open Secondary menu

Electoral Insight – 2004 General Election

Electoral Insight – January 2005

Making Votes Count: How Well Did Our Electoral System Perform?

Lawrence LeDuc
Professor of Political Science, University of Toronto

Over the past decade, an active debate has evolved in Canada about the desirability of electoral reform. While the issue is not new to Canadian academic interest, it took on greater currency following the 1993 election, which produced a highly distorted and regionalized Parliament. Footnote 1 This distortion persisted to some extent in the two following elections, with the Liberals in 1997 forming a majority government with only 38.5% of the total vote and in 2000 with 41%. In all three instances, critics of our first-past-the-post (FPTP) electoral system called attention to some of the negative effects on Canadian democracy that seemed traceable to the electoral system regional distortion, wasted votes, under-representation of the diversity of Canadian society, and declining competitiveness among Canadian political parties, to mention only a few. Footnote 2 The electoral system may also be implicated in the declining voter turnout for recent elections, since Canadians may not be motivated to go to the polls if they feel that their votes do not really "count" under the current system. Footnote 3 In early 2004, the Law Commission of Canada tabled its report on electoral reform, in which it recommended that Canada replace the current electoral system with a mixed-member proportional (MMP) system similar to that used in Germany and New Zealand. Footnote 4 In making its recommendation, the Law Commission stated:

"... a growing number of Canadians are no longer satisfied with our current electoral system. It is becoming increasingly apparent that the current electoral system no longer responds to 21st century Canadian democratic values ... Many Canadians desire an electoral system that better reflects the society in which they live one that includes a broader diversity of ideas and is more representative of Canadian society." Footnote 5

The 2004 election results and the case for electoral reform

The recent federal election provides us with the opportunity to re-examine some of the elements of this debate with new evidence. On the surface, the result of the 2004 federal election was quite different from those of the previous three mentioned above. From the beginning, the 2004 contest was more competitive, the outcome remaining in doubt until the very end. The election produced a minority government, rather than a majority one the Liberals' 37% of the popular vote translated into only 135 of the 308 parliamentary seats. Regional distortion appeared less extreme than previously, as the new Conservative Party won 24 seats in Ontario (in contrast to 2 won by the Canadian Alliance in the 2000 election), and the Liberals improved their representation in British Columbia (8 seats, compared to 5 in the 2000 election). The question thus arises: has the result of this election weakened the case for electoral reform? Is it possible that our electoral system performed better than expected in the recent election, and that some of the pressures to reform it will begin to abate? In this article, I will examine the 2004 election results in light of the arguments that have been advanced in favour of electoral reform, applying some of the tests that have been used on previous election results in Canada and in other countries to measure the degree of representative distortion in the new Parliament in comparison with those of the recent past. Footnote 6

A common method for measuring the representativeness of an electoral system in terms of the political parties is to compare the proportion of seats won by each party with the proportion of the votes that it received. A party that received, for example, 20% of the votes but only 10% of the seats would be said to be under-represented by 10% according to this measure, while a party that obtained 25% of the seats with the same proportion of votes would be overrepresented by 5%. A summary measure created by totalling these discrepancies in the proportion of seats and votes is called the Index of Disproportionality. Footnote 7 A value of this Index close to zero indicates a perfect fit between seats and votes. Higher values suggest greater distortion of a party's representation in comparison with the intentions of voters.

Table 1
Disproportionality of Seats and Votes in Canadian Federal Elections: 1980-2004
  2004 2000 1997 1993 1988 1984 1980
Total seats 308 301 301 295 295 282 282
Liberal – % votes 37 41 38 41 32 28 44
Liberal – % seats 44 57 51 60 28 14 52
Liberal – Difference 7 16 13 19 -4 -14 8
Conservative – % votes 30            
Conservative – % seats 32            
Conservative – Difference 2            
Reform/Canadian Alliance – % votes   26 19 19      
Reform/Canadian Alliance – % seats   22 20 18      
Reform/Canadian Alliance – Difference   -4 1 -1      
Progressive Conservative – % votes   12 19 16 43 50 33
Progressive Conservative – % seats   4 7 1 58 75 37
Progressive Conservative – Difference   -8 -12 -15 15 25 4
New Democratic Party – % votes 16 9 11 7 20 19 20
New Democratic Party – % seats 6 4 7 2 15 11 11
New Democratic Party – Difference -10 -5 -4 -5 -5 -8 -9
Bloc Québécois – % votes 12 11 11 14      
Bloc Québécois – % seats 18 13 15 18      
Bloc Québécois – Difference 6 2 4 4      
Green Party – % votes 4            
Green Party – % seats 0            
Green Party – Difference -4            
Index of Disproportionality 15 18 17 22 12 24 11

Table 1 makes this calculation for five political parties that fielded candidates in the 2004 election, comparing these results with similar calculations for federal elections back to 1980. Overall, the electoral system did perform slightly better in 2004 than in the three previous elections, but many of the distortions that have been evident since 1993 continued to exist. Some have even worsened. On the positive side, Liberal overrepresentation was substantially less in 2004 than previously the difference in the Liberal share of seats over votes dropping to 7% from 16%. The new Conservative Party performed more or less in line with its share of the vote, obtaining 2% more seats than it deserved, compared with a modest degree of under-representation for one of its predecessor parties (Canadian Alliance) and a more serious problem of underrepresentation for the other (Progressive Conservative) in 2000. The negative aspects of the electoral system come through more clearly in the showing of the other parties. The New Democratic Party (N.D.P.) continues to be severely disadvantaged by the FPTP system: it won only 6% of the seats in this Parliament, although it obtained 16% of the vote. The Bloc Québécois, on the other hand, continues to be overrepresented, simply due to the regional concentration of its vote. The Green Party, in spite of fielding candidates in all 308 constituencies and obtaining 4% of the national vote, failed to win a seat. Notwithstanding a slightly better overall performance in 2004, there is little in these results to suggest that the problems associated with the FPTP electoral system in Canada have receded in any significant way.

The failure of the electoral system to accurately reflect the votes of Canadians becomes more evident when the 2004 results are examined by region (Table 2). Here, I report only the summary Index of Disproportionality, which is easily compared across regions, with a value calculated for previous elections. The national improvement in the performance of the electoral system in terms of seats and votes is entirely attributable to Ontario, where the new Conservatives succeeded in winning 23% of the province's seats with 32% of the vote, still a net under-representation of 9%. The N.D.P., likewise, continues to be seriously under-represented in Ontario (11%), a position even worse than its showing in 2000 (7%). It is only the partial correction of Liberal overrepresentation that accounts for the province's improved electoral performance in 2004. But the Liberals, nevertheless, continue to be overrepresented in Ontario by a substantial 26%, having won 71% of the province's seats with only 45% of its vote.

Table 2
Index of Disproportionality, by Region: 1980-2004*
  2004 2000 1997 1993 1988 1984 1980
Atlantic 25 18 9 39 16 25 17
Quebec 23 11 20 23 30 24 29
Ontario 26 47 48 40 18 13 24
West 28 22 25 20 19 20 21
Canada 15 18 17 22 12 24 11

* Regional calculations are for the same sets of parties in each election as shown in Table 1.

In all other parts of the country, the performance of the electoral system was actually worse in 2004 than in 2000, in large part because of the dominance of single parties in a particular region. In the West, the Conservatives are wildly overrepresented, as was the Canadian Alliance in the previous election and before that, the Reform Party. In 2004, the Conservatives won 74% of western seats with only 45% of the vote. In Quebec, the electoral system favours the Bloc Québécois, which won 72% of Quebec's seats in 2004 with 49% of the vote. In the Atlantic provinces, it was the Liberals that gained the advantage, obtaining 69% of the seats in that region with 44% of the vote in 2004 a net overrepresentation in Parliament of 25%. Nearly all of these patterns are continuations of the types of distortions found in the previous three elections, the effect being a Parliament that is much more regionalized than it should be, or than is healthy in a country in which representation of diverse interests is so important.

The consequences of electoral distortion

The consequences of these distortions caused by the FPTP electoral system in Canada are important ones. Minority voices are muted or, in some cases, shut out entirely. Parties that concentrate their votes in a single province or geographic area are rewarded, while those that appeal to voters nationally are systematically disadvantaged. Were the allocation of seats truly proportional to votes cast on a regional basis, the new Conservatives would have obtained 7 seats in Quebec and the N.D.P. 4, and both parties would be better able to portray themselves as truly national entities, which in the minds of voters they clearly are. On the same basis, the Liberals would have an additional 11 seats in the West and the N.D.P. another 9, thereby more accurately reflecting the actual votes cast by citizens in that region of the country. Instead, we have a Parliament in which, once again, the new Conservatives will claim a mandate to speak for the West, the Bloc for Quebec, and the Liberals for the rest of the country, even though none of these parties obtained a majority of the votes in the region where they dominate. In all three instances, the majority of voters who did not cast their ballots for the dominant party in a region will rightly tend to feel that their views are not properly represented. As Alan Cairns noted more than 35 years ago, the electoral system in Canada continues to magnify the regional and linguistic divisions of the country, rather than providing a means to address them. Footnote 8

Table 3
Index of Disproportionality for Selected Countries (most recent election)
Total seats 659 308 150 603 349
Index 20 15 13 5 2
Turnout (%) 59.4 60.9 95.4 79.1 80.1
Electoral system* FPTP FPTP AV MMP PR

* FPTP – First past the post
AV – Alternative vote
MMP – Mixed-member proportional
PR – Proportional representation

How does the Canadian electoral system fare by these measures in comparison with other countries with which we might draw comparisons? In Table 3, I compute indices of disproportionality for the most recent election in four other parliamentary democracies. This comparison suggests that Canada shares some of the characteristics of other FPTP electoral systems, with Britain on balance performing worse than Canada in terms of party representation and Australia only slightly better. Footnote 9 Countries such as Germany or Sweden that employ proportional representation for all or part of their parliamentary seats do much better, because the system itself guarantees a better fit between seats and votes in those two countries. Voters in such systems are far less likely to feel that their votes are "wasted", as tends to be the case in many constituencies in Canada and in Britain. Proportional systems also tend, in general, to have higher voter turnout, although this varies considerably from one case to another, since there are other factors that can affect voter turnout above and beyond the system of representation. Footnote 10

Some countries with proportional representation systems tend to have higher voter turnout than Canada does, although the system of representation is only one of the factors affecting participation.

An idea whose time has come

What might we conclude from this preliminary examination of the 2004 election results in the context of the ongoing electoral reform debate? It is clear that the issue of electoral reform will continue to be widely discussed in Canada in any event, since there is already considerable momentum behind the reform movement. Although the 2004 election was a closer contest, the outcome more uncertain, and the result a minority government, there is little in the outcome that might slow the movement towards reform of an electoral system that remains desperately in need of change. All of the problems of representation that have been evident in the last four elections continue to plague our political system, and to undermine the quality of our democracy. Electoral reform is an idea whose time has clearly come in Canada, as it did in New Zealand in the early 1990s. In the end, the outcome of the most recent election is more likely to advance the cause of reform than to hinder it.


Footnote 1 Among the earlier sources on this topic, see William P. Irvine, Does Canada Need a New Electoral System? (Kingston, Ontario: Institute of Intergovernmental Relations, Queen's University, 1979) and Alan C. Cairns, "The Electoral System and the Party System in Canada, 19211965," Canadian Journal of Political Science Vol. I, No. 1 (March 1968), pp. 5580.

Footnote 2 See, for example, Henry Milner, "The Case for Proportional Representation in Canada," and Donley Studlar, "Will Canada Seriously Consider Electoral System Reform? Women and Aboriginals Should," both in Henry Milner, ed., Making Every Vote Count: Reassessing Canada's Electoral System (Toronto: Broadview Press, 1999), pp. 3749 and pp. 123132. See also Nick Loenen, Citizenship and Democracy: a Case for Proportional Representation (Toronto: Dundurn Press, 1997).

Footnote 3 On the turnout decline in recent elections, see André Blais, Elisabeth Gidengil, Richard Nadeau and Neil Nevitte, Anatomy of a Liberal Victory: Making Sense of the Vote in the 2000 Canadian Election (Peterborough, Ontario: Broadview Press, 2002), pp. 4563; and Jon H. Pammett, "The People's Verdict," in Jon H. Pammett and Christopher Dornan, eds., The Canadian General Election of 2000, Chapter 13 (Toronto: The Dundurn Group, 2001), pp. 293317.

Footnote 4 New Zealand reformed its electoral system in 1993, following an extended national debate and two referendums. On the performance of the MMP electoral system in New Zealand, see Jack Vowles, "Introducing Proportional Representation: the New Zealand Experience," Parliamentary Affairs Vol. 53 (2000), pp. 680696.

Footnote 5 Law Commission of Canada, Voting Counts: Electoral Reform for Canada (Ottawa: Department of Public Works and Government Services, 2004), p. 172.

Footnote 6 For further discussion of some of the issues surrounding electoral reform in Canada, including analyses of the 1993 and 1997 elections and comparisons with other countries, see Milner, Making Every Vote Count. See also Leslie Seidle, "The Canadian Electoral System and Proposals for Reform," in A. Brian Tanguay and Alain-G. Gagnon, eds., Canadian Parties in Transition, 2nd ed. (Toronto: Nelson Canada, 1996), pp. 282306.

Footnote 7 The Index is created by summing the absolute differences between the proportion of seats and votes obtained by the parties and dividing by two. The calculation here is for the five largest parties (in total votes obtained) for each election, and figures shown in the tables are rounded to the nearest whole number. See Michael Gallagher, "Proportionality, Disproportionality and Electoral Systems," Electoral Studies Vol. 10, No. 1 (March 1991), pp. 3351.

Footnote 8 Cairns, "The Electoral System and the Party System in Canada, 19211965," p. 66.

Footnote 9 Although Australia also has single-member districts, it permits voters to order the candidates on the ballot according to preference and then counts these "second choice" votes to determine a winner in constituencies where no candidate has obtained a majority of the votes cast. Thus, although minorities are not necessarily any better represented in Australia than in Canada, the two systems are not really comparable in terms of the effectiveness of individual votes.

Footnote 10 Australia has compulsory voting, accounting for the higher turnout in that case. On the issue of turnout and electoral systems more generally, see Mark N. Franklin, "The Dynamics of Electoral Participation," in Lawrence LeDuc, Richard G. Niemi and Pippa Norris, eds., Comparing Democracies 2: New Challenges in the Study of Elections and Voting (London: Sage Publications, 2002), pp. 148168.


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