A response to the C.D. Howe Institute’s “Room to Grow: Building More of the Housing Homebuyers Want“
A new E-Brief from the C.D. Howe Institute argues that planning policy in the Greater Toronto and Hamilton Area has become disconnected from the housing preferences of its residents. The prescription: open up more greenfield land, because surveys show homebuyers prefer single-detached homes and townhouses.
It’s a familiar argument, and it rests on a familiar methodological problem: the surveys can’t support the conclusion.
What the surveys actually measure
The brief draws on three datasets — CMHC’s Mortgage Consumer Survey, TRREB’s survey of likely buyers, and Sagen’s survey of first-time buyers. Every one of them samples people who bought a home, or intend to buy one soon.
That matters for two reasons.
First, what people buy reflects what’s available and what they can afford; not what they’d choose in the abstract. Single-detached houses remain the largest single category of housing stock in the Toronto CMA— 39 per cent of all occupied dwellings in the 2021 Census— and they dominate the stock outside the City of Toronto, where most of the region’s transactions happen. It’s not surprising the resale market skews the same way the existing stock does.
Second, asking people what they’d like to buy tells you about aspiration, not planning demand. The single-detached home has been marketed as the symbol of arrival for seventy years; of course it polls well. Ask people whether they’d prefer a bigger home with a yard and most will say yes. They’d probably prefer a mansion with a pool. These answers can’t govern how we build cities.
When surveys do present trade-offs, the picture changes considerably. The 2014 RBC–Pembina Home Location Preference Survey (cited in the brief) found that 81 per cent of GTA respondents would choose a walkable, transit-connected neighbourhood over a larger detached home with a yard, when costs were held equal. Pembina’s companion report, Priced Out, found that more than 70 per cent of GTA residents live where they do because of affordability, not preference.
Who’s missing from the data
Even among the buyers these surveys do capture, “preference for ownership” deserves scrutiny. In Ontario, buying is often less a housing preference than an insurance policy.
Consider what renting offers a family with a child in school and three after-school programs. Ontario has vacancy decontrol: rent rules apply within a tenancy, not between tenancies, which means the system’s incentives quietly work against long-term tenants. Units first occupied after November 15, 2018 are exempt from the provincial rent increase guideline altogether. And losing a tenancy doesn’t just mean moving: it means re-entering the market at vacant-unit prices. The City of Toronto’s Housing Data Book puts numbers on that penalty: a family moving into a vacant purpose-built rental with three or more bedrooms would pay about 50% more than what sitting tenants currently pay, with average asking rent at $3,109. When the rental system cannot promise a family they’ll be in the same school catchment next September, a mortgage becomes the only tenure security money can buy. That shows up in a buyer survey as a “preference” for ownership. It’s closer to a flight from precarity.
When the rental system cannot promise a family they’ll be in the same school catchment next September, a mortgage becomes the only tenure security money can buy.
The second problem is that a family-sized apartment (which might be a reasonable alternative to a suburban townhome) barely exists. City of Toronto pipeline data found that of 78,339 apartment units proposed between May 2006 and April 2011, just 3,558— 3.1 per cent— had three or more bedrooms. Toronto’s 2020 guidelines now encourage new developments to include at least 25 per cent larger units (15 per cent two-bedroom, 10 per cent three-bedroom); the author of the brief concedes that developers have built relatively few. You cannot express a preference for a three-bedroom apartment in a walkable neighbourhood if that unit was never built. Surveys measure choices among the options that exist; and for families, the apartment with enough space option mostly doesn’t.
So when the brief observes that families “prefer” ground-oriented homes, it’s reading the endpoint of a process and calling it a starting point. Fix tenure security and build family-sized homes in existing neighbourhoods, and the revealed “preferences” would look rather different.
The cost of single family housing
The brief points to a striking figure: at the end of 2022, the GTHA’s approved and proposed development pipeline totalled 911,748 units — 86 per cent apartments, 14 per cent ground-related homes. This is presented as evidence that planners have ignored buyers.
But that pipeline is what happens when land is expensive and growth needs to pay for itself. Apartments are what pencils on costly urban land, and what pencils for the city servicing them. Ottawa’s 2021 fiscal analysis found that low-density greenfield homes costs the city $465 per person per year beyond what they return in taxes and fees, while high-density infill generates a $606 per-capita annual surplus. The apartment-heavy pipeline isn’t an ideological quirk of the planning profession. It’s arithmetic.
Building outward is the most expensive way to grow: for buyers carrying the price of new detached construction, for municipalities carrying the pipes, roads, and transit forever after, and for everyone who inherits the fiscal gap. A region adding hundreds of thousands of households cannot subsidize its way to affordability one cul-de-sac at a time.
Where the brief is right
None of this means the current housing mix is fine. The brief is correct that the GTHA underbuilds ground-oriented housing: townhouses, stacked towns, duplexes through sixplexes, garden apartments.
But that’s an argument for missing middle housing in the neighbourhoods we’ve already built and serviced, which the brief itself concedes is “a step in the right direction”, not for treating farmland as the release valve. The choice on offer is not actually “high-rise condo or detached house on a greenfield”; there’s an entire spectrum in between, and it’s precisely the part of the spectrum that decades of exclusionary zoning made illegal to build where people most want to live.
Preferences are not needs
City-building has never been the aggregation of individual purchase preferences, and it shouldn’t be. It’s how communities meet collective needs across incomes, tenures, and generations, within real fiscal and land constraints. The market does not, by nature, produce housing for low-income people, free community gathering spaces, or parks: we have to plan for that.
If you want to see what your community actually needs— by income, household size, and priority population— our Housing Needs Assessment Tool covers every community in Canada. The data is free, and unlike a preference survey, it counts everyone.
