¹ “The Electoral System and the Party System in Canada, 1921–1965,” this Journal, I, no. 1 (March 1968), 55–80.

² From 1921 to 1965, the period of Cairns’ study, there were two double-member constituencies in Canada, Queen's and Halifax. For the elections of 1921 to 1930, there was a third, Ottawa. At present all constituencies are single-member.

³ Ibid., 64.

⁴ Ibid., 64–8.

⁵ Ibid.,63.

⁶ Ibid., 68–71.

⁷ Ibid., 71–2.

⁸ Ibid., 76.

⁹ Ibid., 74–5.

¹⁰ Ibid., 56–7. The data have been reworked to include the results of the 1968 federal general election.

¹¹ Ibid., 77–8.

¹² The relationship between the parliamentary and extra-parliamentary parties while the ccf was in power in Saskatchewan suggests that the difference in this respect between the ndp and the other parties may well diminish de facto if that party should win office federally. See Eager, Evelyn, “The Paradox of Power in the Saskatchewan c.c.f., 1944–1961,” in Aitchison, J. H., ed., The Political Process in Canada (Toronto, 1963), 118–35.Google Scholar

¹³ Since the electoral system is only partly responsible for the regional composition of the parliamentary party, a relationship between the latter variable and regional policy bias short of rough congruence would make it difficult to maintain Cairns's contention that the electoral system exercises an “important” influence on party policy.

This point is reinforced when one considers that part of the distortion in the votes-seats relationship may not be attributable to the electoral system at all, but to deviations from the one-man-one-vote principle of apportionment that are remediable without changing the electoral system. How much of the distortion is traceable to this last factor does not seem measurable.

¹⁴ Cairns, “The Electoral System,” 69.

¹⁵ For an attempt at measuring the role perceptions of Canadian mps in 1962, see Allan, Kornberg, Canadian Legislative Behavior (New York, 1967), chap.6.Google Scholar

¹⁶ These references to alternative strategies should not be taken as suggesting that party policy is invariably the result of careful calculation. They are intended only as a convenient means of distinguishing between different types of policy-sets, however arrived at.

¹⁷ Cairns, “The Electoral System,” 64–8. John Meisel's data for the 1957 general election did not indicate a close relationship between the party leaders’ campaign itineraries and the size of the margin separating the parties in previous elections. The Canadian General Election of 1957 (Toronto, 1962), 153–5. The fit might be closer, however, if one were to compare leaders’ itineraries with contemporary evaluations of the margins separating the parties from seat to seat.

¹⁸ Cairns, “The Electoral System,” 67. A model of party behaviour along these lines is presented in Downs’, Anthony well-known An Economic Theory of Democracy (New York, 1957).Google Scholar

¹⁹ The reader's attention may have been caught by the small differences in Table I between column 1 and column 2 for the Liberal party, again suggesting that in this instance the potential significance for policy of the electoral system as such, even assuming pursuit of a “patronage” stategy, cannot be rated very high. Comparable data prepared by Cairns for Ontario and Quebec for 1921 to 1965 show, however, that such small differences have not been typical. “The Electoral System,” Table V, 61.

Column 4, which combines columns 2 and 3, has been included in Table I because intuition suggests that this last distribution may be a more likely correlate for party policy than either of the other two. All other things being equal, and to the extent that electoral considerations influence policy at all, it seems as unlikely that the policy claims of “safe” and “marginal” areas should carry equal weight as that the policy claims of “safe” areas should be ignored altogether. On the one hand, threats of loss of electoral support if policy demands are rejected are bound to be taken more seriously if they come from “marginal” areas. On the other hand, intra-party pressures and the need to maintain party unity make it inconceivable that the demands of “safe” areas should be continually rejected.

²⁰ The writer has in progress an historical study of inter-party electoral competition in Canada. An increase in inter-party competition over time (that is, a rise in the proportion of seats classifiable as “marginal”) would probably mean that the differences between columns 2 and 3 were greater in the past than now. A decline in inter-party competition would probably mean the opposite.

²¹ “The Electoral System,” 71 (italics added). The next sentence, after the section quoted above, reads: “From our perspective the basic point is that both [of these strategies] reflect the politics of sectionalism as stimulated by the single-member constituency system.” This may be true (although it must be conceded that only one or neither of these strategies may have been adopted for the reasons stated). None the less, it may be observed that to the extent that these strategies are combined over time their associated regional policy biases will tend to cancel out. As was argued earlier (n. 19 above), it is highly unlikely that either strategy would be pursued to the total exclusion of the other.

²² These descriptions of party policy are highly impressionistic, being subject to all the earlier-mentioned difficulties of measuring regional policy bias, but are based on the following considerations. As far as the needs of the economically depressed areas of the country are concerned, these were acknowledged very soon after the 1968 general election by creation of a Department of Regional Economic Expansion, by assigning responsibility for its affairs to one of the (reputedly) most influential members of the cabinet, and by passage of the Regional Development Incentives Act. By comparison, the government's attitude towards demands for help from the large urban areas seemed (at least up to the end of 1969) less sympathetic, as was also alleged publicly by the Honourable Paul Hellyer upon his resignation from the cabinet. Globe and Mail, Toronto, April 25, 1969, especially p. 8. For the reactions of the Prime Minister and the leader of the opposition to Mr Hellyer's views, see Can. H. of C. Debates, April 24, 1969, pp. 7893–5, and April 25, 1969, pp. 7976–84. Some of Mr Hellyer's complaints have been echoed by Liberal backbenchers from urban constituencies like Barnett Danson (York North) and Philip Givens (York West). Ibid., Oct. 27, 1969, pp. 108–9; Dec. 8, 1969, pp. 1715–18; Globe and Mail, Nov. 22, 1969, and Feb. 12, 1970. Conservative policy, meanwhile, was to demand more active federal involvement in the cities and the creation of a Commons Committee on Urban Affairs. Debates, Dec. 8, 1969, pp. 1700–36. The Conservative party's contemporary desire to please and accommodate Quebec seemed manifest in the leader's apparent support for the “two founding peoples” concept in his 1968 campaign and, for instance, in the support which he and most of his followers subsequently gave to the Official Languages Act. Ibid., May 16, 1969, pp. 8790–4; May 27, 1969, pp. 9123–4.

²³ Qualifiers like “major,” “primary,” and “important” are notorious obstacles to clarity of description. To reduce this hazard, the word “major” as used above should be interpreted to mean “from one-third to one-half” (of the responsibility for regional policy bias), and thus as weaker than words like “primary” and “principal.” In Cairns' analysis, the word most often used to classify the effects of the electoral system is “important.” Given the thrust of the context in which it appears, I consider it justified to interpret his term as equivalent to “major.”

²⁴ That, at any rate, appears to be the meaning of the assertion already quoted above, p. 498.

²⁵ Cairns, “The Electoral System,” 74–5.

²⁶ This very question is begged by Maurice Duverger's statement that “Divisions of public opinion arise not only from natural differences between citizens but also from external factors, of which the most potent is the electoral system. To that extent political divisions represent a pattern imposed on public opinion from without rather than a reflection of differences existing within public opinion.” Political Parties, trans. Barbara and North, Robert (London and New York, 1954), 386Google Scholar (emphasis added).

²⁷ There are, of course, many kinds of proportional representation, all of which distort the votes-seats relationship to some extent, although less than do plurality-vote, single-member constituency systems. Rae, Douglas, The Political Consequences of Electoral Laws (New Haven and London, 1967), especially chaps. 2, 4, and 5Google Scholar. In Canada, it is unlikely that spatial considerations would ever be totally abandoned (as in “pure” or nation-wide proportional representation). But even a moderate form of proportional representation (which might, for instance, divide the country into ten-member constituencies not crossing regional boundaries) would bring votes and seats into quite close correspondence.

²⁸ Regenstreif, S. Peter, “Some Aspects of National Party Support in Canada,” Canadian Journal of Economics and Political Science, XXIX, no. 1 (Feb. 1963), 59–74CrossRefGoogle Scholar; Alford, Robert R., Party and Society (Chicago, 1963), especially chap. 9Google Scholar; Meisel, John, ed., Papers on the 1962 Election (Toronto, 1964)Google Scholar; Courtney, John C. and Smith, David E., “Voting in a Provincial General Election and a Federal By-election: A Constituency Study of Saskatoon City,” Canadian Journal of Economics and Political Science, XXXII, no. 3 (Aug. 1966), 338–53CrossRefGoogle Scholar; Simmons, James W., “Voting Behaviour and Socio-economic Characteristics: The Middlesex East Federal Election, 1965,” ibid., XXXIII, no. 3 (Aug. 1967), 389–400Google Scholar; Wilson, John, “Politics and Social Class in Canada: The Case of Waterloo South,” this Journal, I, no. 3 (Sept. 1968), 288–309.Google Scholar

Professor Cairns’ view of this question is ambiguous. In one instance he asserts that “the emphasis on sectional divisions engendered by the electoral system has submerged class conflicts…” In another he says that “given the historical (and existing) state of class polarization in Canada the electoral system has made sectionalism a more rewarding vehicle for amassing political support than class.” In a third instance he maintains that the single-member constituency system discourages class-based politics “when class identities are weak or submerged behind sectional identities.” Further on there is a reference to “the strong tendency to sectionalism found in the very nature of Canadian society…” “The Electoral System,” 74–5.

²⁹ For an interesting discussion of the development process from predominantly cross-class, territorial conflict patterns to predominantly cross-constituency, class conflict patterns, see Rokkan, Stein, “Electoral Mobilization, Party Competition and National Integration,” in LaPalombara, Joseph and Weiner, Myron, eds., Political Parties and Political Development (Princeton, 1966), 241–65Google Scholar. He points out that the pattern of change has varied from country to country, and has been slow where there has been a strong local consensus, a tradition of rural territorial representation, and no deep urban-rural cleavage. Ibid., 256–65.

³⁰ The implicit assumption that the influence of the electoral system on party policy is equally strong in the two cases considered may be wrong, and the curves may in fact differ in shape. This contingency will not affect the argument, however, so long as the curves continue to converge sharply towards each end of the cleavage spectrum.

³¹ See n. 10 above.

³² See, for instance, Schattschneider, E. E., Party Government (New York, 1942), 74–5Google Scholar; Smiley, Donald V., “The Two-Party System and One-Party Dominance in the Liberal Democratic State,” Canadian Journal of Economics and Political Science, XXIV, no. 3 (Aug. 1958), 316.Google Scholar

³³ In 1957 the Liberal party received a greater share of the total vote than, the Progressive Conservative party but won a smaller number of seats. As Table iii shows, the Conservative party owed its capture of the government in part to discrimination in its favour by the electoral system.

³⁴ I realize that not everyone considers majority and one-party government as a “benefit,” but to defend my use of this term within the limited scope of this article is impossible.

³⁵ I have defined “regional third parties” as those that failed to poll at least 5 per cent of the valid vote in at least five provinces. Other third parties have been classified as “national.” An alternative criterion for differentiating between these two types of party might be the number of seats contested. If the cutting point in this case is set at 100 seats contested, the results are virtually identical to those reported in the text.

³⁶ In Great Britain and the United States the relationship between vote and seat shares has been found to be roughly cubic in nature. In other words, if a and b are parties in a two-party situation, seat share A: seat share b = vote share a³: vote share b³. A similar relationship has been found to exist in Canada, provided two-party, three-party, four-party, and five-party contests are considered separately. See Qualter, Terence H., “Seats and Votes: An Application of the Cube Law to the Canadian Electoral System,” this Journal, I, no. 3 (Sept. 1968), 336–44Google Scholar, and sources there cited on the British and American experience. Compare, however, a recent study by Duff Spafford, who shows that a major source of the fluctuation in major-party seat snares in Canada has been the variation in the number of third-party candidates. “The Electoral System of Canada,” American Political Science Review, LXIV (March 1970), 168–75.

³⁷ The number of Conservative mps from Quebec has varied since 1921 as follows: 1921, 0; 1925, 4; 1926, 4; 1930, 24; 1935, 5; 1940, 1; 1945, 2; 1949, 2; 1953, 4; 1957, 9; 1958, 50; 1962, 14; 1963, 8; 1965, 8; 1968, 4.

³⁸ Cairns, “The Electoral System,” 71–2, 76–7.

³⁹ For discussion of the “intensity” problem in democratic theory, see Dahl, Robert A., A Preface to Democratic Theory (Chicago and London, 1956), especially chap. 4Google Scholar; Kendall, W. and Carey, G. W., “The ‘Intensity’ Problem and Democratic Theory,” American Political Science Review, LXII (March 1968), 5–24.Google Scholar

⁴⁰ See, for instance, Dye, T. R, Politics, Economics and the Public (Chicago, 1966)Google Scholar; Dawson, R. E., “Social Development, Party Competition, and Policy,” in Chambers, W. N. and Burnham, W. D., eds., The American Party Systems (London and Toronto, 1967), 203–37Google Scholar; Pulsipher, A. G. and Weatherby, J. L. Jr., “Malapportionment, Party Competition and the Functional Distribution of Governmental Expenditures,” American Political Science Review, LXII (Dec. 1968), 1207–19, and LXIII (June 1969), 528–31CrossRefGoogle Scholar; Cnudde, C. F. and McCrone, D. J., “Party Competition and Welfare Policies in the American States,” ibid., LXIII (Sept. 1969), 858–66Google Scholar; I. Sharkansky and R. I. Hofferbert, “Dimensions of State Politics, Economics, and Public Policy,” ibid., 867–79; J. L. Walker, ‘The Diffusion of Innovations among the American States,” ibid., 880–99; Fry, B. R. and Winters, R. F., “ The Politics of Redistribution,” ibid., LXIV (June 1970), 508–522.Google Scholar

⁴¹ For some of the most useful studies, see Schattschneider, Party Government, chap. 4; Duverger, Political Parties, 203–55; critiques of Duverger's work by Leys, Colin and Wildavsky, Aaron B., reprinted in Eckstein, Harry and Apter, David F., eds., Comparative Politics (New York, 1963), 305–15Google Scholar and 368–75 respectively; Lakeman, Enid and Lambert, James D., Voting in Democracies (London, 1955)Google Scholar; Grumm, John G., “Theories of Electoral Systems,” Midwest Journal of Political Science, II (1958), 357–76CrossRefGoogle Scholar; Lipset, S. M., The First New Nation (Garden City, ny, 1967), chap. 9Google Scholar; Giovanni Sartori, “European Political Parties: The Case of Polarized Pluralism,” in LaPalombara and Weiner, Political Parties and Political Development, 137–76; Rae, The Political Consequences of Electoral Laws.

⁴² Political Oppositions in Western Democracies (New Haven, 1966), 351. Rae has subsequently found a “strong association” between plurality formulae and two-party systems, but his definition of a two-party system (“those in which the first party holds less than 70% of the legislative seats, and the first two parties together hold at least 90% of the seats”) weakens this finding by including countries like the United Kingdom among the two-party systems. The Political Consequences of Electoral Laws, 93–5.

⁴³ As is argued by Sartori, the adoption of proportional representation is unlikely to open the way to “extreme” party pluralism where it is met “by an already structured and stabilized party system.” “European Political Parties,” 167–76.

⁴⁴ See pp. 506–7 above.

⁴⁵ Grumm, “Theories of Electoral Systems”; Epstein, Leon D., Political Parties in Western Democracies (New York, 1967), 37–45Google Scholar; Lipson, Leslie, “Party Systems in the United Kingdom and the Older Commonwealth: Causes, Resemblances and Variations,” Political Studies, VII (1959), 12–31.CrossRefGoogle Scholar

⁴⁶ Political Parties, 37. For a similar view, see Sorauf, Frank J., Political Parties in the American System (Toronto, 1967), 27–32.Google Scholar

⁴⁷ “The Electoral System,” 77–8.

⁴⁸ Cairns thinks he knows already why Canada has one electoral system and not another: Canadians share a “cultural bias” against systems of proportional representation. Moreover, habit has so ingrained the present system that any hope of change would be illusory. Ibid., 55–6. It may be noted that evidence supporting these propositions is virtually nonexistent, and that Cairns himself calls them into question by advocating study of the merits of adopting proportional representation in Canada.