Bill C-16 proposes a number of measures to increase the convenience of voting in Canada. These provisions are: increased advanced polling opportunities, the existence of two consecutive regular polling days rather than only one, and the possibility of voting on a Sunday. The combination of the three measures provided for in Bill C-16 would make Canada's electoral legislation very unique. Only Sweden offers a similar combination of convenience measures to encourage greater voter participation. The case of Italy seems to resemble that of Canada. A period of two consecutive days, Saturday and Sunday, was recently established in Italy for voting. However, advanced polling is not permitted in that country.
In the past, several studies considered whether administrative measures to increase the convenience of voting would have an impact on voter turnout. However, existing empirical research has yielded conflicting results. Certain studies indicate that measures such as increased advanced polling opportunities or the possibility of voting on a holiday can increase voter turnout; others observed no impact.
With regard to the existence of two regular polling days, certain studies report a negative impact, which seems hard to believe. Although these studies are rigorous, all have their limitations. That is why we have conducted a new empirical transnational study including elections held in all democratic countries between 1990 and 2001. Our sample included 151 elections in 61 democratic countries. Our objective was to examine the impact that administrative measures to increase the convenience of voting would have on overall voter turnout.
The findings of our analyses indicate that, in countries in which electoral legislation facilitates voting by allowing advanced polling, voter turnout is approximately four percentage points higher than in countries in which this option is not available. However, the possibility of voting on holidays and the availability of two consecutive regular polling days had no significant statistical impact.
We have three comments to make. First, the three measures we examined had a positive impact on voter turnout; however, the correlations were not that strong. Second, the availability of two consecutive regular polling days was found in only six elections in our sample, which may explain why it is so difficult to identify a significant statistical trend. Lastly, our findings suggest that, although all of these measures could have a positive impact on voter turnout, the real impact is limited.