¹ See Marvick's, Dwaine essay, “Political Decision-Makers in Contrasting Milieus,” in Marvick, Dwaine (ed.), Political Decision-Makers: Recruitment and Performance (Glencoe: The Free Press, 1961), pp. 13–27Google Scholar.

² See, for example, Matthews, Donald R., U. S. Senators and Their World (New York: Vintage Books, 1960), pp. 11–67Google Scholar; Mattei Dogan, “Political Ascent in a Class Society: French Deputies 1870–1958,” in Dwaine Marvick (ed.), op. cit., pp. 57–90; and Richards, Peter G., Honourable Members (London: Faber & Faber, 1959)Google Scholar. A recent study of a non-western elite is Frey's, FrederickThe Turkish Political Elite (Cambridge: M.I.T. Press, 1965)Google Scholar. See, also, comparative studies such as Lasswell, Harold D., The Comparative Study of Elites (Stanford: Stanford University Press, 1952)Google Scholar and Matthews, Donald R., The Social Background of Political Decision-Makers (Garden City: Doubleday, 1954)Google Scholar.

³ See, for example, Eulau, Heinz, “The Legislative Career,” in Wahlke, John, Eulau, Heinz, Buchanan, William and Ferguson, Leroy, The Legislative System (New York: John Wiley & Sons, 1962), pp. 69–121Google Scholar; Dwaine Marvick and Charles R. Nixon, “Recruitment Contrasts in Rival Campaign Groups,” in Dwaine Marvick (ed.), op. cit., pp. 193–217; Barber, James D., The Lawmakers: Recruitment and Adaptation to Legislative Life (New Haven: Yale University Press, 1965)Google Scholar; Wright, William E., “Local Party Leadership in West Berlin: SPD and CDU,” (Unpublished doctoral dissertation, Vanderbilt University, 1966)Google Scholar; and Bowman, Lewis and Boynton, G. R., “Recruitment Patterns Among Local Party Officials: A Model and Some Preliminary Findings in Selected Locales,” this Review, 60 (1966), 667–676Google Scholar.

⁴ See, for example, Key, V. O., Southern Politics in Stale and Nation (New York: Vintage Books Ed.), pp. 386–442Google Scholar, and American State Politics (New York: Alfred Knopf, 1956), pp. 133–196Google Scholar; Rossi, Peter H. and Cutright, Phillips, “The Impact of Party Organization in an Industrial Setting,” in Janowitz, Morris (ed.), Community Political Systems (Glencoe: The Free Press, 1961), pp. 81–116Google Scholar; Seligman, Lester, “Political Recruitment and Party Structure,” this Review, 55 (1961), 77–86Google Scholar; Rokkan, Stein and Valen, Henry, “The Mobilization of the Periphery: Data on Turnout, Party Membership and Candidate Recruitment in Norway,” Acya Sociologica, 6 (1962), 115–158Google Scholar; and Snowiss, Leo M., “Congressional Recruitment and Representation,” this Review, 60 (1966), 627–639Google Scholar,

⁵ Eldersveld, Samuel J., Political Parlies: A Behavioral Analysis (Chicago: Rand McNally, 1964)Google Scholar.

⁶ Ranney, Austin, Pathways to Parliament (Madison: University of Wisconsin Press, 1965)Google Scholar.

⁷ Valen, Henry, “The Recruitment of Parliamentary Nominees in Norway,” Scandinavian Political Studies, 1 (1966), 121–166CrossRefGoogle Scholar.

⁸ V. O. Key, for example, in his study of Southern state parties, found that the favored Democratic parties or (in some instances) factions were able to recruit candidates and workers of relatively high status. The less favored Republicans, however, had great difficulty recruiting “respect able” people. Leo M. Snowiss claimed that the characteristics of Congressional candidates in the Chicago area varied with inter-party competition. Thus, in non-competitive districts in which they were certain to win, the Democrats recruited “old-pro” types who were longtime activists in the party organization. In non-competitive districts in which the Republicans were certain to win, however, the Democrats tended to recruit younger and, presumably, better-educated men who most often were lawyers. In marginal districts they frequently recruited candidates with only tenuous links to the party organization rather than the old-line politicians sent to Congress from their inner-city stronghold.

⁹ Eldersveld, and Marvick and Nixon, in their respective studies of party activists in Detroit and Los Angeles, found that the Republican subjects enjoyed more prestigious occupations than their Democratic counterparts. Both Austin Ranney and Henry Valen found that the socio-economic status of candidates for Parliament in Britain and Norway, respectively, varied with the right-left positions of the parties.

¹⁰ Stein Rokkan and Henry Valen demonstrated that the proportion of Norwegian women recruited as party workers increased markedly in urban areas. Valen subsequently found that the proportion of women candidates for the Norwegian Parliament in part was a function of the relative urbanism of constituencies. Leo Snowiss claimed that differences in the backgrounds, the skills, and even the political “styles” of Congressional candidates were associated with relative urbanism (i.e., whether they were seeking office in the core city, the outer city, or the suburban townships of Chicago).

¹¹ (Ottawa: Queen's Printer). This is the most convenient and reliable source for the occupations of both incumbent and non-incumbent candidates. The Parliamentary Guide has considerable bibliographic information but only for incumbents. In seven instances, when a Report listed the occupation of a candidate as “gentleman” or “retired” we also used the Parliamentary Guide, Who's Who in Canada, and the National Reference Book to ascertain occupations. Although there were not enough cases to significantly affect the analysis, it is still possible that some of the published and “real” occupations of candidates differed.

¹² (Ottawa: Dominion Bureau of Statistics). The data for 1946 are available only for Manitoba, Saskatchewan, and Alberta and hence, have not been employed in the analysis.

¹³ Revised Statutes of Canada, 1952, Chapter 23 (2:38)Google Scholar.

¹⁴ See the discussion in Janda, Kenneth, Data Processing (Evanston: Northwestern University Press, 1965), pp. 174–183Google Scholar.

¹⁵ Except for a two-member constituency in Halifax, Nova Scotia and Queens, Prince Edward Island. The two-member districts were abolished and the number of constituencies was increased to 264 after the 1966 redistribution of seats.

¹⁶ Mark Kesselman has employed an “index of multipartyism” to measure competition in French national and local elections. According to Kesselman, the index does not measure competition but rather the convergence of votes among competing parties. See his “French Local Politics: A Statistical Examination of Grass Roots Consensus,” this Review, 60 (1966), 963–973Google Scholar.

¹⁷ We cumulated the absolute value of all possible differences among parties and divided this sum by the total number of differences. Since the resultant measure increases with disparity among candidates, we have signed it negatively in order that a larger number represent greater competition. Figure 1 is a schematic representation of a situation in which all four parties contest a constituency. The resultant measure is Gini's mean difference. For bibliography concerning this measure see Glasser, Gerald J., “Variance Formulas for the Mean Difference and Coefficient of Concentration,” Journal of the American Statistical Association, 59 (1962), 648–654CrossRefGoogle Scholar.

¹⁸ The rationale and method for computing these scores are described in Reiss, Albert J.et al., Occupation and Social Change (Glencoe: The Free Press, 1961)Google Scholar.

¹⁹ We did not employ the Canadian occupational class scale developed by Bernard Blishen because, in our view, the Duncan scale was more detailed. However, Blishen in describing his scale said that rank correlations were computed between the ratings of occupational prestige in each of several countries previously studied by Alex Inkeles and Peter Rossi (i.e., U.S.A., Great Britain, New Zealand, Japan, the U.S.S.R., and West Germany). According to Blishen, “The highest ranked correlation, .94, was found between Canada and the U.S.” See Blishen, Bernard, “The Construction and Use of an Occupational Class Scale,” Canadian Journal of Economics and Political Science, 24 (1958), 523CrossRefGoogle Scholar. It should be noted that Inkeles and Rossi had found that in any population there was very little variation in the evaluations of occupations by sub-groups. Thus, Inkeles later observed, “Whether a worker or a professor does the rating, both place the doctor, lawyer, and engineer very near the top of the list, the ordinary worker about two-thirds of the way down, the shoe-shine boy or garbage man at the bottom.” See Inkeles, Alex, “Industrial Man: The Relation of Status to Experience, Perception and Value,” American Sociological Review, 66 (1960), 8Google Scholar. The earlier study referred to by Blishen, is Inkeles, Alex and Rossi, Peter, “National Comparisons of Occupational Prestige,” American Journal of Sociology, 61 (1956), 329–339Google Scholar.

²⁰ Like the measure of inter-party competition, status difference is the sum of all absolute value differences in SES scores for the candidates in every constituency over the total number of possible differences. Since we shall call this measure “status difference” it has been signed positively. Figure 2 represents a status difference score for four candidates.

²¹ Samuel J. Eldersveld's finding that a considerable proportion of the party activists he studied claimed they generated their own political careers suggests that the relationship between recruiter and recruitee is not a neat asymmetrical one. See Eldersveld, op. cit., p. 128. However, the assumption that candidate recruitment, even in a parliamentary system with disciplined parties, is left largely to the local party organizations is based upon a study of legislative behavior in a Canadian House of Commons. See Kornberg, Allan, Canadian Legislative Behavior: A Study of the 25th Parliament (New York: Holt, Rinehart and Winston, 1967), pp. 63–66Google Scholar. See, also, Dawson, R.M. and Ward, Norman, The Government of Canada, 4th ed., (Toronto: University of Toronto Press, 1963), pp. 480–486Google Scholar. Ranney's investigations in Great Britain and Valen's research in Norway also point to the important role local party organizations play in the process of candidate recruitment.

²² Both our willingness to accept this ideological continuum and the less-than-complete consensus among observers can be understood by considering the observations of various experts. Some of these follow:

Parliamentary scholar R.M. Dawson has written that historically there have always been differences between Conservatives and Liberals (e.g., “The Conservatives have been the party of free enterprise while the Liberals have taken the lead in public ownership and social legislation”). See Dawson, R.M., The Government of Canada (Toronto: University of Toronto Press, 1948), p. 506Google Scholar. Data from an empirical study of the 25th Parliament tend to support this point of view. See Kornberg, Allan, “Caucus and Cohesion in Canadian Parliamentary Parties,” this Review, 60 (1966), 83–92Google Scholar. Also, Robert Alford has described the New Democratic party as a “Social-Democratic-Left” party; the Liberals as a “Center-Left” party; the Conservatives as a “Conservative-Right” party; and the Social Credit party as a “Reactionary-Right” party. See Alford, Robert, Party and Society (Chicago: Rand McNally, 1963), p. 13Google Scholar. Somewhat differently, Gad Horowitz sees the Conservatives as the party of the right, the NDP as the left, and the Liberals as the “classless” center party. Further, the electoral success of the Liberals, their “refusal to appear as a class party forces both right and left to mitigate their class appeals and to become themselves, in a sense, centre parties.” See Horowitz, Gad, “Conservatism, Liberalism and Socialism in Canada: An Interpretation,” in Thorburn, Hugh G. (ed.), Party Politics in Canada, 2nd ed., (Scarborough: Prentice-Hall of Canada, 1967), pp. 55–74Google Scholar.

On the other hand (and despite the fact that a series of Liberal Administrations have passed most of the social welfare legislation Canadians enjoy) sociologist John Porter has argued there are few, if any, basic differences between the Liberals and Conservatives: both espouse the same conservative values; both parties are oriented toward business; and both perform essentially brokerage functions. See Porter, John, The Vertical Mosaic: An Analysis of Social Class and Power in Canada (Toronto: University of Toronto Press, 1965), pp. 373–379CrossRefGoogle Scholar.

As for the two minor parties, the Social Credit and the CCF, both grew out of the short-lived Progressive party. See Morton, William L., The Progressive Party in Canada (Toronto: University of Toronto Press, 1950)Google Scholar. The Social Credit party began as a radical, rural, fundamentalist movement that was pledged to implement extremely unorthodox financial policies. See Irving, John, The Social Credit Movement in Alberta (Toronto: University of Toronto Press, 1959)Google Scholar. Over time, electoral success at the provincial level and the quest for respectability seemingly have caused the Western wing of the party to abandon the financial panaceas and radical ideology advocated during the 30's. The Alberta and British Columbia contingent appears to have moved from a populist, almost neo-fascist party, to a solidly conservative one. In fact, both former national leader Robert Thompson and Mr. Ernest Manning, long-time Premier of Alberta, a number of times have called upon Conservative party supporters to join with their party in a truly conservative coalition against socialist (i.e., Liberal and NDP) forces in Canada. See Mr.Manning's, recent book Political Realignment: A Challenge to Thoughtful Canadians (Toronto: McClelland and Stewart Limited, 1967)Google Scholar. Mr. Thompson, perhaps because few Conservatives were joining him in the Social Credit party, recently left Social Credit, joined the Conservatives, and won the Conservative nomination for Red Deer, Alberta, in the June, 1968, national election. Spectacular defections such as those by Messrs. Thompson and Olson have seriously undermined Social Credit's claim to be a national party. Certainly, the nomination of candidates in less than a hundred constituencies in the 1968 election suggests that such defections have a deletorious effect on candidate recruitment.

The initial intention of the CCF was to eradicate capitalism in Canada and to replace it with a socialist society. In 1961, the party's close ties with organized labor were formalized in a union from which emerged the New Democratic Party. The CCF-NDP also appears to have lost some of its ideological militancy. The desire for electoral success and, relatedly, the necessity of broadening the base of its middle class support, seemingly, have turned the CCF-NDP from a party advocating a full-blown socialist ideology to a leftist party of moderate reform. See Zakuta, Leo, A Protest Movement Becalmed: A Study of Change in the CCF (Toronto: University of Toronto Press, 1964)Google Scholar.

A kind of ideolological convergence seemingly has taken place on the Canadian left. The Liberals, supposedly, under pressure from the CCF-NDP, have moved left while the CCF-NDP, in turn, appears to have moved somewhat to the right. That the Liberal party's drift to the left has not been unopposed, however, is suggested by the recent struggle in the Liberal Cabinet and caucus between the supporters of Messrs. Sharp (the right) and Gordon (the left). In our view, however, the NDP still must be regarded as the party of the left, while the Liberals may be best described as a left-center party.

On the other half of the continuum at least the Western wing of Social Credit has moved from a radical conservative to a traditional conservative position. Still, they are certainly to the right of the Conservatives since the latter party moved somewhat to the left (e.g., they passed legislation which increased farm subsidies and extended benefits paid to old-age pensioners) during the tenure of their recently deposed leader, Mr. John Diefenbaker. Thus, the Social Credit still can be regarded as the Canadian party of the right while the Conservatives can be best described as a center-right party.

²³ However, the status of Canadian candidates still did not vary with their right-left ideological positions since the losing right-wing Social Creditors had mean status scores which were considerably lower than both the unsuccessful Conservatives and Liberals.

²⁴ For the period 1945–65, 28.5% of the candidates of the two major Canadian parties were lawyers. In comparison, Ranney found that 15% of the candidates of the Labor and Conservative parties were lawyers (i.e., barristers and solicitors). Ranney's data support the findings of an earlier study of incumbent British MP's by Barrington and Finer. See Barrington, H.E. and Finer, S. E., “The British House of Commons,” in Meynaud, Jean (ed.)Google Scholar, “The Parliamentary Profession,” International Social Science Journal, 13 (1961), 601–605Google Scholar.

²⁵ See Kornberg, op. cit., p. 44. Dawson and Ward have estimated that approximately a third of the members of the several Canadian Parliaments have been lawyers. See Dawson, R. M. and Ward, N., The Government of Canada, 4th ed. (Toronto: University of Toronto Press, 1964), p. 346Google Scholar. Their estimate is supported by data generated by Roman R. March in his longitudinal study of the characteristics of Canadian Parliamentarians. See March, Roman R., “An Empirical Test of M. Ostrogorski's Theory of Political Evolution in a British Parliamentary System,” (Unpublished doctoral dissertation, Indiana University, 1967)Google Scholar.

²⁶ Brady, Alexander, “Canada and the Model of Westminister,” in Hamilton, William B. (ed.), The Transfer of Institutions (Durham, N.C.: Duke University Press, 1964), pp. 59–80Google Scholar. Brady argues, “In the absence of anything like the modern British governing class, the legal profession in both the French and English sections of the country is a partial substitute. From the birth of the federation the role of lawyers in the national Parliament has been conspicuously more important than at Westminister.” (p. 74)

However, in his intensive analysis of the distribution of power in Canadian society, John Porter claims there is an effective Canadian ruling class—the members of the Canadian corporate elite. They have not (generally) bothered to make their way in national politics but have preferred to govern through the political elite. Most frequently the lawyers in the corporate elite have provided the link between the corporate and political worlds. They are the most likely to have overt political affiliations, to hold elected or appointed public office, and to hold top offices in the national and provincial party organizations. See Porter, op. cit., pp. 296–298.

²⁷ Although, as has been indicated, Valen found that the occupational status of candidates for Parliament varied with the right-left positions of Norwegian parties, he also found by far the largest proportion of low status blue collar workers among Communist candidates.

²⁸ It could be argued that a comparison between the French “far right” and the Social Credit party is not a particularly appropriate one. It should be noted, however, that the Poujadist element of the French far right and the Quebec Creditistes were both “protest” movements and drew support from relatively similar social groups within their respective electorates.

²⁹ See Porter, op. cit., pp. 91–96. See, also, Lovink, J.A.A., “The Politics of Quebec: Provincial Political Parties, 1897–1936,” (Unpublished doctoral dissertation, Duke University, 1967)Google Scholar. The Creditiste successes in Quebec, particularly in 1962, have been interpreted as a protest against the traditional elite domination of both the Liberal and Conservative parties. See, for example, Guindon, Hubert, “Social Unrest, Social Class and Quebec's Bureaucratic Revolution,” Queen's Quarterly, 71 (1964), 150–162Google Scholar.

³⁰ See Robert R. Alford, op. cit., pp. 250–286.

³¹ A 1965 survey of four metropolitan areas showed that people voting Conservative tend to follow religious lines, whereas NDP voters follow class lines. Liberal voters appear to be crosspressured, with some following class and others religious lines. Further, religious affiliation would seem to be a more important determinant of voting preference among working class than among middle class voters. See Gagne, Wallace and Regenstreif, Peter, “Some Aspects of New Democratic Party Urban Support in 1965,” Canadian Journal of Economics and Political Science, 23 (1967), 529–550CrossRefGoogle Scholar.

³² In fact, the CCF-NDP entered one candidate less in 1953 than it had in 1949, while the Social Credit party contested six fewer sample constituencies in 1953.

³³ In 1953, the Liberals increased the number of their 1949 candidates by 11. The Conservatives increased their 1949 total of candidates by 10 in 1953.

³⁴ That this strategy is likely to be employed by parties is suggested by Ranney's study of the recruitment of British MP's. Ranney arrayed British constituencies along a “Winability” continuum and found that previous electoral experience was significantly related to the desirability of the constituencies in which non-incumbents stood. Austin Ranney, op. cit., p. 94. Further, Leo Snowiss reports that in the suburban townships in which there was no prospect of winning, the Chicago Democratic organization recruited Congressional candidates from among ambitious young lawyers seeking to establish themselves politically and professionally. Here they might be groomed for future more important (i.e., local) public office. See Snowiss, op. cit., pp. 633–635.

³⁵ Canadian party organizations have considerable patronage at their disposal if they hold ex-ecutive office at either the national or provincial levels. Lawyers, in particular, are likely to benefit by being given substantial legal work by national and/or provincial Administrations. In Chicago, Snowiss reports that the considerable patronage available to the Democratic machine undoubtedly enabled them to recruit candidates and workers in areas in which the party was weak. See Snowiss, ibid., pp. 627–639.

³⁶ For an explication of their technique see Simon, Herbert A., Models of Man (New York: John Wiley and Sons, Inc., 1957)Google Scholar, chapters I-III; and Blalock, Hubert M. Jr., Causal Influences in Non-experimental Research (Chapel Hill: University of North Carolina Press, 1964)Google Scholar, chapters I-III.

³⁷ Social scientists were introduced to path analysis by geneticist Sewell Wright in a series of articles dating from the 1920's. See, for example, Wright, Sewell, “Correlation and Causation,” Journal of Agricultural Research, 20 (1921), 557–585Google Scholar; “The Method of Path Coefficients,” Annals of Mathematical Statistics, 5 (1934), 161–215CrossRefGoogle Scholar; and “Path Coefficients and Path Regressions: Alternative or Complementary Concepts?” Biometrics, 16 (1960), 189–202CrossRefGoogle Scholar.