open Secondary menu

Potential Impacts of Extended Advance Voting on Voter Turnout

The Canadian Evidence: Who Are Advance Voters?

In ascertaining the effects of expanded opportunities for advance voting, it is important to consider who votes in advance. Two types of variables can be used to distinguish advance voters from those who abstain and from those who choose to vote on election day. First, one can examine socio-demographic characteristics, such as age, gender and income. The comparative literature (Highton 1997, Oliver 1996, Patterson and Caldeira 1985, Karp and Banducci 2001, Neeley and Richardson 2001, Stein and Garcia-Monet 1997, Stein 1998) has demonstrated that these factors are often important predictors of the decision both to vote and to vote in advance.

Second, one can examine political characteristics, such as interest in politics and identification or contact with a political party. Just as with socio-demographic characteristics, the interaction of individuals with the political system has been shown to predict the propensity to vote in advance (Oliver 1996, Karp and Banducci 2001).

It is possible to compare the socio-demographic and political profile of advance voters with those of election day voters and of abstainers. The analyses presented below come from the 2006 Canadian Election Study (CES). The CES polled more than 4,000 Canadians over the course of the 2006 election campaign and re-interviewed more than 3,000 of them after the election. In the last 10 days of the election campaign, the CES asked respondents if they had voted in advance.Footnote 14 The analysis below is restricted to those respondents who were interviewed for the first time in the last 10 days of the campaign. We distinguish three groups: those who indicated in the campaign survey that they had voted in advance (advance voters), those who said in the post- election survey that they voted and had not told us that they had voted in advance in the campaign survey (election day voters), and those who answered in the post-election survey that they had not voted (abstainers).

By comparing those who voted in advance with those who voted on election day or did not vote at all, we can ascertain which socio-demographic and political variables are related to the decision to vote in advance. Table 3 shows that a mixture of factors distinguishes advance voters from those who abstain or vote on election day (for a description of the variables, see Appendix C).

  • As voters age, their probability of voting in advance rather than voting on election day or abstaining increases.
  • As political interest increases, the probability of a citizen voting in advance rather than voting on election day or abstaining increases.
  • Compared with those who do not identify with a party, those who do are more likely to vote in advance than abstain.
  • Compared with those who are not contacted by a political party, those who are contacted by a party are more likely to vote in advance rather than on election day.

The first result suggests that the huge turnout gap between the youngest and oldest generations is unlikely to be reduced by an extension of advanced voting, since it is the oldest citizens who are most prone to take advantage of such measures.Footnote 15 The findings concerning political interest and party identification indicate that, so far, the main consequence of easier voting in advance in Canada has been to facilitate things for those who are already integrated into the political process. Whether advance voting is made more available does not matter much for those who do not follow politics. In addition, the findings concerning the relationship between contact with a political party and voting in advance suggest (in line with the findings reported by Oliver in the U.S.) that parties have an important role to play in the mobilization of people to vote in advance. As opportunities for advance voting expand, this mobilizing function of parties could become crucial to realizing gains in participation.

The factors that do not predict voting in advance are as important as the factors that do. We note that those who live in urban settings are no more likely to vote in advance than those who live in rural settings. Moreover, gender, education and income do not distinguish advance voters from the citizens who choose to cast a ballot on election day.Footnote 16

Table 3: Socio-demographic and Political Predictors of Election Day Voting and Advance Voting
Independent variables Dependent variable:
Choice to abstain, vote on election day, or vote in advance
Regression coefficient (Error)
Comparison group is advance vote.
Contact with a party -0.58 (0.36)
Interest in politics -5.12*** (0.68)
Partisan identification -0.87** (0.33)
Age -0.06*** (0.01)
Female -0.02 (0.08)
Income -0.05 (0.06)
Education -0.05 (0.09)
Urban -0.03 (0.36)
Constant 6.75*** (1.00)
Election day vote
Contact with a party -0.43* (0.26)
Interest in politics -0.80 (0.49)
Partisan identification -0.24 (0.22)
Age -0.02*** (0.01)
Female 0.06 (0.05)
Income 0.05 (0.04)
Education -0.05 (0.07)
Urban -0.28 (0.24)
Constant 4.20*** (0.74)

Number of cases 1,023

* significant at 0.10 (two-tailed test)
** significant at 0.05 (two-tailed test)
*** significant at 0.01 (two-tailed test)

Note: Missing data are imputed through 10 imputations. Model is multinomial logit.

Footnote 14 The question asked, "Did you vote in the advance poll?"

Footnote 15 This does not mean that extended advance voting will necessarily have no effect on youth turnout. The implication is that it is unlikely to have a larger effect on youth than it does on older citizens.

Footnote 16 Education has an effect when the analysis includes only socio-demographic characteristics. Better educated electors are somewhat more prone to vote in advance, but this is only because they tend to be more interested in politics.