Tuesday, 3 May 2011

Election 2011: Who's missing?

Much has been made of the dramatic vote swings witnessed in Monday’s historic federal elections, and party leaders and pundits have spent hours interpreting the results in search of what Canadians “really said” on Election Day.  Unfortunately, we’ve missed the main message by focussing on the wrong sets of numbers.

Let’s start by looking at what we usually do:  seats and votes.  According to these statistics, the “blue wave” brought Stephen Harper’s Conservative Party of Canada 24 additional seats and its long-awaited majority, while the “orange crush” saw Jack Layton and the New Democrats vault into second place, earning 65 more seats than they had at dissolution.  These victories came at the expense of the Liberals and Bloc Quebecois (both down 43 seats), and overshadowed the Green Party’s first-ever victory with the election of leader Elizabeth May. 

Beyond the seat totals, most observers point to the popular vote earned by the various parties as an indication of the outcome.  With the support of 39.6 percent of voters, the Conservatives now form one of the most “manufactured” majorities in Canadian history – second only to the Chretien Liberals’ 1997 victory (when they won 155 of 301 seats based on 38.5 percent of the popular vote).  These results have renewed calls for electoral reform, with critics claiming that the opposition parties were punished – and the Conservatives rewarded – by our first-past-the-post electoral system.  (One should note, however, that the New Democrats actually benefitted from the FPTP system, too.  They received 33 percent of the seats on the strength of 30.6 percent of the vote.)

While they await final verification by Elections Canada, these numbers do accurately reflect the mainstream view of the election’s outcome.  One important statistic is missing, however:  voter turnout.

Preliminary figures released by Elections Canada estimate Monday’s turnout at 61.4 percent.  This rate was up slightly from 2008, when participation reached an all-time low of 58.8 percent.  But it was a far cry from the predictions offered by many pundits and politicians, and remains the third-lowest turnout on record.  

This means that 38.6 percent of eligible Canadians chose not to participate in Monday’s election, yet their role in the outcome receives no attention in traditional reports of the results.  This is puzzling considering non-voters make up the largest segment of the Canadian electorate – larger than voters supporting the Conservative government, and larger than all of the opposition parties combined.  This is not the first time, of course.  Non-voters made up the largest segment of the electorate in 2004, as well.  In terms of provincial election results, the same is presently true across most of the country (save for Saskatchewan, Newfoundland and Labrador, and pei).

Some may see little reason for concern in these figures.  After all, recent research indicates that there is very little distinguishing the political views of voters and non-voters.  These studies offer little reason to believe that the outcomes of Canadian elections would have differed drastically, had more citizens turned out to vote.  Moreover, provided non-voting is evenly distributed across the entire population, and provided Canadians appear satisfied with the status quo, long-term decreases may be an indication of the health of Canadian democracy.  Canadians may be content that their vote will not make much of a difference, given that a change in government is unlikely to result in a change to their everyday lives; or they may feel comfortable allowing someone else to make the necessary decision on Election Day.  In this sense, many non-voters may be considered “couch” supporters of the governing party (who see little reason to turnout, given that their party of choice will win the election, regardless of their vote); “closet” supporters of the opposition (who see only futility in casting a ballot for a losing party); or “political drop-outs” (who are so ill-informed that their participation in the election could only serve to distort the outcome).

By contrast, others are likely to be concerned with low voter turnout.   That our first-past-the-post electoral system “manufactures majorities” is worrisome enough, they claim; the fact that legislative majorities and mandates are being based on the expressed support of so few eligible voters is even more disconcerting.  Other critics may point to several surveys at the federal level that challenge the assumption that non-voters are drawn proportionately from all segments of the population.  Studies consistently report an under-representation of youth, Aboriginals, recent immigrants, members of visible minority communities, people with lower levels of education, and lower-income individuals among Canadian voters.  This “participatory distortion” suggests that only certain types of people are involved in the electoral process.   This may establish a situation in which active citizens are over-represented in terms of their influence on public policy, yet unrepresentative of the general population with respect to some politically relevant characteristic(s).  Because they are more likely to vote (or otherwise engage in politics), public officials are more likely to hear, respond to, and behave proactively in the interests of the most active citizens and groups.  By the same token, parties have fewer incentives to engage with those people and communities who are least likely to participate.  These distortions may create a vicious cycle, in that marginalized groups may feel even less inclined to vote, thus further distancing themselves from the electoral process. 

Without further research, it is impossible to decide between these two normative evaluations of the 2011 federal election.  Previous studies appear to support both causes.  Beyond doubt, however, it is time we start reconsidering our use of “popular vote” totals to depict the outcome of Canadian elections.  Non-voters should be counted in our calculations.

1 comment:

  1. Hi Jared,

    Interesting as always. Seems to me you want to do two things. One, highlight that there is low voter turnout. Second, highlight that a government somehow lacks legitimacy because so few people actually voted for them. We can leave aside disagreements over whether non-voters differ in their aggregate preferences from voters (though I'll note that the only systematic, peer-reviewed evidence in Canada suggests they do not).

    Of course, we can highlight that there is low voter turnout merely by noting that there is low voter turnout. This, however, is not done by reporting some product of vote share and turnout. Instead, this brings less clarity to whether participation is lower. For example, if I tell you that in one election Jean Chretien got 25.75% according to your proposed measure, and in another he received 25.04%. In which election was turnout higher? And in which did Chretien receive more support among those who voted. Unfortunately, you can't really answer either of those questions if you so wished. You need one more piece of information: either turnout or the vote share. So, we're back to reporting both.

    I should think that a basic rule of reporting data is that it's better to report the component parts of something rather than the whole, if those two parts have independent value. In this case, both do. So, I think it's hardly beyond doubt that should reconsider popular vote.