Why Rob Ford won Toronto: Car lovers

Long plane rides can be a drag — I'm on one as I type this between Ottawa and Vancouver en route to cover the NDP convention this weekend — but having a few political science papers on the ol' hard drive to pass the time with sure helps.

And so it was that I finally had a chance to read a fascinating analysis of Toronto Mayor Rob Ford's electoral success by Zack Taylor, a doctoral candidate in the department of political science at the University of Toronto. Taylor presented a paper [pdf] this spring to the Canadian Political Science Association annual conference and, while he cautions that the techniques he uses are “exploratory” and his analysis “preliminary”, it all smells about right to me.

Taylor is the first to use what he calls an “ecological” analysis of a municipal electoral event. Rather than take ward-by-ward or census tract data and match that to how the votes went, Taylor breaks it down further on a poll-by-poll basis and then uses GIS software to match poll data with a variety of economic, social, and geographic data to try to find some common characteristics about the kind of voter that swept the populist, small-c conservative Ford to victory over the big-name, big-spending, Bay Street favourite George Smitherman.

Taylor's conclusion:

Perhaps the most important finding is that location of residence — urban versus suburban — is the strongest predictor of Ford support. The underlying factors driving this effect only partially conform to expectations. The propensity to commute by automobile is a strong predictor of Ford support, while property-oriented variables (the home ownership rate and percentage of housing in detached form) are shown to have a negligible influence on candidate support.

That last bit — whether you owned a home or not didn't have much to do with how Torontonians voted for mayor — seems an interesting bit. I'm no political scientist but I think there is an assumption out there that those who own homes, as oppose to rent a residence, tend to vote in higher numbers for conservative candidates. If that is the case, that rule wasn't in effect for Toronto's last municipal election.

Taylor finds that there was a strong connection between Ford support and automobile use. Indeed, Ford explicitly campaigned on ending the “war on the car”, which included cutting support for some big public transit items. (By contrast, Calgary mayor Naheed Nenshi — another out-of-nowhere candidate like Ford — triumphed in Calgary with a platform that included more public transit support, particularly for that city's light rapid transit service)

Now, as Taylor himself mentions, his analysis is “preliminary”. Nonetheless, I thought it noteworthy first for his approach — poll-by-poll analysis seems more “real” to me — and because his research finds some neat new stuff. But, as he notes at the end of the paper, “A useful next step would be to increase the explanatory power of the models by adding other variables shown to be influential in other studies: size of age cohorts, educational attainment, and immigration, religiosity, and employment by sector.”

Finally, Taylor writes: “This paper was completed prior to the May 2, 2011 federal election. It will be interesting to see to what extent the spatial distribution of the Ford vote is mirrored by the Conservatives in the imminent federal election and October’s provincial election. If so, it would add strength to the argument that urban and suburban voters possess divergent political values that transcend the municipal level.”

Indeed it would, Mr. Taylor! Someone get that man an SSHRC grant so he can present that paper at next year's political science convention.