Secondary menu

Electoral Insight – Review of Electoral Systems

Electoral Insight – June 1999

Proportional and Semi-proportional Electoral Systems: Their Potential Effects on Canadian Politics

Heather MacIvor
Department of History, Philosophy and Political Science, University of Windsor

Canada's single-member plurality (SMP) electoral system is no longer acceptable. Our representative democracy faces serious problems, including declining voter turnout, public discontent with our political institutions, the under-representation of women and minorities in the House of Commons, and growing regional divisiveness.

While proportionality is not the only criterion for evaluating electoral systems, a more proportional system specifically, parallel mixed-member plurality (MMP) would likely alleviate the problems just mentioned. Other alternatives include "top-up" MMP, single transferable vote (STV), single non-transferable vote (SNTV), and proportional representation based on party lists (list-PR). The two MMP systems are the most promising alternatives for Canada. Both produce greater proportionality in translating parties' national and regional vote shares into parliamentary seats. Both offer women, Aboriginal persons and minor-party candidates a better chance of election to the House of Commons. Parallel MMP, the less proportional of the two, would still permit the election of single-party majority governments with less than half the popular vote. While no electoral system is perfect, parallel MMP would likely reduce the problems described above without necessarily making matters worse.

Ideally, an electoral system encourages voter turnout by giving each elector a meaningful influence over the outcome. SMP does not meet this standard. The percentage of eligible voters who exercise their franchise has fallen over the past four Canadian general elections, from 75 percent in 1984 to 67 percent in 1997.Footnote 1 While individual motivation to vote depends on several factors including age, education, political interest, and satisfaction with the available choices it is also influenced by the belief that one's vote is "wasted."Footnote 2 In the 1997 Canadian election, almost 65 percent of the successful candidates won fewer than half of the valid votes in their ridings.Footnote 3 So at least half of the ballots were "wasted," in the sense that they did not elect members of Parliament. Comparative studies conclude that "turnout is over eight percent higher in a PR election than in a plurality one."Footnote 4 Finally, SMP discourages voters who wish to cast their ballots for parties or candidates with little or no chance of winning. This "psychological effect," which punishes "third" parties and discourages their supporters over the long term, reduces their vote share under SMP by about 20 percent relative to PR systems.Footnote 5 The greater the proportionality, the weaker the psychological effect.

Electoral systems directly affect the composition of legislative chambers. The more proportional the system, the more accurately the parties' seat shares in Parliament reflect their respective shares of the popular vote. Disproportional systems, such as SMP, distort party caucuses, both in their relative sizes and in terms of their regional composition. In a 1997 survey of proportionality in national elections from 1945 to 1996, Canada ranked 35th of the 37 cases studied.Footnote 6 Although the distortion of the parties' parliamentary representation is often expressed in abstract mathematical terms, SMP does real harm to some parties while unfairly helping others. In 1993 the federal Conservatives won 16 percent of the national vote, and less than one percent of the seats in the Commons. Although it had received millions of votes, the party was placed on a death watch because it elected only two MPs. In the 1997 election the Reform Party won 66 365 more votes than the Conservatives (out of almost 13 million), but its dominance in the West gave it three times as many seats.

SMP favours the largest party and regionally concentrated "third" parties such as the Bloc Québécois at the expense of the second-largest national party and "third" parties with widely diffused support (like the New Democratic Party [NDP]).Footnote 7 The brokerage function of the national parties in Parliament is seriously compromised. By over-rewarding regionally concentrated votes, SMP gives Canadian parties a powerful incentive to engage in divisive appeals. It also protects a governing party from the political consequences of losing support in smaller regions of the country if it can sweep a populous province such as Ontario, and it deprives entire provinces of cabinet representation. Two-thirds of the current Liberal caucus comes from Ontario; 73 of the West's 88 seats are held by opposition parties, as are 21 of the 32 seats in the Atlantic and 49 of Quebec's 75 seats. Some would argue that if voters outside Ontario wanted to be represented in cabinet, they would have voted Liberal. But millions of voters did vote Liberal, and their votes were wasted. These seat distributions distort the wishes of the voters.

In recent years, the emphasis in the debate over SMP has shifted to the demographic reflection of the electorate in the House of Commons. Women constitute slightly over half of the Canadian electorate, but have never accounted for more than 21 percent of the House of Commons. While women are in a minority in every national legislature, female representation in the Canadian parliament is only in the middle rank of Western democracies.Footnote 8 Canada ranks 11th out of 28 democracies in the proportion of female cabinet ministers.Footnote 9 There are several reasons why women are less likely than men to seek election to Parliament and why, once nominated, they are less likely to win. These include the wage gap, the gendered division of labour at home and at work, and persistent derogatory stereotypes about women.Footnote 10 But cross-national surveys consistently identify the electoral system as the most important variable affecting women's representation in national legislatures.Footnote 11 A more proportional Canadian electoral system would level the political playing field between men and women. It would also lower the barriers to members of minority groups who seek election to the House of Commons. At present, SMP works against ethnic, linguistic and other minority groups which are not territorially concentrated. The party lists used in the MMP system would give members of these groups a better chance to win election to the House of Commons. A similar argument has been made for Aboriginal Canadians, whose territorial concentration particularly in northern ridings has not yet sufficed to guarantee them a fair share of Commons seats.Footnote 12

Finally, any discussion of alternative electoral systems must consider public and judicial opinion, and the practicality of various systems in a given setting. Electoral reform is not a burning issue in Canada, but there is evidence of public disenchantment with SMP. A 1990 national survey found that 42 percent of those who expressed an opinion accepted the distorting effects of SMP, while 75 percent preferred the principle of MMP (casting one ballot for a local candidate and a second ballot for a provincial party list).Footnote 13 While the Supreme Court of Canada has not yet assessed the constitutionality of SMP, its 1991 ruling on electoral boundaries suggests that SMP might fail a court challenge. Madam Justice McLachlin, writing for the majority, held that "Each citizen is entitled to be represented in government. Representation comprehends the idea of having a voice in the deliberations of the government as well as the idea of the right to bring one's grievances and concerns to the attention of one's government representative. ..."Footnote 14 SMP denies a majority of citizens the "effective representation" described in this passage, although it does give each citizen a single "government representative." A more proportional "mixed" system, which wastes fewer votes while preserving single-member constituencies, could guarantee the Charter right to "effective representation" as defined by the Court.

In the same Supreme Court judgment, the justices held that any alternative electoral system must be practical in a country with vast, underpopulated regions.Footnote 15 Any proposed electoral system which required the creation of multi-member constituencies in sparsely populated areas, where some ridings are already larger than most European countries, might not survive a court challenge under s. 3 of the Charter. Because the district magnitude largely determines the proportionality of a given STV or list-PR system and a minimum of four seats per constituency is required for a tolerably proportional outcome such a system would be impractical in Canada. If we wish to enhance proportionality, we must do so without sacrificing single-member constituencies, at least in northern and rural areas.

While the effects of any electoral reform cannot be predicted with certainty, and one cannot apply the experience of one country to another, the literature on electoral systems offers some firm conclusions about the relative benefits of the systems listed in the introduction to this paper. STV is an unacceptable alternative, for four reasons. First, it requires multi-member constituencies. Second, it forces candidates from the same party to compete against each other directly. This can divide local party organizations, and it can also foster serious and persistent factional divisions in national parties. Third, the way in which surplus votes are redistributed can have a powerful and arbitrary impact on the outcome. Fourth, it can take days or weeks to count the votes and elect a government.

SNTV is equally unacceptable, both because it would require multi-member constituencies and because it would not guarantee sufficient proportionality.Footnote 16 List-PR systems, which also require multi-member constituencies, could deprive local party associations of the right to nominate candidates, depending on who prepares the lists. However, list-PR does have an excellent record of electing women and minorities.

Canada should adopt some form of MMP. Public and judicial opinion already appear to support the principle of a "mixed" system, which offers the advantages of SMP and list-PR while mitigating their disadvantages. Under top-up MMP, which is presently used in Germany and New Zealand, each party nominates two groups of candidates. The first group runs in SMP constituencies, just as candidates for the Canadian House of Commons do now. The second group competes on party lists, either national or regional. When the ballots are counted, each party's share of parliamentary seats is calculated on the basis of its "second-ballot" vote. The second-ballot results are used to correct the disproportionality of the first-ballot results. This is the most proportional of the "mixed" systems.

Under top-up MMP, some party caucuses include substantially more "list" MPs than others. This need not be a problem, if all MPs are given substantial and clearly defined responsibilities.Footnote 17 SMP creates high turnover in the Canadian House of Commons, which is weakened by inexperienced members who lack policy expertise and a working knowledge of the rules. A more stable membership would make the Commons a more powerful legislative body,Footnote 18 as would a "class" of MPs with the time and energy for serious committee work, departmental oversight and legislative review. The real obstacle to adopting top-up MMP is the extreme proportionality of the seat allocations. Defenders of SMP argue that an excessively proportional electoral system would rule out single-party majority governments, and require the formation of "unstable coalitions." There are several flaws in this argument, not the least of which is the stability of many coalitions, but it must be taken seriously by those who would introduce a more proportional electoral system in Canada.

The second type of mixed system is parallel MMP. The crucial difference between top-up and parallel MMP is that in the latter system the seat total for each party is the sum of the seats won on both ballots (instead of being determined by the second-ballot result, as in top-up MMP). The seats won on the second ballot are added to those won on the first ballot; they are not used to correct the disproportional first-ballot results. Parallel MMP is therefore substantially less proportional than top-up MMP. While this may seem like a failing, it does answer those critics who prefer single-party majority governments. By favouring the largest party nationally, parallel MMP permits the election of majority governments with a large minority of the popular vote probably between 43 and 49 percent, as Table 1 suggests. It also gives every voter a chance to elect a second-ballot MP, reducing the problem of wasted votes and thereby boosting voter turnout, and it gives all of the major national parties a good chance to win seats in every region. It also benefits women and minority candidates.

Table 1

Approximate Seat Distributions Under SMP and Parallel MMP, 1984-1993
Party 1984
Lib. 40 54 83 87 177* 157*
P.C. 211* 189* 169* 151* 2 24
N.D.P 30 37 43 52 9 12
Ref. - - - - 52 52
B.Q. - - - - 54 47
Other 1 2 0 5** 13  
Total 282 282 295 295 295 295

* Indicates a single-party majority government.

** Includes the Reform vote.

Table 2

Party Seat Totals Under SMP and Three Variants of MMP
Party Parallel MMP/Provincial Lists Parallel MMP/National List Top-up MMP SMP
Lib. 139 138 116 155
Ref. 60 60 58 60
P.C. 36 36 57 20
B.Q. 38 39 32 44
N.D.P. 28 27 33 21
Other 0 1 5 1
Total 301 301 301 301

Note: The above tables use actual voting data from Canadian federal elections for illustrative purposes; they should not be taken as projections of actual outcomes, because the voting pattern would almost certainly have differed under an alternative electoral system.

To gauge the relative proportionality of the two MMP systems and SMP, Table 2 compares the seat allocations under the three systems. While the Liberals are always the winners, their share of Commons seats varies from an artificial majority of 51.5 percent (SMP) to an extremely proportional 38.5 percent (top-up MMP). The two parallel MMP systems award the Liberals around 45 percent of the seats not enough for a single-party majority government, but a comfortable plurality over the second- and third-place parties (Reform and the Progressive Conservatives, with 20 and 12 percent of the seats respectively).

While the effect of any electoral reform should not be overstated, and cannot be fully predicted, parallel MMP has enough potential to warrant serious official investigation and public discussion. The federal government should establish a commission of inquiry into the electoral system, with a mandate to recommend an alternative to SMP. A binding decision between the two systems should be left to a national referendum, preceded by an impartial campaign of public education about the issues involved in the choice. In recent years, several other established democracies including New Zealand, Japan and Italy have adopted variants of MMP in response to the failings of their existing electoral systems. The conditions are right for a similar leap of faith in Canada.


Abbreviation Full name Description
SMP Single-member plurality The country is divided into single-member constituencies. The voter chooses one of the candidates on the ballot. The candidate with more votes than any other wins the seat.
STV Single transferable vote The country is divided into multi-member constituencies. The voter ranks some or all of the candidates on the ballot. A Hare or Droop quota is calculated on the basis of the valid votes cast, and the first-preference votes are counted. Any candidate with more first preferences than the quota is declared elected, and his or her surplus votes are redistrib- uted among the remaining candidates on the basis of second preferences. The counting continues, eliminating the lowest candidates and redistributing the surplus votes of the winners, until all the seats have been filled.
SNTV Single non-transferable vote As STV, except that each voter casts only one vote (instead of a maximum of one per available seat).
List-PR List-proportional representation The country is divided into multi-member constituencies (or is designated as a single multi-member constituency). Each of the parties lists its candidates on the ballot, and the voter chooses one of those lists. The lists can be open (the voter can express a preference among the candidates of her favourite party) or closed (a simple vote for a party). The seats are allocated on the basis of either a highest- average or a largest-remainder formula. When the number of seats for each party has been determined, they are filled by the candidates at the top of the lists.
MMP Mixed-member plurality Each voter casts two ballots: one for a candidate in an SMP constituency and one for a party list. In top-up MMP, the number of seats for each party is calculated on the basis of the list-PR votes and the number of SMP seats is subtracted to determine the number of list-PR seats to which each party is entitled. In parallel MMP the two groups of MPs are elected separately, and their combined totals determine each party's seat allocation


Footnote 1 Chief Electoral Officer of Canada, Thirty-Sixth General Election, 1997: Official Voting Results Synopsis (Ottawa: Elections Canada, 1997), Table 4.

Footnote 2 Jackman, Robert W., "Political Institutions and Voter Turnout in the Industrial Democracies", American Political Science Review 81:2 (1987), 405-423.

Footnote 3 Alan Frizzell and Jon H. Pammett, eds., The Canadian General Election of 1997 (Ottawa: Carleton University Press, 1998), 254-284.

Footnote 4 Blais, André and R. Kenneth Carty, "Does Proportional Representation Foster Voter Turnout?", European Journal of Political Research 18 (1990), 179.

Footnote 5 Blais, André and R. Kenneth Carty, "The Psychological Impact of Electoral Laws: Measuring Duverger's Elusive Factor", British Journal of Political Science 21 (1991), 91.

Footnote 6 Farrell, David M., Comparing Electoral Systems (London: Prentice-Hall/Harvester Wheatsheaf, 1997), 146-147.

Footnote 7 Cairns, Alan C., "The Electoral System and the Party System in Canada, 1921-1965", in Douglas E. Williams, ed., Constitution, Government, and Society in Canada (Toronto: McClelland and Stewart, 1988), 114-115.

Footnote8 Studlar, Donley, "Will Canada Seriously Consider Electoral System Reform? Women and Aboriginals Should", in Milner, Henry, ed., Making Every Vote Count: Canada's Electoral System Reassessed (Peterborough: Broadview, 1999), 129.

Footnote 9 Arseneau, Thérèse, "Electing Representative Legislatures: Lessons From New Zealand", in Milner, ed., Making Every Vote Count, 144.

Footnote 10 MacIvor, Heather, Women and Politics in Canada (Peterborough: Broadview, 1996), Chapters 7-9.

Footnote 11 Rule, Wilma, "Electoral Systems, Contextual Factors and Women's Opportunity for Election to Parliament in Twenty-Three Democracies", Western Political Quarterly 40:3 (1987), 494-495; Norris, Pippa, "Conclusions: Comparing Passages to Power", in Norris, Pippa, ed., Passages to Power: Legislative Recruitment in Advanced Democracies (Cambridge: Cambridge University Press, 1997), 214-215.

Footnote 12 Studlar, "Will Canada Seriously Consider Electoral System Reform?", 130.

Footnote 13 Blais, André and Elisabeth Gidengil, Making Representative Democracy Work: The Views of Canadians, volume 17 of the research studies for the Royal Commission on Electoral Reform and Party Financing (Toronto: Dundurn, 1991), 55.

Footnote 14 Reference re Provincial Boundaries (Saskatchewan), 1991 2 S.C.R.

Footnote 15 Reference re Provincial Boundaries (Saskatchewan), 1991 2 S.C.R.

Footnote 16 Lijphart, Arend, Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies, 1945-1990 (Oxford: Oxford University Press, 1994), 41.

Footnote 17 Jenkins, Lord, et al. (1998), The Report of the Independent Commission on the Voting System (London: UK Home Office), Chapter 7.

Footnote 18 Franks, C.E.S., The Parliament of Canada (Toronto: University of Toronto Press, 1987), 73-79.

Note: The opinions expressed are those of the authors; they do not necessarily reflect those of the Chief Electoral Officer of Canada.