¹ Polsby, N.W. and Wildavsky, A.B., Presidential Elections, 2nd ed. (New York, 1968), 280–1Google Scholar, (italics added).

² Southern Politics in State and Nation (New York, 1949), esp. 302–10.

³ Politics, Parties and Pressure Groups, 5th ed. (New York, 1964), 298–9.

⁴ Some of these studies have also examined the association between the level of electoral competition and other policies, not directly related to the so-called “have–have-not struggle” in society. One, by Fry and Winters (cited in n. 3), uses as its policy variable not the level of social welfare expenditures per recipient but the allocation of the burdens and benefits of state revenue and expenditure policies across income classes, that is, the redistributive character of state policies.

⁵ Dawson, R.E. and Robinson, J.A., “Inter-Party Competition, Economic Variables, and Welfare Policies in the American States,” Journal of Politics, 25 (1963), 265–89CrossRefGoogle Scholar; Hofferbert, R.I., “The Relation between Public Policy and Some Structural and Environmental Variables in the American States,” American Political Science Review, 60 (1966), 73–82CrossRefGoogle Scholar; Dye, T.R., Politics, Economics and the Public (Chicago, 1966Google Scholar); Dawson, R.E., “Social Development, Party Competition and Policy,” in The American Party Systems, ed. Chambers, W.N. and Burnham, W.D. (Toronto, 1967), 203–37Google Scholar; Fry, B.R. and Winters, R.F., “The Politics of Redistribution,” American Political Science Review, 64 (1970), 508–22.CrossRefGoogle Scholar

⁶ Pulsipher, A.G. and Weatherby, J.L. jr., “Malapportionment, Party Competition, and the Functional Distribution of Governmental Expenditures,” American Political Science Review, 62 (1968), 1207–19.CrossRefGoogle Scholar

⁷ Cnudde, C.F. and McCrone, D.J., “Party Competition and Welfare Policies in the American States,” ibid., 63 (1969), 858–66.Google Scholar

⁸ I. Sharkansky and R.I. Hofferbert, “Dimensions of State Politics, Economics, and Public Policy,” ibid., 867–79, at 878.

A measure of support for this second set of findings has also been provided by a recent study by Prewitt and Eulau. Based on interviews with some 400 city councilmen in the San Francisco Bay area, this research led its authors to conclude that the more seriously councilmen were threatened by electoral defeat the more likely they were to see themselves as responding to the wishes of their constituents rather than to their own image of the public interest. K. Prewitt and H. Eulau, “Political Matrix and Political Representation: Prolegomenon to a New Departure from an Old Problem,” ibid., 427–41. No data are presented on the fit between the responsiveness of councilmen as seen by themselves and as judged by their constituents. See also Prewitt, K., “Political Ambitions, Volunteerism, and Electoral Accountability,” ibid., 64 (1970), 5–17.Google Scholar

⁹ In most of these studies, the measurement of competition is based on one or more state offices, an appropriate restriction given that only state policies are being considered. In two cases (Hofferbert, “Public Policy and Structural and Environmental Variables,” and Fry and Winters, “Politics of Redistribution”), however, the measure of competition includes the outcomes of elections for U.S. president and senator as two of its three components, thereby casting considerable doubt on the validity of the relationships observed.

¹⁰ The exception is the study by Sharkansky and Hofferbert, “Dimensions of State Politics,” whose index of competition includes several components measured at a single point in time. Some of their other components are measured over time, however, without any explanation of the rationale involved.

¹¹ “Toward Functionalism in Political Science: The Case of Innovation in Party Systems,” American Political Science Review, 57 (1963), 570–83, at 575.

¹² Ibid.

¹³ Ambition and Politics (Chicago, 1966), 121–5.

¹⁴ “The Diffusion of Innovations among the American States,” American Political Science Review, 63 (1969), 880–99, at 882–3.

¹⁵ Ibid., 885–7.

¹⁶ In support of this conclusion, doubts may also fairly be expressed about Walker's measures of electoral competition (borrowed from Hofferbert), which rely on gubernatorial elections in one instance, and on a combination of presidential, senatorial, and gubernatorial elections in another. Short of a demonstration that competition for these offices is a good proxy for competition for the state legislature, this does not seem satisfactory.

As far as the coincidence between the content and timing criteria of innovation is concerned, Professor Walker has pointed out in a letter to the writer that secondary sources he has consulted assert that in many cases the content of state programs is very similar (sometimes being based on model bills that are circulated among the states). He is confident, on that basis, that in “the majority of cases” no significant differences exist.

In so far as his list of programs includes many relatively uncontroversial items like accountants licensing, this confidence is probably well founded. It is less plausible, however, with respect to the more politically salient programs like fair housing, resale price maintenance (fair trade laws), and right-to-work legislation, precisely those to which the level of electoral competition has been presumed most significant.

¹⁷ Pinard, M., The Rise of a Third Party (Scarborough, Ont., 1971), 37.Google Scholar In the author's own words, one-party dominance is “conducive” to the rise of new parties. Ibid.

¹⁸ Ibid., 39.

¹⁹ Wakefield, Kimberly, “Measuring One-Party Dominance” (MA thesis, Queen's University, 1972Google Scholar).

²⁰ Porter, John, The Vertical Mosaic (Toronto, 1965), 405–6.CrossRefGoogle Scholar In accounting for the scarcity of political careerists and what he sees as the consequent lack of autonomy in the Canadian political system, Porter is not satisfied with the uncertainty of a Canadian political career as an adequate answer. He also cites the absence of a financially secure élite willing to make a career of politics, as well as the allegedly bureaucratic outook of many Liberal leaders, beginning with Mackenzie King.

²¹ Cf. Downs, An Economic Theory of Democracy, chap. 4, esp. 62–3.

²² American State Politics: An Introduction (New York, 1956), 229.

²³ Ibid., 228–9.

²⁴ Ibid., 240–41.

²⁵ Ibid., 230–42.

²⁶ Reporting a clearcut relationship between the two variables are: Golembiewski, R.T., “A Taxonomic Approach to State Political Party Strength,” Western Political Quarterly, 11 (1958), 494–513CrossRefGoogle Scholar; Eulau, H., “The Ecological Basis of Party Systems: The Case of Ohio,” Midwest Journal of Political Science, 1 (1957), 125–35CrossRefGoogle Scholar; Cutright, P., “Urbanization and Competitive Party Politics,” Journal of Politics, 25 (1963), 552–64CrossRefGoogle Scholar; Dawson and Robinson, “Inter-party Competition …,” 283–5; Ranney, A., “Parties in State Politics,” in Politics in the American States, ed. Jacob, H. and Viner, K. N. (Toronto, 1965), 61–99, at 68–9.Google Scholar

The following arrive at contrary findings: Gold, D. and Schmidhauser, J.R., “Urbanization and Party Competition: The Case of Iowa,” Midwest Journal of Political Science, 4 (1960), 62–75CrossRefGoogle Scholar; T.R. Dye, Politics, Economics and the Public, 57–8; Bonjean, G.M. and Lineberry, R.L., “The Urbanization-Party Competition Hypothesis: A Comparison of All United States Counties,” Journal of Politics, 32 (1970), 305–21.CrossRefGoogle Scholar

For conflicting views on the relationship in Canada, see Scarrow, H.A., “Patterns of Voter Turnout in Canada,” Midwest Journal of Political Science, 5 (1961), 351–64, at 360CrossRefGoogle Scholar; Grossman, L.R., “‘Safe’ Seats: The Rural-Urban Pattern in Ontario,” Canadian Journal of Economics and Political Science, 29 (1963), 367–71CrossRefGoogle Scholar; Kornberg, A., “Parliament in Canadian Society,” in Legislatures in Developmental Perspective, ed. Kornberg, A. and Musolf, L. (Durham, N.C, 1970), 55–128, at 71.Google Scholar

²⁷ For a convenient review article, see Coulter, P. and Gordon, G., “Urbanization and Party Competition: Critique and Redirection of Theoretical Research,” Western Political Quarterly, 21 (1968), 274–88.CrossRefGoogle Scholar

²⁸ Particularly relevant to this point are D.E. Stokes, “Party Loyalty and the Likelihood of Deviating Elections,” and Converse, P.E., “Information Flow and the Stability of Partisan Attitudes,” in Campbell, A.et al., Elections and the Political Order (New York, 1966), 125–35 and 136–57Google Scholar respectively.

²⁹ Campbell, A.et al., The American Voter (New York, 1966Google Scholar), chap. 15.

³⁰ A. Campbell and H. Valen, “Party Identification in Norway and the United States,” in Campbell et al., Elections and the Political Order, 245–68. See also Rose, R. and Urwin, D., “Social Cohesion, Political Parties, and Strains in Regimes,” in European Politics: A Reader, ed. Rose, R. and Dogan, M. (Toronto, 1971), 217–37, at 228Google Scholar, where the authors present data suggesting that for heterogeneous political parties the vote fluctuates much more than for parties which have a more clearly defined, restricted appeal. It has been noted, on the other hand, that once party identifications are firmly established a diminution in the ideological or political distinctiveness of the existing political parties need not (in fact, does not usually) bring about a decline in the prevalence or intensity of party identification. See, generally, Political Oppositions in Western Democracies, ed. Dahl, R.A. (New Haven, 1966), and esp. 380–6.Google Scholar

Variations in the pattern of party identification have also been related independently to differences in the politicization process. See P.E. Converse and G. Dupeux, “Politicization of the Electorate in France and the United States,” in Campbell et al., Elections and the Political Order, 269–91; but cf. Cameron, D.R. and Summers, L., “Non-Family Agents of Political Socialization: A Reassessment of Converse and Dupeux,” this Journal, 5 (1972), 418–32.Google Scholar

³¹ This point is made more fully in Campbell and Valen, “Party Identification in Norway and the us,” 265–8.

³² For a brief discussion of the ways in which the parties have approached this task, see below, pp. 372–3.

³³ Alternative approaches to measuring electoral competition are discussed below, pp. 356–61.

³⁴ Canada Votes (New Orleans, 1962).

³⁵ “A Research Note on Canadian Party Systems,” Canadian Journal of Economics and Political Science, 33 (1967), 98–106.

³⁶ “The Electoral System of Canada,” American Political Science Review, 64 (1970), 168–76.

³⁷ For CCF supporters, comparable proportions for the second and third dimensions were 68 per cent and 4 per cent, respectively. No meaningful comparison is possible for the first dimension, because in many constituencies CCF supporters had no candidate of their own on the ballot. Regenstreif, S.P., “Some Aspects of National Party Support in Canada,” Canadian Journal of Economics and Political Science, 29 (1963), 59–74.CrossRefGoogle Scholar Also, by the same author, The Diefenbaker Interlude (Toronto, 1965).

³⁸ Jenson, Jane, “Party Identification in Canada: A Rationally Limited Allegiance,” (paper presented at the Annual Meeting of the Canadian Political Science Association, St John's, Nfld., June, 1971), 6–7.Google Scholar See also Converse, P.E., Dupeux, G., and Meisel, J., “Continuities in Popular Political Cultures: French and Anglo-Saxon Contrasts in Canada,” (paper prepared for the international conference on comparative electoral behaviour, University of Michigan, 1967), 28–35.Google Scholar

³⁹ Campbell and Valen, “Party Identification in Norway and the us,” 264, whose data indicate that among strong party identifiers in the United States two-thirds have always voted for the same party for President; while among weak identifiers about 45 percent have always done so. For both categories, the comparable Norwegian percentages are higher (Christian party identifiers excepted).

⁴⁰ Working Papers on Canadian Politics (Montreal, 1972), 67. For further evidence on the weakness of Canadian party identifications, see the growing literature on federal-provincial vote switching, notably John Wilson and Hoffman, David, “The Liberal Party in Contemporary Ontario Politics,” this Journal, 3 (1970), 177–204Google Scholar; Perlin, George and Peppin, Patti, “Variations in Party Support in Federal and Provincial Elections,” ibid., 4, no. 2 (1971), 280–6.Google Scholar Whether such “ticket-splitting” is more common in Canada than in other federations is a moot point. Scarrow's analysis suggests that the reverse may be true, at least for the United States and Australia, but since his findings are based on changes in winning party (not on the popular vote) their validity remains in doubt. Scarrow, H.A., “Federal-Provincial Voting Patterns in Canada,” Canadian Journal of Economics and Political Science, 26 (1960), 289–98.CrossRefGoogle Scholar

⁴¹ “Party Identification in Canada,” 9 and passim.

⁴² See, for example, Scarrow, H.A., “Distinguishing Between Political Parties – the Case of Canada,” Midwest Journal of Political Science, 9 (1965), 61–76.CrossRefGoogle Scholar

⁴³ For those interested in examining this literature (most of it bearing on the United States), many of the more important studies are cited in Pfeiffer, D.G., “The Measurement of Inter-Party Competition and Systematic Stability,” American Political Science Review, 61 (1967), 457–67.CrossRefGoogle Scholar

⁴⁴ I am well aware of Pfeiffer's claim that electoral competition can be adequately measured using only the party distribution of the vote. In my view, Pfeiffer fails to provide the systematic empirical evidence that would be required to meet the criticisms of such a simple measure that are advanced in this article. Without such evidence, there is no reason to think that vote pluralities by themselves offer a valid basis of comparison, be it over time, between different offices within the same political system, or between comparable offices in different political systems. (Pfeiffer, “The Measure of Inter-Party Competition and Systemic Stability.”)

An important contemporary motivation for simplicity (also advanced by Pfeiffer) is the desire to facilitate statistical manipulation of the results. Thus, Bonjean and Lineberry think that the “most compelling” argument for adopting Pfeiffer's simple measure is “that it meets the requirements of all commonly used parametric tests”. Bonjean and Lineberry, “The Urbanization-Party Competition Hypothesis,” 309. This consideration loses whatever force it may have when a more complex alternative measure (such as the one proposed in this article) is equally easily manipulatable.

Finally, some reference should be made to the evidence recently provided that a variety of existing measures to compare the competitiveness of the American states all yield roughly similar results. This finding does not mean that a simple measure such as Pfeiffer proposes would necessarily also yield similar results, for the measures compared in the study cited are all relatively complex. It also does not obviate the need for the measure to be developed in this paper, for the measures examined were all tailored to institutional conditions peculiar to the United States. See Zody, R.E. and Luttbeg, N.R., “Evaluation of Various Measures of State Party Competition,” Western Political Quarterly, 21 (1968), 723–4.CrossRefGoogle Scholar

⁴⁵ Jones, C.O., “Inter-Party Competition for Congressional Seats,” Western Political Quarterly, 17 (1964), 461–76.CrossRefGoogle Scholar

⁴⁶ Rasmussen, J., “The Implication of Safe Seats for British Democracy,” ibid., 19 (1966), 516–29, at 517–9Google Scholar, updating findings reported in Jones, C.O., “Inter-Party Competition in Britain, 1950–59,” Parliamentary Affairs, 17 (1963–1964), 50–6.Google Scholar In both of these studies, only the 453 seats unaffected by boundary revisions are considered.

⁴⁷ In this period there were no federal boundary revisions in Canada, so that all 263 constituencies have been included in the calculation.

⁴⁸ For a short explanation of this point, see Jones, “Inter-Party Competition for Congressional Seats,” 465, n. 14.

⁴⁹ Rasmussen, “The Implication of Safe Seats for British Democracy,” 519–29.

⁵⁰ For an explanation of these options see the appendix to this paper.

⁵¹ David J. Elkins has recently proposed an approach to measuring electoral competition which would use survey data on each party's share of “standpatters” and “switchers” (in V.O. Key's language) to arrive at the statistical probability that the in-party will be displaced at the next election. “The Measurement of Party Competition in Multi-Party Systems” (unpublished manuscript, Department of Political Science, University of British Columbia, n.d. (1973?)). Although theoretically more satisfying than the present approach, the proposal's dependence on survey data disqualifies it for longitudinal studies reaching into the pre-survey era, and makes it too costly for use at the constituency level.

⁵² The proportion of Canada's population classified in census statistics as “urban” has increased from 49.5 per cent in 1921 to 76.1 per cent in 1971. Because of a change in definitions in 1956, these figures slightly overstate the actual rate of urbanization. Canada, Dominion Bureau of Statistics, 1961 Census of Canada, Bull. 7.1–2, Tables III and IV; Statistics Canada, 1971 Census of Canada, Bull. 1.1–9.

⁵³ I am grateful to Professor Meisel for letting me see these results of his work. The findings cited were drawn from cross-tabulating variable 271 (“Where grew up”) with variables 183 (“Strength of identification”) and 214 (“Party vote consistency-federal”) of the 1965 survey (N = 2721), and variables 002 (“Community size”) and 362 (“During schooling where did you live”) with variables 123 (“Degree of strength felt to chosen party”) and 174 (“What federal party have you always voted for”) of the 1968 survey (N = 2767).

⁵⁴ Because turnover is only one of the three components making up the index, and because involuntary retirement through electoral defeat is only one of the causes of turnover, the relationship is obviously neither necessary nor likely to be very close. Moreover, its reverse would probably often fail to hold true: in practice, a high rate of turnover has frequently coincided with relatively uncompetitive conditions. T.A. Dye, “State Legislative Politics,” in Politics in the American States, ed. Herbert Jacob and Kenneth Vines, 151–206, at 170; Blondel, J., Voters, Parties, and Leaders (Baltimore, 1963), 153–8.Google Scholar Also Schlesinger, Ambition and Politics, 60–3.

⁵⁵ Cf. Hoffman, David and Ward, Norman, Bilingualism and Biculturalism in the Canadian House of Commons (Ottawa, 1970), 27.Google Scholar By comparison, the average parliamentary experience of British MP S at the start of each Parliament was reported in 1963 to be eight years. Blondel, Voters, Parties, and Leaders, 155–6.

⁵⁶ See above, pp. 342–350.

⁵⁷ It is interesting to note that Canada stands midway between Britain and the United States in the proportion of lawyers in its national legislature. Of Kornberg's sample of the 25th Parliament of Canada, 33 per cent were lawyers, while Ward found exactly the same percentage in 1931. Kornberg, Allan, Canadian Legislative Behavior (Toronto, 1967), 43Google Scholar; Ward, Norman, The Canadian House of Commons: Representation (Toronto, 1950), 135.Google Scholar The comparable British and American figures are about 20 and 60 per cent respectively. The Times House of Commons, 1966 (London, The Times, n.d.), 242; Butler, David and King, Anthony, The British General Election of 1966 (London, 1966), 208–9CrossRefGoogle Scholar; Butler, David and Pinto-Duschinsky, Michael, The British General Election of 1970 (London, 1971), 302Google Scholar; Matthews, Donald, The Social Background of Political Decision-makers (Garden City, 1954), 30Google Scholar (US data for 1949–51); Davidson, Roger H., The Role of the Congressman (New York, 1969), 37Google Scholar (based on a sample of 87 Congressmen of the 88th Congress, 1962–4). On the face of it, this finding suggests that the link between the level of electoral competition and the occupational make-up of the legislature is rather weak. It may be, however, that closer inspection of each legislature may nonetheless show a certain association between these two variables.

⁵⁸ In the early sixties, Kornberg found constituency competitiveness as perceived by MPS to be totally unrelated to their “legislative goals” and “representational focus,” and to be only tenuously related to their “representational style.” Substituting constituency competitiveness as measured by Kornberg produced no associations whatever. Canadian Legislative Behavior, 92–3, 109, 111. For American data on this point, see Davidson, Role of the Congressman, 126–33, and Wahlke, J.C.et al., The Legislative System (New York, 1962), 291–4.Google Scholar

⁵⁹ The report that follows is based on open-ended interviews with 90 MPS (frontbench and backbench), out of a sample of 120, stratified by region and party and chosen at random except for minor modifications to ensure that very small minorities (like Quebec Conservative MPS) would be represented. The regional and party composition of the MPS actually interviewed closely resembled that of the sample (and of the House) as a whole.

Of the 90 MPS interviewed 10 were frontbenchers, of whom 7 were currently in the cabinet while the other 3 were former ministers. To supplement the resulting information on frontbench perceptions and attitudes, interviews were also arranged with 5 other frontbenchers to whom the writer happened to have some access. As a result, 95 MPS were actually interviewed.

All interviews were conducted in June and July, 1969, with the very able assistance of John Dingwall, then a graduate student at Yale University, and of André Carrier, then a graduate student at the University of Ottawa.

⁶⁰ Cf. Wilson, James Q., The Amateur Democrat (Toronto, 1966Google Scholar) for a description and analysis of this orientation to politics in the American context; also Roche, John P. and Sachs, Stephen, “The Bureaucrat and the Enthusiast,” in Policy-Making in Britain, ed. Rose, Richard (Toronto, 1969), 81–96.Google Scholar Cf. also the late V.O. Key's judgment: “The prediction of public reaction to new courses of action is so primitive an art that thoughtful politicians remark that the only course to follow is to take the action that seems right, on the assumption that it will be defensible at the next election.” Politics, Parties, and Pressure Groups, 544, n. 28.

⁶¹ For a brief discussion of the problems of measuring and explaining regional favouritism in party policies, see my “On Analysing the Impact of the Electoral System on the Party System in Canada,” this Journal, 3 (1970), 497–516, at 499–506.

⁶² Such conflict is likely to be avoided only in cases where the distribution of a party's safe and competitive seats closely resembles the distribution of its marginal seats won and lost, when pursuit of a seat-maximizing strategy would not be prejudicial to the interests of the party's less competitive constituencies.

⁶³ See Lowi, Theodore, “American Business, Public Policies, Case Studies and Political Theory,” World Politics, 16 (1964), 677–715.CrossRefGoogle Scholar

⁶⁴ See Meisel, John, The Canadian General Election of 1957 (Toronto, 1962), 166–8.Google Scholar