WHY ISSUES AREN’T THAT IMPORTANT IN ELECTIONS
Parties’ policies do matter but mostly because they help define the leaders
Are campaigns really a poor time to talk about issues? The day-today political coverage of policy announcements would have us believe that issues matter a lot for vote choice, perhaps more than other considerations like the direct evaluations of leaders and long-term political affiliations. After all, why else would leaders invest such time in talking about them and why would the media dedicate so many column inches? The systematic evidence suggests something diferent, however.
In my last column, I attempted a sketch of the ideology of the major party leaders. On the part of Stephen Harper, there is a belief in not only a limited role but also a limited capacity of government. What results from this are targeted policies aimed at changing the behaviour of citizens at the margin. These policies are tied together with a common narrative about the need to support families, rather than just individual citizens. For Justin Trudeau, there is an almost classically liberal belief in markets corrected by government redistribution. He is no obvious statist. For Thomas Mulcair, there is less clarity. This was not aided this week by the revelation of his 14-year-old defence of the liberalism of Margaret Thatcher. There’s good reason to believe Mr. Mulcair is more of a social democrat than either of his competitors, but it is also reasonable to believe there is perhaps not all that much space between our parties.
If there is not a comparatively large amount of ideological space between our leaders, there are certainly diferences in the actual policies they are proposing. In short, they difer on the issues. But does this matter much for vote choice?
Recently, my colleague Patrick Fournier and I undertook a systematic analysis of the factors afecting vote choice in Canada. The Canadian Election Study (of which Fournier is the lead) interviewed close to 1,000 of the same voters in each of the 2004, 2006, 2008 and 2011 elections. We examined the movement of their partisan identification, their values, their leader evaluations, their economic perceptions, and their issue positions, and the efect of all of these on vote choice. This allows us both to estimate the stability of these things (i.e. how much do they change election over election), but also their causal relationship with vote choice (formally, using something called Granger causality).
The results of this are important. While partisan identification and political values (which we can think of as broadly like ideology) are causally prior to vote choice, issues rarely are. Instead, voters take cues from their preferred parties about what positions they should take on issues. When systematic data are brought to bear, issues have little direct efect on vote choice.
Why, then, would party leaders focus on issues at all? There are at least four reasons, each of which recommends the discussion of issues in elections. First, issues may not steer vote choice directly, but they do matter for other elements of the vote. The policies leaders propose can play a partial role in helping voters develop their evaluations of the leaders. Indeed, this was true even in the most notable single issue election of our time, the 1988 contest over free trade. As Richard Johnston and colleagues demonstrated in Letting the People Decide (the best Canadian political science work ever), the function of that issue was to affirm Mulroney’s leadership advantage and to distract from another issue, namely Meech Lake. Mulroney was abetted in this by John Turner, whose own evaluations on Meech Lake made him vulnerable to Ed Broadbent. Even this single issue election was more about what the issue said about each leader than about the issue itself. So it is, I would venture, with Mr. Mulcair’s support of a limited minimum wage increase and Mr. Harper’s continued commitment to policies meant to increase home ownership, not to mention various foreign commitments. These issues are about their leaders as much as they are about the voters they are meant to help.
Second, a discussion of issues followed closely by the media and other opinion leaders (including, importantly, various Twitterati), is a test of leaders and their mettle.
Can they propose a coherent policy package and defend it from continued scrutiny? If they cannot, they will not withstand the scrutiny of government.
Third, leaders do have to govern following an election, and it is useful to have a list of policies to fill the first months and years of government.
This is something diferent than a mandate, to be sure. There’s little theoretical reason to think that governments elected by a mere plurality of voters have a mandate for particular policies. But it is equally useful for a government caucus to have some signposts on how to proceed.
Fourth, absent a discussion of issues, on what would our campaigns turn?
We can only take so much discussion of vision and vague images of the future, especially given the comparatively uninspired nature of Canadian political rhetoric. Issues are as good as anything else as a basis of campaign materials and news reports, even if voters do not move much on them.
Let’s just not pretend they matter all that much.
Issues may not steer vote choice directly, but they do matter for other elements of the vote. The policies leaders propose can play a partial role in helping voters develop their evaluations of the leaders. — Peter Loewen Voters take cues from their preferred parties about what positions they should take on issues.