Hostname: page-component-cd9895bd7-dk4vv Total loading time: 0 Render date: 2024-12-28T06:31:48.082Z Has data issue: false hasContentIssue false
  • English
  • Français

Class and Region in Canadian Voting: A Dependency Interpretation*

Published online by Cambridge University Press:  10 November 2009

Elisabeth Gidengil
Elisabeth Gidengil
Affiliation:
McGill University
Get access Rights & Permissions [Opens in a new window]

Abstract

This study develops a dependency interpretation of the interplay between class and region in influencing Canadian voting. The weakness of any national class cleavage in voting is linked to the socially disintegrative effects of regional dependency. Class cleavages are not consistently manifested in Canadian voting because consistent class interests are lacking. Log-linear analyses confirm that class does affect voting but this effect differs in both form and intensity depending on a region's location in the centre-periphery system. The impact of union membership and language on the interplay between class and region is also examined.

Résumé

Cet article présente une interprétation, basée sur la théorie de la dépendance, de l'interaction entre classe et région en tant que facteurs déterminants du vote au Canada. La faiblesse du clivage de classe dans le vote national est attribuée aux effets perturbateurs de la dépendance régionale. Les clivages de classe ne se manifestent pas de façon cohérente dans le vote canadien parce que la cohérence des intérêts de classes est absente. Une analyse des modèles log-linéaires confirme que l'appartenance de classe a bel et bien un effet sur le vote mais que cet effet est différent en nature et en intensité selon la situation de la région dans le système centre-périphérie. L'auteur analyse également l'impact de l'appartenance syndicale et celui de la langue sur l'interaction classe/région.

Type
Research Article
Information
Canadian Journal of Political Science/Revue canadienne de science politique , Volume 22 , Issue 3 , September 1989 , pp. 563 - 587
Copyright
Copyright © Canadian Political Science Association (l'Association canadienne de science politique) and/et la Société québécoise de science politique 1989

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 Lambert, Ronald D., Curtis, James E., Brown, Stephen D. and Kay, Barry J., “Social Class and Voting,” in Curtis, James, Grabb, Edward, Guppy, Neil and Gilbert, Sid (eds.), Social Inequality in Canada (Scarborough: Prentice-Hall, 1988), 456.Google Scholar

2 Alford, Robert R., Party and Society: The Anglo-American Democracies (Chicago: Rand McNally, 1963).Google Scholar

3 Johnston, J. Paul, “Some Methodological Issues in the Study of ‘Class Voting’: A Critique of Erickson's Analysis,” Canadian Journal of Sociology 6 (1981), 145.CrossRefGoogle Scholar

4 Ornstein, Michael D., Stevenson, H. Michael and Williams, A. Paul, “Region, Class and Political Culture in Canada,” this Journal 13 (1980), 227.Google Scholar

5 Their concern with the relative magnitudes of regional versus status and/or class effects, for example, leads Ornstein and his colleagues to overlook the fact that neither region nor class/status is actually a very good predictor of the political attitudes examined. See also Matthews, Ralph and Davis, J. Campbell, “The Comparative Influence of Region, Status, Class and Ethnicity on Canadian Attitudes and Values,” in Brym, Robert J. (ed.), Regionalism in Canada (Toronto: Irwin, 1986), 97.Google Scholar

6 Simeon, Richard and Elkins, David J., “Regional Political Cultures in Canada,” this Journal 7 (1974), 397.Google Scholar “Regionalism” is used here to refer simply to “the political salience of the spatial dimension.” See Elkins, David and Simeon, Richard, Small Worlds (Toronto: Methuen, 1980), xi.Google Scholar

7 Chi, N. H., “The Regression Model of Regionalism: A Critique,” this Journal 5 (1972), 91Google Scholar (emphasis in the original).

8 Simeon and Elkins, “Regional Political Cultures,” 399.

9 For an extended and thoughtful discussion of the issues involved in defining and identifying regional differences, see Matthews, Ralph, The Creation of Regional Dependency (Toronto: University of Toronto Press, 1983)Google Scholar, chap. 1.

10 Elkins and Simeon, Small Worlds, xi.

11 Schwartz, Mildred A., Politics and Territory (Montreal: McGill-Queen's University Press, 1974), 5CrossRefGoogle Scholar (emphasis added).

12 Lipset, Seymour Martin, Political Man (Garden City: Doubleday, 1960).Google Scholar

13 Lipset, Seymour Martin and Rokkan, Stein, “Cleavage Structures, Party Systems, and Voter Alignments: An Introduction,” in Lipset, Seymour M. and Rokkan, Stein (eds.). Party Systems and Voter Alignments (New York: Free Press, 1967), 13.Google Scholar

14 Ibid., 10.

15 Lipset and Rokkan did not, in fact, view this evolutionary trajectory as unproblematic: archaic forms of conflict could become institutionalized in the party system, enduring long after the original conflicts had lost their salience or, if unresolved, the conflicts of nation-building could continue into later evolutionary stages.

16 Wilson, John, “The Canadian Political Cultures: Towards a Redefinition of the Nature of the Canadian Political System,” this Journal 7 (1974), 438–83.Google Scholar

17 In addition to Wilson, “The Canadian Political Cultures,” see for example Jenson, Jane, “Party Systems,” in Bellamy, David J., Pammett, Jon H. and Rowat, Donald C. (eds.), The Provincial Political Systems (Toronto: Methuen, 1976), 118–31.Google Scholar

18 Blake, Donald E., “The Measurement of Regionalism in Canadian Voting Patterns,” this Journal 5 (1972), 54–81.Google Scholar

19 Cairns, Alan C., “The Electoral System and the Party System in Canada, 1921–1965,” this Journal 1 (1968), 74.Google Scholar

20 Brym, Regionalism in Canada, vii–viii.

21 Blalock, Hubert M. Jr., Theory Construction (Englewood Cliffs, New Jersey: Prentice-Hall, 1969), 35.Google Scholar Blake's use of ecological regression to study regionalism in Canadian voting patterns is an important exception to this tendency to neglect possible interactions between class and region. See Blake, “The Measurement of Regionalism in Canadian Voting Patterns.”

22 Economic Council of Canada, Living Together (Ottawa: Supply and Services Canada, 1977), 25.Google Scholar

23 Panitch, Leo, “Dependency and Class in Canadian Political Economy,” Studies in Political Economy 6 (1981), 7–36.CrossRefGoogle Scholar

24 Ibid., 28.

25 Santos, Theotonio Dos, “The Structure of Dependence,” American Economic Review 60 (1970), 231.Google Scholar

26 Cardoso, Fernando, “Associated-Dependent Development: Theoretical and Practical Implications,” in Stepan, Alfred (ed.), Authoritarian Brazil (New Haven: Yale University Press, 1973), 163.Google Scholar

27 Booth, David, “André Gunder Frank: An Introduction and Appreciation,” in Oxaal, Ivar, Barnett, Tony and Booth, David (eds.), Beyond the Sociology of Development (London: Routledge and Kegan Paul, 1975), 50–85.Google Scholar See also Friedmann, Harriet and Wayne, Jack, “Dependency Theory: A Critique,” Canadian Journal of Sociology 2 (1977), 399–416.CrossRefGoogle Scholar

28 Clement, Wallace, “A Political Economy of Regionalism,” in Glenday, Daniel, Guindon, Herbert and Turowetz, Allan (eds.), Modernization and the Canadian State (Toronto: Macmillan, 1978), 89–110.Google Scholar

29 In addition to Dos Santos, see especially Cardoso, Fernando and Faletto, Enzo, Dependency and Development in Latin America (Berkeley: University of California Press, 1979)Google Scholar; and Sunkel, Osvaldo, “Transnational Capitalism and National Disintegration in Latin America,” Social and Economic Studies 22 (1973), 132–76.Google Scholar

30 Galtung, Johan, “A Structural Theory of Imperialism,” Journal of Peace Research 8 (1971), 81.CrossRefGoogle Scholar While the discussion here is in terms of a simple dichotomy of centre and periphery, this is an analytical simplification adopted for ease of exposition.

31 Santos, Theotonio Dos, “The Crisis of Development Theory and the Problem of Dependence in Latin America,” in Bernstein, Henry (ed.), Underdevelopment and Development (Harmondsworth: Penguin, 1973), 78.Google Scholar

32 Marchak, Patricia, “Class, Regional and Institutional Sources of Social Conflict in B.C.,” BC Studies 27 (1975), 46.Google Scholar This study does recognize the possible interaction between class and region and actually links it to a centre-periphery perspective. The latter is not developed, however, except by reference to the work of Frank, and Marchak too readily assumes peripheral solidarity.

33 Sunkel, “Transnational Capitalism and National Disintegration in Latin America,” 170.

34 Galtung, “A Structural Theory of Imperialism,” 84.

35 Schwartz, Mildred, “Canadian Voting Behavior,” in Rose, Richard (ed.), Electoral Behavior (New York: Free Press, 1974), 592.Google Scholar

36 See Caporaso, James, “Introduction to the Special Issue of International Organization on Dependence and Dependency in the Global System,” International Organization 32 (1978), 1–12.CrossRefGoogle Scholar

37 For an extended discussion of this conceptualization of dependency, as well as the indicators that were used to operationalize the component characteristics of the dependency syndrome, see Gidengil, Elisabeth, “Diversity within Unity: On Analyzing Regional Dependency,” Studies in Political Economy 29 (1989), 91–122.CrossRefGoogle Scholar

38 This time period was chosen to allow the use of the 1965, 1968 and 1974 National Election Studies upon which a number of the earlier analyses of class voting were based.

39 To ensure that the results were not simply an artifact of the particular clustering algorithm chosen, a cross-validation was performed. The interpretation of the cluster solution was externally validated by comparing the six clusters on variables that were not used in the clustering, but which on theoretical grounds should be related to them (age structure and growth in the labour force).

40 The strong East-West division evident between the depressed and industrial peripheries, on the one hand, and the vulnerable and advantaged peripheries, on the other, helped to limit the possible confounding effects on the subsequent analysis of differences in the context of electoral politics between the Atlantic provinces and the West.

41 As argued, for example, in Schwartz, Politics and Territory, 288.

42 See especially Archer, Keith, “The Failure of the New Democratic Party: Unions, Unionists and Politics in Canada,” this Journal 18 (1985), 353–66.Google Scholar

43 Simeon and Elkins used a merged data set created from the same three surveys in the Small Worlds version of their study of provincial political cultures. One drawback of this procedure is the possible introduction of temporal variation to the extent that transitory events affected particular election outcomes. The National Election Studies used in this analysis were made available by the Inter-University Consortium for Political and Social Research. Neither the Consortium nor the original investigators bears responsibility for the analysis or interpretation presented here.

44 Lambert et al., “Social Class and Voting,” 460..

45 Schreiber, E. M., “Class Awareness and Class Voting in Canada: A Reconsideration of the Ogmundson Thesis,” Canadian Review of Sociology and Anthropology 17 (1980), 41.Google Scholar

46 Kahan, Michael, Butler, David and Stokes, Donald, “On the Analytical Division of Social Class,” British Journal of Sociology 17 (1966), 128.CrossRefGoogle Scholar Of the three objective measures examined, only occupational status suggested a natural class dichotomy when related to subjective class identification.

47 Schwartz, “Canadian Voting Behavior,” 585–89.

48 Gagne, Wallace and Regenstreif, Peter, “Some Aspects of New Democratic Party Urban Support in 1965,” Canadian Journal of Economics and Political Science 33 (1967), 549.CrossRefGoogle Scholar

49 Fingleton, B., Models of Category Counts (Cambridge: Cambridge University Press, 1984), 4.CrossRefGoogle Scholar

50 A log-linear model can be converted into additive form by taking natural logarithms of all the terms. The term “log-linear” derives from the fact that the equations are linear in their logarithms. While the additive form is mathematically equivalent, the multiplicative form is presented here because it has the conceptual advantage of “closeness to the odds and odds ratio basis underlying the motivation of the models” ( Knoke, David and Burke, Peter J., Log-Linear Models [Beverly Hills: Sage, 1980], 19CrossRefGoogle Scholar).

51 Swafford, Michael, “Three Parametric Techniques for Contingency Table Analysis: A Nontechnical Commentary,” American Sociological Review 45 (1980), 676.CrossRefGoogle Scholar The automatic inclusion of these interactions is not without objections, including the loss of degrees of freedom when insignificant interactions are included.

52 See Fingleton, Models of Category Counts, 71. Just how much these are facts of Canadian life and not just artifacts of the sampling schemes used is demonstrated by the fact that an adequate fit actually could not be achieved without these terms.

53 A similarly conservative approach was applied to the problem of “sampling zeros” which can occur due to the small probabilities attached to some categories. Where sampling zeros were encountered here, the widely recommended solution was adopted of adding 0.5 to every observed cell frequency. On the problem of sampling zeros, see Knoke and Burke, Log-Linear Models, 63–65.

54 See Johnston, “Some Methodological Issues in the Study of ‘Class Voting,’” 151–52.

55 Fingleton, Models of Category Counts, 65.

56 Since the standardized residuals approximate standard normal deviates, any standardized residual exceeding ± 1.96 would be considered significant at the .05 level.

57 A test that all effects greater than a given order are equal to zero is performed by assessing the fit of a model containing all the effects of that order and their implied subordinate interactions. Since the models of full order are nested, the significance of the differences between these models can also be tested.

58 Numbers permitted the inclusion of Social Credit/Creditiste voters in this analysis. Whether these voters are included or not [SRLU][VSRL] still emerges as the “best” model.