In April 1998, Policy Options carried an article in which I argued that Canada’s electoral system could be greatly improved by the addition of run-off elections, to ensure that Parliament would only consist of MPs who had actually won a majority of the vote. If adopted as proposed, runoff elections would be held about one week after the first election in all constituencies where no candidate had received a majority of the vote, between the two candidates with the most votes in the first ballot. This article briefly revisits that argument in light of the results of the federal general election in 2000.
The extent to which MPs are elected with a minority of the vote under the present electoral system can be gauged from the results of the federal general elections in 1997 and 2000. In 1997, only 34.5 per cent of the newly elected MPs had a majority of the vote; 50.2 per cent got between 40 and 50 per cent; and 15.3 per cent received less than 40 per cent. In 2000, the comparable distribution was 51.5 per cent, 40.5 per cent and 8.0 per cent; i.e., in nearly half of the federal constituencies, most voters did not support their present MP.
Although these outcomes do not fit comfortably in a system in which majority-rule is the norm, proposals to change the electoral system have had little success. That is probably attributable, at least in part, to concerns about the consequences of proposed reforms, which are often difficult to predict and potentially far-reaching. Those concerns could apply to run-off elections as well, because of their uncertain implications for voter preferences, party strategies and coalition-building. But the risks seem far less than under more radical reforms.
To assess the risks of run-off elections in a limited way, my 1998 proposal presented the results of a “what-if” analysis, aimed at estimating how the addition of run-off elections might have affected the results of the 1997 federal general election. The same analysis has now been done for the federal general election of 2000.
For both elections, the analysis assumed that: (a) runoff elections would have been held in all of the constituencies where the first ballot had given none of the candidates a majority of the vote; (b) all candidates with more than 40 per cent of the vote in the first ballot and more than a 10 percentage-point lead over his/her nearest competitor would have ended up with a majority in the run-off election; and (c) half (randomly selected) of all the other candidates with a plurality in the first ballot would have won a majority in the run-off election, while the other half would have been defeated by the runner-up candidate in the first ballot.
In that highly simplified though reasonably credible scenario, run-off elections would have changed the outcome of both elections in only minor ways. The resulting distribution of power would have been about the same, and only 38 and 36 constituencies would have ended up with a different MP, in 1997 and 2000 respectively (see Table 1).
Run-off elections represent only one of many suggested remedies for the drawbacks of the present electoral system. Unfortunately, the advocates of other electoral reforms tend to be unduly single-minded in their pursuit of an ultra-precise recording of popular voting preferences, and not to worry too much about the resulting increase in the complexity of the voting process, and/or the reduced understandability/transparency of the results (which would then tend to emerge from a black box, following calculations that are hard to explain publicly); and/or the loss of a direct, personal connection between MPs and their constituents; and/or a resulting propensity for minority—or coalition— government, with the associated dispersion of political power and enhancement in the influence and veto-capacity of well-organized single interest groups.
As is implied by these objections to other suggested electoral reforms, the criteria for evaluating electoral systems should not be limited to their fairness and accuracy in translating votes into seats. Equally important are the implications of alternative electoral systems for the transparency of the process, the nature of the connection between voters and legislators, the government’s ability to make decisions and the clarity of political accountability.
From that broader perspective, the best remedy for the main weakness of the present electoral system would be to add run-off elections. Apart from their main benefit—ensuring that everyone in Parliament had won a majority— they would also make it more likely that governments with a majority of the seats would have won a majority of the national vote as well.