1 Taagepera, Rein and Shugart, Matthew Soberg, Seats and Votes: The Effects and Determinants of Electoral Laws (New Haven: Yale University Press, 1989), 2–3Google Scholar; see also Rae, Douglas W., The Political Consequences of Electoral Laws (New Haven: Yale University Press, 1967 and 1971).Google Scholar

2 This does not exhaust criticism of Canada's current electoral system. For example, Alan Cairns argues that “the emphasis on sectional divisions engendered by the electoral system has submerged class conflicts” so that “to the extent that our politics have been ameliorative it has been more concerned with the distribution of burdens and benefits between sections than between classes” (“The Electoral System and the Party System in Canada,” this Journal 1 [1968], 74). Blais and Carty argue that plurality electoral rules depress voter turnout in comparison with proportional representation rules ( Blais, André and Carty, R. K., “Does Proportional Representation Foster Voter Turnout?” European Journal of Political Research 18 [1990], 167–81).CrossRefGoogle Scholar These criticisms will not be addressed directly in this article.

3 Taagepera and Shugart, Seats and Votes, 4–5.

4 For an argument that the Canadian party system is operated as a cartel by the established parties, see McIvor, Heather, “Do Canadian Political Parties Form a Cartel?” this Journal 29 (1996), 317–33.Google Scholar

5 The best-known critique is Lijphart, Arend, Democracies: Patterns of Majoritarian and Consensus Government in Twenty-One Countries (New Haven: Yale University Press, 1984).CrossRefGoogle Scholar For two recent critiques, see Royal Commission on the Electoral System, Towards A Better Democracy (Wellington, New Zealand: Government Printer, 1986)Google Scholar; Harper, Stephen and Flanagan, Tom, “Our Benign Dictatorship,” Next City (Winter 1996/1997), 34–57.Google Scholar

6 Cairns, “The Electoral System and the Party System,” 56. Assessments of the relative efficacy of majority and minority government by politicians, academics and the public have been neither unanimous nor stable, however. See Wiseman, Nelson, “Cairns Revisited—The Electoral System and the Party System in Canada,” in Fox, Paul W. and White, Graham, eds., Politics: Canada (7th ed.; Toronto: McGraw-Hill Ryerson, 1991), 265–74Google Scholar, esp. 267, and Leduc, Lawrence, “Political Behaviour and the Issue of Majority Government in Two Federal Elections,” this Journal 10 (1977), 311–39.Google Scholar For cross-national evidence on the effectiveness of single-member plurality electoral rules in producing majority governments, see Blais, André and Carty, R. K., “The Impact of Electoral Formulae on the Creation of Majority Governments,” Electoral Studies 6 (1987), 209–18CrossRefGoogle Scholar, and Blais, André and Carty, R. K., “The Effectiveness of the Plurality Rule,” British Journal of Political Science 18 (1988), 550–53.CrossRefGoogle Scholar

7 See Table 5 for data on proportionality in vote-to-seat conversions for individual parties.

8 John Courtney points out that Canadian constitutional and political practice allows enormous flexibility with respect to who gets to form a government after an election in which no single party wins a clear majority in the House of Commons. In the recent past, however, the party that has won the most seats has formed the government. See the discussion in Courtney, John C., “Reflections on Reforming the Canadian Electoral System,” Canadian Public Administration 23 (1980), 427–57.CrossRefGoogle Scholar

9 See Table 5; see also Seidle, F. Leslie, “The Canadian Electoral System and Proposals for Reform,” in Tanguay, A. Brian and Gagnon, Alain G., eds., Canadian Parties in Transition (2nd ed.; Scarborough: Nelson Canada, 1996), 284.Google Scholar

10 The Reform party ran candidates nationwide outside Quebec but received only 42 per cent of its vote east of Manitoba in the 1993 national election.

11 Seidle, “The Canadian Electoral System and Proposals for Reform,” 282–306, esp. 288.

12 Johnston, Richard and Ballantyne, Janet “Geography and the Electoral System,” this Journal 10 (1977), 857–66.Google Scholar

13 See Wiseman, “Cairns Revisited,” 269–70.

14 See, for example, Cairns, “The Electoral System and the Party System”; Leduc, Lawrence, “Performance of the Electoral System in Recent Canadian and British Elections: Advancing the Case for Electoral Reform,” in Holler, M. J., ed., The Logic of Multiparty Systems (Dordrecht: Martinus Nijhoff, 1987), 341–58CrossRefGoogle Scholar; Canada, Task Force on Canadian Unity, A Future Together: Observations and Recommendations (Ottawa: Supply and Services Canada, 1979)Google Scholar; and Courtney, “Reflections on Reforming the Canadian Electoral System.”

15 Cairns claims that disparities in regional representation may also lead to regional biases in parties’ policies in favour of those regions where their representation is most numerous and where additional favours granted may create constituency pluralities (“The Electoral System and the Party System,” 68–72). Lovink, J. A. A. argues that such claims about biases in party policies are almost impossible to verify empirically (“On Analyzing the Impact of the Electoral System on the Party System in Canada,” this Journal 3 [1970], 497–521Google Scholar; see also Courtney, “Reflections on Reforming the Canadian Electoral System”). In any case, claims about the actual extent of party bias in policy are not essential to the argument of this article: it is enough that regional exclusion leads to public perceptions that a governing party does not represent a region's interests.

16 For a critical evaluation of claims that the combination of electoral volatility (which is also associated with SMP electoral rules) and geographic balance in cabinet building has created problems for governing parties in Ottawa, see Sutherland, S. L., “The Consequences of Electoral Volatility: Inexperienced Ministers, 1949–1990,” in Bakvis, Herman, ed., Representation, Integration and Political Parties in Canada, Research Studies of the Royal Commission on Electoral Reform and Party Financing, Vol. 14 (Toronto: Dundurn Press for Supply and Services Canada, 1991), 303–54.Google Scholar

17 Trudeau, Pierre Elliott, Memoirs (Toronto: McClelland and Stewart, 1993), 272.Google Scholar

18 Cairns, “The Electoral System and the Party System,” 57. For a critique of Cairns on this point, see Wiseman, “Cairns Revisited,” 266.

19 The 1958 seat figure for the Liberals includes one Liberal Labour seat in Ontario.

20 For a discussion, see Dobrowolsky, Alexandra and Jenson, Jane, “Reforming the Parties: Prescriptions for Democracy,” in Phillips, Susan D., ed., How Ottawa Spends, 1993–1994: A More Democratic Canada…? (Ottawa: Carleton University Press, 1993), 43–81Google Scholar; and Canada, Royal Commission on Electoral Reform and Party Financing, Reforming Electoral Democracy (Ottawa: Supply and Services Canada, 1991).Google Scholar On representation of Aboriginal peoples, see also Milen, Robert A., ed., Aboriginal Peoples and Electoral Reform in Canada, Research Studies of the Royal Commission on Electoral Reform and Party Financing, Vol. 9 (Toronto: Dundurn Press for Supply and Services Canada, 1991)Google Scholar; and Schouls, Tim, “Aboriginal Peoples and Electoral Reform in Canada: Differentiated Representation versus Voter Equality,” this Journal 29 (1996), 728–49.Google Scholar

21 Sylvia Bashevkin, “Political Parties and the Representation of Women,” in Tanguay and Gagnon, eds., Canadian Parties in Transition, 479–95.

22 A third option is to have a combination of single-member districts and “additional seats” that are awarded separately on the basis of proportional representation, rather than to compensate parties that have failed to get single-member district seats proportionate to their share of the vote. Because the additional seats system does very poorly at solving the problem of regional underrepresentation unless the ratio of these seats to single-member districts is relatively high, I will not consider it here. For reviews of proposals to introduce proportionality in the House of Commons, see Smiley, Donald V. and Watts, Ronald L., Intrastate Federalism in Canada (Toronto: University of Toronto Press, 1985)Google Scholar, chap. 6; Fox, Paul W., “Should Canada Adopt Proportional Representation?” in Fox, Paul W. and White, Graham, eds., Politics: Canada (7th ed.; Toronto: McGraw-Hill Ryerson, 1991), 343–50Google Scholar; Irvine, William P., “A Review and Evaluation of Electoral System Reform Proposals,” in Aucoin, Peter, research coordinator, Institutional Reforms for Effective Government (Toronto: University of Toronto Press, 1985), 71–109Google Scholar; and Seidle, “The Canadian Electoral System and Proposals for Reform.”

23 William P. Irvine argues that because “it is unlikely that any single political party will be able to constitute the government by itself” under proportional representation, party “leaders will therefore seek to lower the temperature of political controversy and to maximize a party's acceptability as a coalition partner” (“Measuring the Effects of Electoral Systems on Regionalism,” Electoral Studies 7 [1988], 16). But there is no inherent conflict between making strong regional appeals and being able to enter into a coalition government as a defender of regional interests.

24 Lijphart, Arend, Democracy in Plural Societies: A Comparative Exploration (New Haven: Yale University Press, 1977).Google Scholar

25 Neil Sutherland has shown that a system of two-member districts for House of Commons elections could maintain a probability of single-party majority government similar to that of the current SMP system while lowering regional disparities among parties (“Regionalism, Cabinet Stability and Canada's Electoral System: The Effect of District Magnitude,” this Journal 29 [1996], 497–519). However, the potential of a two-member system to reward moderately large regionally oriented parties makes it difficult to be confident that the distribution of party choices and votes in such a system would be similar to that under the current system. A two-member-district system might in fact stimulate significantly the growth of regionally oriented parties. Moreover, the paucity of experience with two-member districts in Canada or internationally raises political obstacles to their acceptance.

26 LeDuc, “Performance of the Electoral System,” 351.

27 See the discussion in Seidle, “The Canadian Electoral System and Proposals for Reform.”

28 See Irvine, “A Review and Evaluation of Electoral System Reform Proposals,” for a discussion of Canadian proposals for compensation-seat electoral reforms.

29 See Blais, André and Massicotte, Louis, “Electoral Systems,” in Leduc, Lawrence, Niemi, Richard and Norris, Pippa, eds., Comparing Democracies: Elections and Voting in Global Perspective (Thousand Oaks, Calif.: Sage, 1996), 199Google Scholar; and Blais, André and Massicotte, Louis, “Mixed Electoral Systems: An Overview,” Representation 33 (1996), 115–19.CrossRefGoogle Scholar

30 See, for example, Kaase, Max, “Personalized Proportional Representation: The ‘Model’ of the West German Electoral System,” in Lijphart, Arend and Grofman, Bernard, eds., Choosing an Electoral System: Issues and Alternatives (New York: Praeger, 1984), 155–74Google Scholar; and Hawke, Gary R., Changing Politics? The Electoral Referendum 1993 (Wellington, N.Z.: Institute of Policy Studies, Victoria University of Wellington, 1993).Google Scholar

31 Boston, Jonathan, “Electoral Reform in New Zealand,” Australian Quarterly 663 (1994), 67–90.CrossRefGoogle Scholar

32 For cross-national evidence to the contrary, see Lijphart, Arend, “Democracies: Forms, Performance and Constitutional Engineering,” European Journal of Political Research 25 (1994), 1–17.CrossRefGoogle Scholar

33 Roberts, Geoffrey K., “The ‘Second-Vote’ Campaign Strategy of the West German Free Democratic Party,” European Journal of Political Research 16 (1988), 317–37.CrossRefGoogle Scholar

34 In most elections, the rank ordering of major parties by seat share and vote share is the same. There are occasional exceptions, notably 1957, 1962 and 1979, when the Liberals outpolled the Progressive Conservatives but won fewer seats. But in these situations, one of the two largest parties is generally overrepresented in each province anyway, and thus ineligible for compensation seats. Compensation seats awarded with a bias to the largest parties would move between these two parties only if both parties are underrepresented in a province, a relatively infrequent occurrence. In the 1993 election, the Bloc Québécois was ranked second in constituencies won but only fourth in nationwide vote share, but this would not entitle it to any additional compensation seats since it had no vote share in other provinces.

35 On attachment to the current Single Member Plurality system, see Irvine, William P., Does Canada Need a New Electoral System (Kingston: Queen's University Institute of Intergovernmental Relations, 1979), 67, 71–73.Google Scholar

36 On the importance of strategic voting in Canada, see Blais, André and Nadeau, Richard, “Measuring Strategic Voting: A Two Step Procedure,” Electoral Studies 15 (1996), 39–52.CrossRefGoogle Scholar More generally, on strategic voting in SMP systems, see Cox, Gary W., Making Votes Count: Strategic Coordination in the World's Electoral Systems (Cambridge: Cambridge University Press, 1997)CrossRefGoogle Scholar, chap. 4.

37 Cassidy, Michael, “Fairness and Stability: How a New Electoral System Would Affect Canada,” Parliamentary Government 42 (1992), 3–27.Google Scholar The results of this study were updated through the 1993 election in Seidle, “The Canadian Electoral System and Proposals for Reform.”

38 Occasionally, allocating an additional compensation seat would give any of the parties a greater share of seats than it had of the popular vote in the province. Usually this occurs in provinces with few seats and a close allocation of seats between the two leading parties—for example, if the two leading parties have split New Brunswick's 10 seats equally and have relatively equal vote shares. In this situation, the seat is allocated to the party for which awarding the seat would create the least overcompensation. This method is used to deal with all overcompensation situations that arise in the simulations summarized here.

39 See, for example, Young, Lisa, “Electoral Systems and Representative Legislatures: Consideration of Alternative Electoral Systems” (Ottawa: Canadian Advisory Council on the Status of Women, 1994).Google Scholar

40 It is important to use vote share rather than absolute number of votes received because of the high tolerance for deviations from the “one person-one vote” norm in drawing Canadian electoral boundaries. Using absolute vote totals to fill compensation seats would presumably produce a systematic bias in favour of candidates from more populous urban districts. On difference in district population size, see, for example, John C. Courtney, “Drawing Electoral Boundaries,” in Tanguay and Gagnon, eds., Canadian Parties in Transition, 328–48; and Courtney, John C., Mackinnon, Peter and Smith, David E., eds., Drawing Boundaries: Legislatures, Courts and Electoral Values (Saskatoon: Fifth House, 1992).Google Scholar

41 For an individual province, the number of compensation seats would be set at the whole number that is closest to one eleventh of the total number of seats for the province. This means that Prince Edward Island, as well as Yukon and the North-west Territories, would not have any compensation seats. For the purposes of the simulation, provinces with 5 to 14 seats under the SMP system were awarded 1 compensation seat, 15 to 24 constituencies 2 compensation seats and so on.

42 There is no minimum vote or seat threshold for a party to be eligible for winning compensation seats.

43 The index of proportionality is calculated by adding together the gap between percentage of seats won and percentage of the popular vote won for each party, dividing that sum by 2 and then subtracting that number from 100. See Richard Rose, “Electoral Systems: A Question of Degree or of Principle?” in Lijphart and Grofman, eds., Choosing an Electoral System, 73–81.

44 Irvine argues that in a two-party system, the most disproportionate result would be for one party to win all constituency seats with just over 50 per cent of the votes; thus the minimum IP score would be 50. In a three-party system, the minimum score would be 33.3, in a four-party system 25 and so on. Thus he suggests a “rescaled Index of Proportionality,” based on the distance between the minimum possible IP score in that election and 100 (“Measuring the Effects of Electoral Systems on Regionalism”). The RIP approach has problems of its own, however. In the 1993 federal election, for example, five parties won seats, but the Bloc Québécois and the Reform party did not compete outside and inside Quebec, respectively. Moreover, other parties (such as the National party and the Natural Law party) did compete and had their results tallied by the Chief Electoral Officer. In short, the number of parties “really” competing in an election is to some extent a judgment call by the observer. To avoid this problem, the standard Index of Proportionality is used here.

45 Irvine, “Measuring the Effects of Electoral Systems on Regionalism.”

46 The formula used is:

Weighted Seat Share = Σ [(IP_r) * V_r/V_c] where IP_r is the Index of Proportionality for a region r, V_r is the total valid vote in region r and V_c is the total valid vote in the country as a whole. This approach differs from Irvine's, which bases its weighting on regions’ seat shares of the national total rather than vote shares.

47 A national-level tier of compensation seats is a much more efficient mechanism for ensuring that the political party that is the plurality-vote winner is also the plurality-seat winner than awarding all constituency and compensation seats at the provincial level, as proposed here. See especially the proposal in Lijphart, Arend, Electoral Systems and Party Systems: A Study of 27 Democracies, 1945–1990 (New York: Oxford University Press, 1994), 146–48.CrossRefGoogle Scholar National-level compensation seats are a more substantial constitutional change than those considered here, however.

48 The Liberal total already includes one Independent Liberal from Quebec.

49 In New Brunswick, the Conservatives finished with a margin of 1,512 votes over the Liberals, and each party won five seats. Thus each party is slightly overcompensated. The tie-breaker mechanism then comes into play, awarding New Brunswick's compensation seat to the Conservatives because it would result in less disproportionality than awarding it to the Liberals, CCF or Social Credit. In Saskatchewan, the Liberals won 30.3 per cent of the vote of parties passing the threshold, but awarding both of Saskatchewan's compensation seats to the Liberals would have given them a higher seat share (31.6 per cent) than vote share in the province; thus the Conservatives would be awarded the second compensation seat.

50 It is worth noting in this regard that a shift of 5 per cent of the Reform vote and just over 2 per cent of the Liberal vote in the 1993 election to the Progressive Conservatives would have moved the Conservatives up to second place in popular vote, and thus qualified them for the 12 compensation seats won by the Reform party under the compensation seat system proposed here. A shift of this magnitude does not seem implausible, given the “negative bandwagon” effects of the current electoral system on Conservative voters, who confronted widespread press speculation that their party would be wiped out in the 1993 election. A shift of only 28,000 voters in 1997 from Reform to Conservative (1.1% of Reform voters) would have moved the Conservatives ahead of Reform in the national popular vote, entitling them to Reform's 10 compensation seats in Ontario.

51 See the discussion of the importance of nominating procedures affecting parliamentary representation by women in Matland, Richard E., “Institutional Variables Affecting Female Representation in National Legislatures: The Case of Norway,” Journal of Politics 55 (1993), 737–55.CrossRefGoogle Scholar

52 This is consistent with Matland's finding that party magnitude as well as district magnitude matters in determining parliamentary representation of women, at least until a norm that there should be equal representation of women is established (see Matland, “Institutional Variables Affecting Female Representation”). See also Welch, Susan and Studlar, Donley T., “Multi-Member Districts and the Representation of Women: Evidence from Britain and the United States,” Journal of Politics 52 (1990), 391–412CrossRefGoogle Scholar; and Rule, Wilma and Zimmerman, Joseph, eds., Electoral Systems in Comparative Perspective: Their Impact on Women and Minorities (Westport, Conn.: Greenwood, 1994).Google Scholar

53 Section 42 of the Constitution Act, 1982 requires that “amendments to the Constitution of Canada… in relation to… the principle of proportionate representation of the provinces in the House of Commons prescribed by the Constitution of Canada” be made in accordance with the general amending formula (resolutions of the Parliament and two thirds of provincial legislatures comprising at least 50 percent of the population). For a discussion of constitutional restrictions on electoral reform, see Irvine, “A Review and Evaluation of Electoral System Reform Proposals,” 103–06.

54 Boston, Jonathan, Levine, Stephen, McLeay, Elizabeth and Roberts, Nigel S., “Why Did New Zealand Adopt German-Style Proportional Representation?” Representation 33 (1996), 134–40CrossRefGoogle Scholar, esp. 137.

55 See Irvine, Does Canada Need a New Electoral System? 71–73.

56 See Boston, Jonathan, Levine, Stephen, McLeay, Elizabeth and Roberts, Nigel S., “The 1996 General Election in New Zealand: Proportional Representation and Political Change,” Australian Quarterly 69 (forthcoming 1997).CrossRefGoogle Scholar

57 See Bakvis, Herman and Macpherson, Laura G., “Quebec Block Voting and the Canadian Electoral System,” this Journal 28 (1995), 659–92.Google Scholar

58 Of course, if Quebeckers voted disproportionately for a single nationwide party in federal elections in the future, as they did prior to 1993, that party would have an advantage in federal elections, and Quebeckers would once again likely be overrepresented in the governing party caucus relative to their share of the population. However, a system of compensation seats would lessen that imbalance. For data on Quebec overrepresentation in the governing party caucus in the House of Commons, see ibid.