A Grain of Salt: It is Not Clear that Electoral Reform Can Improve Voter Turnout

A key part of the discussion of electoral reform has been voter turnout. One of the arguments often put forward in favour of electoral reform has been that proportional systems can increase voter turnout. The argument goes that when voters feel more certain that their votes will count towards electing a member of parliament, they will be more likely to vote. Examining election results across different countries, however, provides little evidence to support this argument. There are a number of countries with proportional systems that have higher levels of voter turnout than Canada, but there are also a number with lower turnout. Perhaps more strikingly, electoral reform in New Zealand did nothing to reverse that countries’ declining turnout.

If one looks at the most recent elections in countries (displayed in the table below) across North America and Europe, one finds a mix of evidence linking electoral systems to voter turnout. On one hand, the five countries with the highest levels of voter turnout; Belgium, Denmark, Sweden, Iceland, and Norway are all list proportional representation (PR) systems. On the other, some of the countries with the lowest levels of voter turnout; Switzerland, Poland, and Slovenia are also list PR systems. They are followed by two countries with quasi-proportional systems in Japan and Lithuania (both these countries include a proportional element in their electoral systems, but they are not fully proportional). In contrast, the three countries still using single member plurality (SMP- otherwise known as first past the post) find themselves in the middle of the table. Canada does the best of the SMP countries at 13th, but the worst of the SMP countries, the United States, is still ahead of a significant number of list PR systems. If one looks at the averages, as shown in the graph below, turnout in PR systems is slightly higher than in SMP ones, but only by 2-5 percentage points (depending on whether one compares SMP to list PR, mixed member proportional (MMP), or all PR systems).

Country

Electoral System

Turnout

Country

Electoral System

Turnout

Belgium

List PR

89.37

USA*

SMP

64.44

Denmark

List PR

85.89

Estonia

List PR

64.23

Sweden

List PR

85.81

Greece

MMP

63.60

Iceland

List PR

81.44

Slovakia

List PR

59.82

Norway

List PR

78.23

Czechia (Czech Republic)

List PR

59.48

New Zealand

MMP

76.95

Latvia

List PR

58.80

Italy

MMP

75.19

South Korea

Parallel

58.03

Austria

List PR

74.91

Serbia

List PR

56.07

Netherlands

List PR

74.56

Portugal

List PR

55.84

Israel

List PR

72.34

France

Two Round Run-Off

55.40

Germany

MMP

71.53

Lithuania

Parallel

52.93

Spain

List PR

69.84

Japan

Parallel

52.66

Canada

SMP

68.28

Slovenia

List PR

51.73

Finland

List PR

66.85

Poland

List PR

50.92

UK

SMP

66.12

Switzerland

List PR

48.40

Ireland

STV

65.09

average-voter-turnout

The low impact of electoral systems on turnout holds when the analysis is extended to look at all elections going back to 2000. Voter turnout is higher in proportional systems, particularly in MMP systems, but the difference tends not to be large. There is only a 5-6 point difference between SMP systems and List PR systems or all PR systems taken together. MMP systems do have levels of turnout that are a fair bit greater than SMP systems, but it is unclear why these systems would increase voter turnout in a way that list PR would not.  In the absence of a theory as to why voter turnout would be higher in MMP systems than other PR systems, it is likely that the particularly high turnout in these systems is a result of randomness.  While there is slightly higher voter turnout in PR systems, the differences between PR and SMP systems are not large.

average-going-back-to-2000

The lack of a large difference is troubling because of the extent to which there are confound variables other than electoral systems that can influence turnout. In Bowling Alone**, Robert Putnam points to declining social capital as a cause of declines in voter turnout in that country. Putnam makes the argument that people in countries with higher levels of social trust will feel a greater connection to the community, and therefore be more likely to participate in politics. Many of the list PR countries with particularly high levels of turnout are also countries with high levels of social capital such as Denmark, Norway, and Sweden. The fact that a significant number of the high turnout PR countries are also high social capital countries raises questions as to whether electoral systems or high social capital are causing the difference in levels turnout between PR and SMP countries. This, combined with the small differences in turnout between PR and SMP systems make it hard to say conclusively that PR systems lead to higher levels of voter turnout.

A comparison of New Zealand with Canada offers further evidence of the weakness of the effect that electoral systems have on voter turnout. New Zealand switched from Canada’s SMP electoral system to MMP for the 1996 election, making it a good test case of what happens to voter turnout when a country changes electoral systems. If MMP leads to increased voter turnout, one would expect to see turnout in New Zealand increase immediately after the change in electoral systems. Instead, the graph below shows that turnout in New Zealand continued to decline after electoral reform. It did so at a similar rate to the decline in a voter turnout in Canada, despite the fact that Canada has not seen electoral reform. There is a small, brief spike in voter turnout in the 1996 election, but by 1999 the country had returned to the declining trend in turnout that New Zealand experienced under SMP. The fact that New Zealand has a higher level of turnout that Canada has much more to do with the high levels of turnout the country saw throughout the post-war period than with its electoral reform. The difference in turnout between the two countries was as great when both used the same electoral system as it was after New Zealand changed to a proportional system.

canada-vs-new-zealand-turnout

There is little evidence that a change in a countries’ electoral system will do much to change voter turnout. The differences in turnout between countries with proportional systems and SMP systems is not all that large and could be explained by factors like social capital that have nothing to do with a countries’ electoral system. New Zealand’s experience with electoral reform further offers little evidence that a change to a more proportional system can stem declines in voter turnout. None of this is to argue that proportional systems do not have merit. There are certainly reasons to adopt proportional systems that have nothing to do with turnout. Expecting electoral system reform to increase turnout, may however, be expecting too much of electoral reform.

* American mid-term elections are excluded from this analysis because they do not coincide with the same large scale national campaigns that occur in most countries during a general election.

** Putnam, Robert. (2000). Bowling Alone: The Collapse and Revival of American Community. New York: Simon and Schuster.

*** Turnout data taken from: http://www.idea.int/vt/viewdata.cfm#

Standard