Zack Taylor

The stark geography of housing prices in Vancouver

Posted by | projects | No Comments



I would love to be able to map this kind of data for Toronto and Montreal. Never has the divide between West Van and East Van been more clear. (For the associated article in the Globe and Mail, click here.)

It’s instructive to put this side-by-side with local election returns. Here’s a map from a very good SFU masters thesis by Michelle Vernooy, which shows the areas supporting the centre-left Vision party versus the centre-right NPA:



In search of a winning progressive coalition

Posted by | data-driven musings | 2 Comments

A while back I was asked to give a presentation on what I thought the fundamental divide was in Toronto city politics, and how a “progressive” candidate might win an election. I’ve attached the presentation as a PDF here. The main takeaway messages are:

  • There is a real city-suburb divide that has persisted across multiple elections. To win, a “downtown” candidate must carry his or her base but also appeal to suburban voters. Two-thirds of the votes are in the suburbs; even if the downtown voted as a bloc, it cannot win a citywide election on its own. More than half of Smitherman’s votes came from outside of the old City of Toronto, but they were not enough to win.
  • Examining the aggregate demographics of the parts of the city in which Smitherman and Ford won a majority of the vote reveals that there are two worlds within the City of Toronto. On average, Smitherman Village is wealthier and more educated than Ford Nation, contains more tenants and fewer homeowners, and contains more professionals. Home values are higher and fewer people commute by car. Ford Nation also contains more first-generation immigrants and visible minorities.
  • A more detailed statistical analysis suggests that, neighbourhood by neighbourhood, the propensity to commute by automobile is the greatest predictor of Ford support, while higher levels of educational attainment and visible minorities as a proportion of the population are the strongest correlates of Smitherman support. Interestingly, income and home values were not statistically significant correlates of support for either candidate.
  • This suggests that the 2010 election cannot be read simply as the revenge of the squeezed lower-middle class or an upper-middle class property tax revolt. The common narrative that Ford was buoyed by the rise of a new conservative ethnic voter in suburbs also doesn’t play out. Neighbourhoods with many foreign-born residents leaned toward Ford, while those with many visible minorities (i.e., more recently arrived immigrants) tended to lean toward Smitherman. In other research I have found the same split among support for Conservative and NDP candidates at the federal level.
  • Instead, it seems that there is a lifestyle divide between core and suburban voters than leads them on average to behave differently regardless of whether they are rich or poor, homeowners or tenants.
  • Finally, a comparison between the 2003 and 2010 elections shows that John Tory and Rob Ford had a very different distribution of support, as did David Miller and George Smitherman. This suggests that while there is a core-suburb divide, it is not insurmountable and communities can be activated in different ways depending on the special dynamics of the election campaign. Also, no candidate should forget that in the 2011 provincial and federal elections, the Liberals were the first-place party in the City of Toronto, with the NDP and the Conservatives vying for second place.

The politics and economics of local government fragmentation

Posted by | data-driven musings | No Comments

In a cute little analysis, Richard Florida and U of T’s Martin Prosperity Institute have mapped local government fragmentation by American state (see the map above). It is sobering to see that Illinois, Pennsylvania, and Texas have more local governments than all of Canada (over about 4,000). (It is unclear, however, whether they are including special-purpose local governments  in the total, or are only counting general-purpose municipal governments.)

Read More

Transit lines and density

Posted by | data-driven musings | 2 Comments

I recently stumbled on a map I made last year that plots present and proposed transit lines against population density. Density  shown at the level of the city block using information from the 2011 Census.

Higher-density areas — those where the density is above 50 people per hectare — are shown in pink. Lower-density areas are shown in blue. Very low-density areas, which are typically employment lands and utility corridors, are white or grey. The threshold of 50 people per hectare is considered by some to be the minimum threshold for the provision of frequent-service transit.

What the map shows, I think, is that the case for surface LRT over subways is strong along the corridors proposed by Metrolinx and the TTC. The proposed LRT lines along Eglinton West and East and Finch would largely run through lower-density areas. Much of the existing subway system runs through broad areas of high density corresponding to the prewar city. The underground central portion of the Eglinton Crosstown line also runs through this zone.

While some corridors and transit station areas (but not the broader residential areas that would feed into them) have been intensified over the past 20 years — the North York City Centre area, the Sheppard corridor east of Yonge, and the Kipling/Islington Station areas — I’m not convinced that this can be replicated, at least anytime soon, in the blue zones where Mayor Ford wants to run subways.

Voter turnout in the City of Toronto – Part 2

Posted by | data-driven musings | One Comment

Why does turnout vary across space?

I argue that at least some of this variation is due to demographic and socio-economic differences.

The maps on the left show wards with below-average turnout. (The above-average ones are not shown.) On the right I have mapped census tracts in which there is:

  • a majority population of first-generation immigrant residents;
  • a majority population of visible minority residents, which typically corresponds with more recent immigration; and
  • below-average median household income. (These are census tracts in which the average income of all households in the tract is less than the citywide average.)

Turnout and demographics

Read More

Voter turnout in the City of Toronto – Part 1

Posted by | data-driven musings | One Comment

Last week I was invited to speak on a panel with Dave Meslin on civic engagement in Toronto. I thought it might be interesting to revisit the geography of turnout — how some parts of the city seem more included to vote than others, election after election. I had earlier looked at this in the context of the 2010 election. Using City of Toronto Elections data, I made ward-level maps of turnout for the 2003, 2006, and 2010 elections.

Read More

Project: Ward boundary review in Toronto

Posted by | projects | 3 Comments

The disparity in the populations of the City of Toronto’s wards is large and increasing. This undermines equality of electoral representation, residents’ access to councilors, and the quality of constituency services provided by councilors. All councillors have identical office resources and time constraints, yet some must interact with more constituents than others. This imbalance will be resolved one or way or another before the next election as the City undertakes a review of ward boundaries. The question is — how? And can it be done in such a way as to improve civic democracy, accountability, and responsiveness?

Read More