Redistricting and Incumbency Advantage: Evidence from Toronto and Calgary

• In Event: Understanding Municipal Elections: New Data and Approaches from Canada

Abstract

In Canadian city elections, incumbent candidates dominate. Past research has demonstrated that re-election rates among incumbent council candidates regularly reach rates of 90, 95, even 100 percent, and incumbents tend to dominate fundraising and media attention. These success rates resemble those in other non-partisan city elections, such as in the United States. While incumbent success rates are well documented, however, we have a limited understanding of the distinctive sources of incumbency advantage in city elections.

In this paper, we focus on the role of personal familiarity and experience with incumbent politicians -- often called the "personal vote" -- in municipal elections. To isolate this component of incumbency advantage, we leverage redistricting processes in two recent high-profile city elections: Calgary 2017 and Toronto 2018. In Calgary, a redistricting process prior to the election resulted in substantial reshaping of the electorate in several city wards. In Toronto, a much more contentious redistricting process saw the Ontario provincial government impose a radically reconfigured ward system on the city at the last minute, resulting in a number of ward races which pitted incumbents against one another.

Drawing on data from three surveys, collected in Calgary in 2017 and 2018 and Toronto in 2018, we undertake three analyses which take advantage of these redistricting processes – natural experiments - to understand the personal vote. Using Calgary 2017 survey data, we demonstrate that an incumbent's "new" voters are less likely to support them, and more likely to perceive a race as competitive, than the incumbent's "old" voters. Then, using a unique panel survey of the same respondents in 2018, we provide evidence about the persistence of these effects over time. Finally, we use data on incumbent-vs-incumbent races in Toronto in 2018 to test how familiarity, ideological proximity, and competitiveness relate to vote choice in the unusual context of incumbent-versus-incumbent races. We suggest that the personal vote is an important but relatively small component of municipal incumbency advantage.

While the quasi-random character of redistricting processes has been used productively in American studies of incumbency advantage, it has also been criticized as failing to account for non-random differences between "old" and "new" voters. However, by using individual-level survey data (with a total N of over 6,000), rather than aggregate outcomes, we are able to test and adjust more extensively for balance across old and new voters. Our study therefore also has methodological implications for scholars who may wish to undertake similar studies in other political contexts.

Authors

• Jack Lucas, University of Calgary

• Michael McGregor