In the 2025 federal election, Canadians voted at the highest rate since 1993. Explanations for high or low turnout usually focus on the national election campaign, including how close the race was and what seemed to be at stake. As lead author of a new study published in the Canadian Journal of Political Science, I worked with my co-authors — Daniel Rubenson, Roosmarijn de Geus, Benjamin Allen Stevens and Peter John Loewen — to look close to home and examine how someone’s political views compare with those of the people around them. Our research shows that people whose views closely match the average views in their riding are less likely to vote than those whose views differ moderately. Similarly, people whose views differ extensively from those around them are less likely to vote. In short, voting peaks at moderate levels of disagreement. Measuring disagreement next doorWe used the Local Parliament Project, an online survey of more than 31,000 Canadians conducted during the 2015 federal election campaign, in our research. It included respondents from nearly every riding, which allowed us to compare each person’s views with those of others in the same riding. Research shows labels like “left” and “right” mean different things to different people. We, therefore, measured political views using people’s positions on 37 specific issues, such as taxes, health care and immigration. This allowed us to see how far each person’s views were from their neighbours’ in the same riding, and the diversity of opinions across the riding as a whole. This captured different information. For example, a riding’s voters can be largely like-minded, yet some people living there can still hold differing views about specific issues.A sweet spot for votingWe found that a person whose views are very similar to their riding’s average has a predicted 83 per cent chance of voting. A person at the opposite extreme has a predicted 63 per cent chance, though this estimate is less precise. People halfway between these extremes have the highest likelihood of voting, at 90 per cent.These patterns hold after we account for factors that include age, education, income, how close the local race was and how much parties campaigned. Ridings with tight races in 2015 did not have higher turnout than those where one party won easily.Our riding-level analysis shows a similar picture. Predicted turnout is 65.7 per cent at the lowest level of opinion diversity, rises to 68.7 per cent at the mid-point and falls to 65.3 per cent at the highest. However, these differences are not statistically significant once we account for other factors.The link between a person’s views and voting also varies with the local mix of views. In ridings with a wide range of views, voting peaks at smaller differences and drops off more steeply for people whose views differ extensively. So two people equally far from how most of their neighbours intend to vote can have different chances of showing up to cast a ballot depending on their riding.Disagreement can boost voter turnoutWhen our views match those around us, we rarely hear disagreement. Politics can feel settled, and our vote may seem unnecessary. Some disagreement changes that. It makes political differences more visible and the choice on the ballot feel more important.But too much disagreement can do the opposite. When we’re surrounded by very different views, the competing signals can make politics confusing and overwhelming. This can also feel socially uncomfortable.Earlier research has found that people in highly conflict-ridden social networks sometimes pull back from politics to avoid straining relationships with friends, family and co-workers.What this means for CanadaOur data come from 2015, when Liberal Justin Trudeau ended up winning a majority government, but the campaign was a tight three-way race for much of its run. There was a wide range of views across ridings, making it a suitable election for studying how disagreement relates to voting. We think the same forces are likely at work in later elections, including 2025. So how voters experience disagreement in their own communities may matter for turnout, not just how close the race is.The findings also matter when riding maps change. When boundaries are redrawn, like they were in Québec in 39 provincial ridings ahead of its recent election, the mix of neighbours who voters live among shifts too.Our study also adds to the debate about political division. Research on Canadian voting habits has focused on how warmly or coldly people feel toward the other side. Read more: Canada isn’t deeply polarized — yet. What new research reveals about partisan animosity Our results raise a different concern. People whose views closely mirror those around them are somewhat less likely to vote than those whose views differ moderately.Our study has a clear scope. It covers one election in a system where the candidate with the most votes wins each riding. Our survey results rely on stated intentions to vote, and show an association between how much people’s views differ from those around them and voter turnout. The next test is whether these findings held in 2025.The lesson, however, is worth taking seriously. A healthy democracy may depend not on eliminating political disagreement, but on keeping it within a range people can engage with.Md Mujahedul Islam receives funding from the Social Sciences and Humanities Research Council of Canada (SSHRC) through a postdoctoral fellowship, and has received a book manuscript completion grant from the Institute for Humane Studies.