Thank you. As I know I have very little time, I'll try to be very quick and to the point.
In your deliberation about whether to reform the existing electoral system, you will have to address two questions: the first, what will be the likely consequences of a new system, and, second, are these consequences good or bad for the country?
As a scientist, I can address the first question, and this is what I will do in the next few minutes. I have personal views about the consequences, which ones are good and bad, but I believe that my main contribution should be to tell you what empirical research tells us about the consequences of voting systems.
I will tell you about four empirical studies that I have conducted with other colleagues, each dealing with potential consequences of voting systems. My challenge is to present four of my studies, which are all very complex, and rich, and so on, in 10 minutes. That is two minutes per study. I'll be sweet and short.
The studies about the consequences of proportional representation consist in a comparison of what we have observed in places with PR, proportional representation, and in places under non-proportional systems, which are sometimes called majoritarian. The differences that we observe can result from causes other than the voting system, and these studies attempt to take into account these other factors, and to control for them. However, we are never sure that we have taken into account all of the significant factors, and thus we are never absolutely certain about our conclusions. This will be taken into account.
Furthermore, these studies do not tell us about the specific consequences of specific forms of PR. Still, I would argue that the most important decision you have to make is whether to adopt some form of PR or not. It is thus important to look at what the international comparative evidence tells us, so hopefully you will find these studies helpful.
The research I present deals with the first two principles for electoral reform that have been established by the committee: one, effectiveness and legitimacy; and second, engagement.
The first study is about whether turnout tends to be higher under PR. A study published with Agnieszka Dobrzynska in the European Journal of Political Research deals with turnout in lower house elections, a total of 324 elections in 91 countries.
The dependent variable, what one can explain, is turnout. We consider about a dozen factors that could effect turnout: GDP per capita, illiteracy, population size, and so on. For the voting system, we compare PR with non-PR elections, and we also look at the degree of disproportionality of the voting system, the difference in vote and seat shares for the parties.
We estimate the independent impact of each factor, controlling for all the others. Our finding for PR turnout, everything else being equal, is that it is three percentage points higher under PR. This study suggests that the adoption of PR might slightly increase turnout.
The second study is about whether there is less strategic voting under PR. The study was conducted with Thomas Gschwend from the University of Mannheim, and it deals with strategic desertion, which is defined as not voting for one's preferred party or leader.
The data is from the Comparative Study of Electoral Systems, a series of academic election surveys conducted in 25 democracies. All of these studies include questions about how much each respondent likes or dislikes each of the parties and leaders and which party they voted for. In each survey, we determined how many respondents voted for a party or leader that is not their preferred one. The mean in all of these 25 elections is 22%.
We then compare the proportion of strategic defection in PR and non-PR elections—22% versus 21%—there was no difference. The correlation between defection and the degree of disproportionality is nil. Multivariate analysis confirmed the same result: there is no relationship between PR and strategic defection.
Our conclusion to this study indicates that the adoption of PR is unlikely to reduce strategic voting.
The third study is about whether citizens have more positive evaluations of democracy under PR. This was a study with Peter Loewen, who was a student in Montreal and is now a member at University of Toronto, published in a book by Oxford University Press. The data again is from CSES, a group of academic studies and surveys conducted by academics in 20 different elections across the world.
Again, the dependent variable to explain is basically attitudes about democracy. We have three kinds of attitudes. First is satisfaction with democracy. How satisfied are you with the way democracy works in your country? The second is perceptions of fairness. How fairly or unfairly was the election conducted? The third is perceptions of responsiveness, with a battery of three questions. How much do MPs know about ordinary people in your country; how much do parties care about ordinary people; and how much difference does it make who is in power?
The independent variables, the explanatory factors, include the degree of disproportionality, the degree of democracy, and human development.
The findings are that more proportional systems are clearly perceived to be fairer; they are perceived to be just a little bit more responsive; and people are not more satisfied overall under PR. This study suggests that if PR is adopted, elections are likely to be perceived to be fairer, but it is unlikely that people will be more satisfied overall.
The fourth study is about whether PR produces governments that better represent citizens' ideological orientations. It is a study by Marc André Bodet, a student at UDM at the time, and now at Laval University.
The variable to be explained is what we call “ideological congruence”, which is basically the absence of distance between citizens and government on a left-right ideological self-placement. The respondents have to locate themselves on a scale of 0 to 10, where zero is far left and 10 is far right. They can locate themselves wherever they want, and then they also locate each of the parties on that same scale. So we have an ideological placement for each of the respondents, and also the median perception of each of the parties, meaning where the parties are on that left-right scale.
We look at the distance between each citizen and the government. Of course, if you want representation, we hope that the distance will be as small as possible. The distance is what we try to explain. The explanatory factors are the degree of disproportionality, plus whether it's a new or old democracy.
The finding is that there is no more or less congruence overall under PR. PR does not produce greater or weaker correspondence between the voter and government ideological orientation. PR does not reduce the mean distance between citizens and government, but it does produce a parliament that better represents the diversity of ideological orientations. Similar results have been reported by a few other studies.
I have five conclusions from these four studies. First, the introduction of PR might slightly increase turnout; second, it would almost certainly enhance the correspondence between the distribution of ideological orientations in the electorate and in the House of Commons; third, it would almost certainly enhance voters' evaluations of the fairness of elections; fourth, it would almost certainly not reduce strategic voting; and fifth, it is very unlikely to make Canadians more satisfied overall.