four-category ordinal scales. The use of categorical indicators tends to produce violations of the confirmatory factor assumptions that (1) observed variables have a multinormal distribution, and (2) the population covariance matrix of observed variables is equal to the covariance matrix written as a function of model parameters. Studies indicate, however, that the maximum likelihood fitting function produces parameter estimates which are only slightly biased in the presence of these problems. However, variables with skewnesses and kurtoses greater than 1.0 in absolute value are likely to inflate the chi-square significance test, making models appear to be worse fitting than they actually are. For the 35 variables analyzed in this study, the mean kurtosis equalled −0.5 with 12 variables having a kurtosis greater than 1 in absolute value; and the mean skewness equalled 0.4 with 4 variables having a skewness greater than 1 in absolute value. See Bollen, Structural Equations With Latent Variables, 434–39, and Hayduk, Leslie A., Structural Equation Modeling With LISREL (Baltimore: Johns Hopkins Press, 1987), 328–330Google Scholar.

34 See Jöreskog, Karl G. and Sörbom, Dag, LISREL VI, User's Guide (Mooresville, Ind.: Scientific Software, 1985), 1.24Google Scholar; Bollen, Structural Equations With Latent Variables, 251; and Hayduk, Structural Equation Modeling, 149.

35 The goodness-of-fit measures presented in the Appendix indicate that although the fit of the constrained models is significantly worse than that of the unconstrained models (chi-square difference tests), the difference in fit is substantively unimportant (see the GFI values).

36 Lambert et al., “Left/Right Beliefs,” 557, 561.

37 Peffley and Hurwitz, “A Hierarchical Model,” 883.

38 Laponce, Left and Right, 199–200; Levitin, Teresa E. and Miller, Warren E., “Ideological Interpretations of Presidential Elections,” American Political Science Review 73 (1979), 766CrossRefGoogle Scholar; and Lafferty, William M. and Knutsen, Oddbjørn, “Postmaterialism in a Social Democratic State,” Comparative Political Studies 17 (1985), 417CrossRefGoogle Scholar.

39 Other studies have used the same technique to establish the meaning of the left/right dimension. See Inglehart, Ronald and Sidjanski, Dusan, “The Left, the Right, the Establishment and the Swiss Electorate,” in Budge, et al. , eds., Party Identification, 236Google Scholar, and Johnston, “The Ideological Structure,” 60–64.

40 A series of chi-square difference tests determined whether the estimated covariances for different groups are significantly different. Each test involved constraining two or three phi parameters to equality across groups for a single political attitude. Significant findings are recorded in Table 2. On the logic of chi-square tests of nested models see Hayduk, Structural Equation Modeling, 163–67.

41 As noted by an anonymous reviewer, the relative importance of class and education could best be studied in a nine-group analysis (one for each combination of education stratum and class). Unfortunately, relatively small sample sizes for a number of these subgroups precluded such an approach. However, as a further check on the results reported in the fourth panel of Table 2, education was controlled by studying only those individuals with no post-secondary education in a two-class analysis (managers/supervisors and owners in a pooled group, minimum pairwise N = 160; workers in the second group, minimum pairwise N = 233). The results are consistent with those reported in Table 2, with the covariance between Left/Right Orientation and Economic Justice .03 for workers and .24 for the pooled class group.

42 See, for example, the classic statement by Lukács, Georg, History and Class Consciousness (Cambridge: MIT Press, 1971), 73Google Scholar, and a current version of the Lukácsian position by Ollman, Bertell, “How to Study Class Consciousness, and Why We Should,” Insurgent Sociologist 14 (1987), 57–96CrossRefGoogle Scholar.

43 Compare Gramsci, Selections, 326–37; and Willis, Paul, Learning to Labor: How Working Class Kids Get Working Class Jobs (New York: Columbia University Press, 1981)Google Scholar; with Moorhouse, H. F. and Chamberlain, C. W., “Lower Class Attitudes to Property: Aspects of the Counter-Ideology,” Sociology 8 (1974), 387–405CrossRefGoogle Scholar; and Abercrombie, Nicholas, Hill, Stephen and Turner, Bryan S., The Dominant Ideology Thesis (London: George Allen and Unwin, 1980)Google Scholar.

44 This study also examined the mean difference between these two groups of workers for each of the other 25 political attitude indicators listed in the Appendix. Consistent differences were discovered for only one set of items: for every one of the eight indicators of Alienation from Government, left/right self-identifiers are less alienated (p ≤ .01 in each case). It hardly seems coincidental that workers who lack a left/right identity are much more alienated from government than workers who possess this abstract identity. Further study is needed to see if political sophistication is the common cause of both left/right thinking and a lower level of alienation.