What is the story?
Ascertaining the impact of electoral systems is a tricky business. On the one hand, the long-rooted observational approach consisting in comparing electoral outcomes across countries with different electoral systems has been unable to produce unbiased results because of almost infinite numbers of factors intervening in the explanation of this outcome. On the other hand, although promising, recent experimentations on voting behavior of subjects in simulated election in a lab have been unable to capture real-life complexity.
To address these problems, a ground-breaking quasi-experimental study has been developed by the Making Electoral Democracy Work project. Shortly before the 2011 Ontario election, a Website called ThreeOntarioVotes was launched. The goal was to collect data on the impact of three electoral systems: first past the post (FPTP), the alternative vote (AV), and proportional representation (PR). This was done by inviting Web users to indicate their constituency and to vote under each of the three electoral systems. In doing so, we were able to control for a whole range of intervening factors while keeping the study in a real-life framing. This study leads to 2 major findings.
Finding 1: The impact of electoral rules on votes’ fragmentation
As reported in Table 3, people tend to adapt their voting behavior by considering the potentiality of their vote being wasted under FPTP and AV. As expected, the vote distribution is more fragmented under PR than under FPTP. The fragmentation of the votes under AV is located somewhere in between these two.
Although conventional theories assumes that the incentives in terms of vote-wasting to be similar under FPTP and AV, the fragmentation of the vote distribution appears to be much greater under the latter.
Finding 2: The impact of electoral rules on party success
The distribution of seats obtained by each party considering the distribution of votes presented above has been computed for each electoral system. The FPTP/AV comparison is particularly interesting. As reported in Table 6, AV produces less strategic desertion of weak parties than FPTP. Also,
Compared to FPTP, AV advantages parties that are acceptable to many voters, which are likely to be centrist in most instances.
In the particular case of the 2011 Ontario election that party happened to be the Liberal party, which succeeded in getting more seats under AV than under FPTP in spite of the fact that it got fewer first votes.
For more details, see Blais, André, Héroux-Legault, Maxime, Stephenson, Laura, Cross, William, and Elisabeth Gidengil. 2012. “Assessing the Psychological and Mechanical Impact of Electoral Rules: A Quasi-Experiment.” Electoral Studies 31: 829-837.
