Although the analyses have identified several key characteristics that can potentially explain the greater propensity of younger Canadians and visible minority Canadians to be habitual non-voters, we have not yet evaluated precisely how much each characteristic can help explain the observed gaps in habitual non-voting between these groups. To answer this question, we performed simulations. We begin with younger Canadians.
Based on the multivariate analyses presented in Table 8, we can predict what would be the level of habitual non-voting among younger Canadians if they had the same profile as older Canadians on a key set of characteristics. The simulation is performed for one variable at a time to assess how much each variable can explain the gap in voter turnout. All other variables are kept at the sample mean. The simulations are conducted only for Canadians aged 18-24. Table 10 presents the results of these simulations. For each simulation, the table reports how much the gap in habitual non-voting would shrink should younger Canadians share the same attribute as older Canadians. The capacity of each variable to explain the gap in habitual non-voting between younger and older Canadians is dependent on how strongly the variable correlates with habitual non-voting and how large the difference is between younger and older Canadians in holding this attribute. For instance, if feeling close to a party strongly correlates with habitual non-voting for younger Canadians and younger Canadians are substantively less likely to feel close to a party, then the variable will explain a significant gap in habitual non-voting.
|Capacity to explain gap between 18-24 and 35 and older)||Capacity to explain gap between visible minorities and other Canadians|
|Being a recent immigrants||-4||-3|
|Being a student||-7||n.s.1|
|Closeness to political party||-3||-1|
|Feeling guilty if not voting||-7||-2|
|Confidence in Elections Canada||-4||n.s.1|
|Gap observed in descriptive data – see Table 7||37||15|
|Gap potentially explained||25||10|
Capacity to explain the gap: Explained gaps in habitual non-voting based on predicted probabilities derived from analyses in Tables 8 and 9. Keeping everything else at the sample means, Canadians aged 18-24 were attributed the score of Canadians aged 35 and older (visible minority Canadians were attributed score of other Canadians) for each of the variables listed in the Table. Simulations were performed one at a time for each variable.
1. Variable was not statistically significant for younger Canadians (Table 8) or visible minority Canadians (Table 9) and therefore cannot explain difference in habitual non-voting with other Canadians.
The gap in habitual non-voting observed between the two groups was 37 points (see Table 7). Our simulations indicate that if younger and older Canadians shared the same attributes in terms of the likelihood of being a recent immigrant, being a student, feelings of guilt when not voting, closeness to a political party and confidence in Elections Canada, the gap in habitual non-voting would shrink by about 25 points. This is approximately two thirds of the observed gap in habitual non-voting. The two variables with the greatest potential to explain the gap are the feeling of guilt when not voting and being a student.
First, if younger Canadians were as likely as older ones to feel guilty when not voting, the propensity to be habitual non-voters would decrease by seven points. The importance of trying to create a sense of civic duty, via civic education programs or otherwise, is thus potentially critical.
Second, if the proportion of 18 to 24 year-olds who are students were the same as that among the 35 and older, the proportion of habitual non-voters would shrink by another 7 points. The large effect for the student variable is caused by the large gap in the proportion of 18 to 24 and 35 and older who are students in the sample (49% vs. 1%). Should this finding be replicated by other studies, it would suggest that the challenge of voter turnout among the Canadian youth is not about to disappear as the share of the population staying longer in school is increasing. On a more optimistic note, however, it might suggest that the problem is more temporary and that once finished with school or university, Canadians might start to vote more often; but that remains to be verified.
Third, also important are the different levels of confidence in Elections Canada that could explain about 4 points in the gap in habitual non-voting between younger and older Canadians. The perceived quality of the fairness and transparency of the electoral process and of those in charge of that process is thus potentially critical in maintaining the desire of people to participate in the electoral process.
Fourth, the magnitude of the effect is also of 4 points for being a recent immigrant. The size of the effect for this characteristic is impressive considering that the proportion of younger Canadians who are recent immigrants is not substantively larger than that for older Canadians (8% vs. 3% – see Table 2). This indicates how substantively greater habitual non-voting is among recent immigrants.
Finally, younger Canadians' weaker relationship with political parties also partly explains their greater propensity to be habitual non-voters. Should younger Canadians be as likely as older ones to feel close to a political party, their likelihood of being habitual non-voter would decrease by three percentage points. The solution to youth disengagement with electoral politics thus appears to be in part in the hands of parties who need to find ways to reconnect with the Canadian youth.
The story is somewhat different for visible minority Canadians. First, the gap in habitual non-voting with other Canadians is not as large as that observed between younger and older Canadians (15 points vs. 37 points). Second, being a student here cannot be part of the explanation because this was not a factor associated with habitual non-voting among visible minority Canadians. The same holds for confidence in Elections Canada; although it was a significant correlate of habitual non-voting among the general Canadian population, it was not among visible minority Canadians. In any case, visible minority Canadians expressed greater confidence in Elections Canada than other Canadians, not lower. Third, although some of the correlates of voting are the same as for younger Canadians, the capacity to explain the gap in habitual non-voting between visible minority Canadians and other Canadians is limited because the two groups are essentially as likely to hold these attitudes. This holds for feeling guilty when not voting and the perception that one's vote can make a difference (see Table 3).
To understand the greater propensity to be habitual non-voters among visible minority Canadians, we need to turn our attention toward their socio-economic status, more specifically the fact that a large proportion of them are recent immigrants and that they are younger than the rest of the population.
First, as we observed for our analyses of younger Canadians, one of the reasons why visible minority Canadians are more likely to be habitual non-voters is that a larger proportion of them are recent immigrants (24% vs. 2%), and recent immigrants are more likely to be habitual non-voters. Intriguingly, the relationship is not as strong as that observed among younger Canadians. Although the reason for this weak relationship is not clear, the consequences on its capacity to explain the gap in habitual non-voting are clear. Despite a large difference in sample composition on this characteristic, it only explains the gap in habitual non-voting by 3 points.
Second, visible minority Canadians are younger than the rest of the Canadian population. In our sample, the average age of visible minority Canadians 38 years while for the rest of the population it is 50. Controlling for all other factors included in the analyses in Table 9, should visible minority Canadians be the same age as the rest of the Canadian population (50 instead of 38), their propensity to be habitual non-voters would decrease by 4 points.
Of course, our analyses do not provide a complete explanation as to why younger Canadians and visible minority Canadians are more likely to be habitual non-voters. The analyses allow us to explain only part of the gap. The share of that gap that is explained is nevertheless substantial, about two-thirds for each group. Moreover, the simulations point to different stories for each group. While the explanations for younger Canadians relate both to their socio-economic status and political attitudes, the explanations for visible minority Canadians relate only to their socio-economic status. Our investigations thus reveal two groups of Canadians who are less likely to vote but who abstain from voting for different reasons. On both cases, a better understanding of the challenges for voting among recent immigrants is needed. In the specific case of younger Canadians, however, the challenge is greater as we need a broader understanding of the roots of their disengagement with politics in terms of their perception that voting is a duty, their disconnect with parties, or even their lower level of confidence in Elections Canada.