An ecological analysis of walkability and housing affordability in CanadaModeration by city size and neighbourhood property type composition

Chelsea D. Christie, Christine M. Friedenreich, Jennifer E. Vena, Dany Doiron, Gavin R. McCormack•View original

OverviewBalancednova voice

Here's a question that sounds almost too straightforward: if walkable neighbourhoods are good for your health, why can't everyone just live in one? The answer, it turns out, has less to do with urban planning and more to do with the housing market. A large Canadian study has now put some hard numbers on exactly how complicated that relationship gets. Christie, Friedenreich, Vena, Doiron, and McCormack set out to answer two related questions. First: does living in a walkable neighbourhood cost more? And second: does the answer depend on what kind of city you're in and what kind of housing surrounds you? To find out, they linked a nationally scaled walkability measure to census housing data across more than thirty thousand neighbourhoods in thirty-one Canadian cities. The study uses a cross-sectional, ecological design — meaning it works at the neighbourhood level, not the individual level — and the analysis was approved by the University of Calgary ethics board. The walkability measure they used is called the Canadian Active Living Environments index, or Can-ALE. Think of it as a composite score that captures four things about a neighbourhood's built environment: how dense the street intersections are, how many dwellings are packed into the area, how many destinations like parks, shops, and schools are nearby, and how many transit stops are within reach. The intersection density measure includes off-road footpaths and recreational trails but excludes highways. Each of those four components gets standardized to a z-score, and the Can-ALE score is their sum. It can range from about negative four at the low end to nearly forty-eight at the high end — a wide spread that captures everything from a sparse rural fringe to a dense urban core. Neighbourhoods here are defined as dissemination areas, Statistics Canada's smallest geographic unit, typically housing four hundred to seven hundred people. The housing affordability outcome is the median home value within each neighbourhood, drawn from the two thousand sixteen census long-form survey and log-transformed to handle the skewed distribution of property prices. The study covers thirty-three thousand twenty-six dissemination areas across thirty-one Canadian census metropolitan areas. Four metropolitan areas had to be excluded because they lacked the transit data needed for Can-ALE, and a further two thousand two hundred dissemination areas were removed for missing data. The analysis uses multilevel linear regression with neighbourhoods nested inside cities, and the researchers applied group-mean centering to separate the effect of walkability within a city from broader differences between cities. They also allowed the walkability slope to vary randomly by city — a way of acknowledging that walkability might carry different price implications in different urban markets. So what did they find? At the national level, after adjusting for neighbourhood characteristics like median household income, dwelling age, proportion of homeowners, and the share of non-movers — the answer is essentially nothing. The adjusted association between neighbourhood walkability and log median home value sits at negative zero-point-zero-one, with a confidence interval running from negative zero-point-zero-two to zero-point-zero-zero. No universal walkability premium. No national penalty either. The city-level walkability effect, while positive in unadjusted terms, is also not statistically significant after adjustment. But that flat national average is hiding something important. When the researchers tested whether the walkability-home-value relationship differed by city size and housing composition — specifically, the proportion of single-detached homes in a neighbourhood — a striking pattern emerged. The three-way interaction among walkability, city size, and detached-home share carries a p-value of zero-point-zero-one-four. That's not noise. That's the data telling you the relationship is genuinely conditional. Here's what that means in practice. In neighbourhoods where a high proportion of homes are single-detached — the classic suburban housing form — the walkability story flips depending on city size. In small cities, those with populations between thirty thousand and ninety-nine thousand nine hundred ninety-nine, higher walkability in detached-dominated neighbourhoods is actually associated with lower home values. The coefficient is negative zero-point-zero-five, with a confidence interval running from negative zero-point-one-zero to negative zero-point-zero-one. Medium cities — populations up to one million — show the same direction, a coefficient of negative zero-point-zero-four, with a confidence interval from negative zero-point-zero-six to negative zero-point-zero-two. In both cases, more walkable means less expensive in those housing contexts. For lower-income households in smaller Canadian cities, that's actually encouraging news. Then you get to extra-large cities — Toronto, Montreal, Vancouver — and the pattern reverses completely. In high-detached neighbourhoods within these mega-markets, walkability is positively associated with home values, with a coefficient of positive zero-point-zero-six and a confidence interval from zero-point-zero-one to zero-point-one-zero. More walkable means more expensive. The large-city category — populations between one million and two million — shows no statistically robust effect in the high-detached segment. For neighbourhoods with a low proportion of detached homes, none of these associations were statistically meaningful regardless of city size. The effect is specific to that detached-home context. Take a moment with that inversion. The same physical feature — a neighbourhood you can walk through, with destinations reachable on foot — carries opposite price signals depending on where you are. In small and medium cities, walkable detached neighbourhoods are cheaper. In the biggest cities, they're pricier. The authors' interpretation is that city-scale market dynamics are driving this. In extra-large cities, walkable housing is scarce relative to demand, so the premium gets bid up. In smaller cities, that scarcity pressure doesn't exist in the same way. Gentrification and intense demand concentration are phenomena of large, competitive housing markets — not every Canadian city. This matters enormously for equity. If you're a lower-income household in a small or medium Canadian city, the data suggest that walkable neighbourhoods — at least those dominated by detached housing — are not pricing you out. But if you're trying to find affordable access to a walkable neighbourhood in Vancouver or Toronto, the market is working against you. The walkability advantage that researchers tout for health outcomes is increasingly a luxury good in Canada's biggest cities, at least when measured through home values. The researchers are careful about what this study can and cannot tell us. It's cross-sectional — a snapshot of two thousand sixteen, not a causal story. The housing data come from owner-occupied dwellings, which means renters, who are disproportionately lower-income, are largely invisible in the outcome measure. The Can-ALE index uses a one-kilometre buffer, and different buffer choices might shift the results. There are also unmeasured confounders — neighbourhood amenities, school quality, crime rates — that could be doing some of the work attributed to walkability. Because the home value data derive from census long-form responses about owner-occupied, non-farm private dwellings, neighbourhoods with very few owners may be underrepresented. What the study does establish clearly is that one-size-fits-all walkability policy is inadequate. The authors point to tools like inclusionary zoning — requirements that new developments include a share of affordable units — and mixed-income housing near high-walkability, transit-adjacent areas as mechanisms to preserve equitable access. They cite Canada's National Housing Strategy and the Rapid Housing Initiative as existing policy frameworks that align with this kind of mixed-tenure, transit-adjacent thinking. The argument isn't that walkability is bad. It's that, in large cities especially, the health benefits of walkable design need to be deliberately decoupled from the price pressures that make those neighbourhoods inaccessible to the people who might benefit most. The broader takeaway is both nuanced and actionable. Walkability's relationship with housing costs is not a fixed property of urban design — it's contingent on the scale of the city and the structure of its housing stock. In Canada's smaller cities, walkable neighbourhoods appear financially accessible. In its largest, they are becoming premium products. Understanding that distinction is the first step toward building cities where the ability to walk to work, to school, or to the grocery store doesn't depend on your income.

Christie, Friedenreich, Vena, Doiron, and McCormack set out to answer two related questions. First: does living in a walkable neighbourhood cost more? And second: does the answer depend on what kind of city you're in and what kind of housing surrounds you?

To find out, they linked a nationally scaled walkability measure to census housing data across more than thirty thousand neighbourhoods in thirty-one Canadian cities. The study uses a cross-sectional, ecological design — meaning it works at the neighbourhood level, not the individual level — and the analysis was approved by the University of Calgary ethics board.

The walkability measure they used is called the Canadian Active Living Environments index, or Can-ALE. Think of it as a composite score that captures four things about a neighbourhood's built environment: how dense the street intersections are, how many dwellings are packed into the area, how many destinations like parks, shops, and schools are nearby, and how many transit stops are within reach. The intersection density measure includes off-road footpaths and recreational trails but excludes highways.

Each of those four components gets standardized to a z-score, and the Can-ALE score is their sum. It can range from about negative four at the low end to nearly forty-eight at the high end — a wide spread that captures everything from a sparse rural fringe to a dense urban core. Neighbourhoods here are defined as dissemination areas, Statistics Canada's smallest geographic unit, typically housing four hundred to seven hundred people.

The housing affordability outcome is the median home value within each neighbourhood, drawn from the two thousand sixteen census long-form survey and log-transformed to handle the skewed distribution of property prices. The study covers thirty-three thousand twenty-six dissemination areas across thirty-one Canadian census metropolitan areas. Four metropolitan areas had to be excluded because they lacked the transit data needed for Can-ALE, and a further two thousand two hundred dissemination areas were removed for missing data.

The analysis uses multilevel linear regression with neighbourhoods nested inside cities, and the researchers applied group-mean centering to separate the effect of walkability within a city from broader differences between cities. They also allowed the walkability slope to vary randomly by city — a way of acknowledging that walkability might carry different price implications in different urban markets.

So what did they find? At the national level, after adjusting for neighbourhood characteristics like median household income, dwelling age, proportion of homeowners, and the share of non-movers — the answer is essentially nothing. The adjusted association between neighbourhood walkability and log median home value sits at negative zero-point-zero-one, with a confidence interval running from negative zero-point-zero-two to zero-point-zero-zero.

No universal walkability premium. No national penalty either. The city-level walkability effect, while positive in unadjusted terms, is also not statistically significant after adjustment.

But that flat national average is hiding something important. When the researchers tested whether the walkability-home-value relationship differed by city size and housing composition — specifically, the proportion of single-detached homes in a neighbourhood — a striking pattern emerged. The three-way interaction among walkability, city size, and detached-home share carries a p-value of zero-point-zero-one-four.

That's not noise. That's the data telling you the relationship is genuinely conditional.

Here's what that means in practice. In neighbourhoods where a high proportion of homes are single-detached — the classic suburban housing form — the walkability story flips depending on city size. In small cities, those with populations between thirty thousand and ninety-nine thousand nine hundred ninety-nine, higher walkability in detached-dominated neighbourhoods is actually associated with lower home values.

The coefficient is negative zero-point-zero-five, with a confidence interval running from negative zero-point-one-zero to negative zero-point-zero-one. Medium cities — populations up to one million — show the same direction, a coefficient of negative zero-point-zero-four, with a confidence interval from negative zero-point-zero-six to negative zero-point-zero-two. In both cases, more walkable means less expensive in those housing contexts.

For lower-income households in smaller Canadian cities, that's actually encouraging news.

Then you get to extra-large cities — Toronto, Montreal, Vancouver — and the pattern reverses completely. In high-detached neighbourhoods within these mega-markets, walkability is positively associated with home values, with a coefficient of positive zero-point-zero-six and a confidence interval from zero-point-zero-one to zero-point-one-zero. More walkable means more expensive.

The large-city category — populations between one million and two million — shows no statistically robust effect in the high-detached segment. For neighbourhoods with a low proportion of detached homes, none of these associations were statistically meaningful regardless of city size. The effect is specific to that detached-home context.

Take a moment with that inversion. The same physical feature — a neighbourhood you can walk through, with destinations reachable on foot — carries opposite price signals depending on where you are. In small and medium cities, walkable detached neighbourhoods are cheaper.

In the biggest cities, they're pricier. The authors' interpretation is that city-scale market dynamics are driving this. In extra-large cities, walkable housing is scarce relative to demand, so the premium gets bid up.

In smaller cities, that scarcity pressure doesn't exist in the same way. Gentrification and intense demand concentration are phenomena of large, competitive housing markets — not every Canadian city.

This matters enormously for equity. If you're a lower-income household in a small or medium Canadian city, the data suggest that walkable neighbourhoods — at least those dominated by detached housing — are not pricing you out. But if you're trying to find affordable access to a walkable neighbourhood in Vancouver or Toronto, the market is working against you.

The walkability advantage that researchers tout for health outcomes is increasingly a luxury good in Canada's biggest cities, at least when measured through home values.

The researchers are careful about what this study can and cannot tell us. It's cross-sectional — a snapshot of two thousand sixteen, not a causal story. The housing data come from owner-occupied dwellings, which means renters, who are disproportionately lower-income, are largely invisible in the outcome measure.

The Can-ALE index uses a one-kilometre buffer, and different buffer choices might shift the results. There are also unmeasured confounders — neighbourhood amenities, school quality, crime rates — that could be doing some of the work attributed to walkability. Because the home value data derive from census long-form responses about owner-occupied, non-farm private dwellings, neighbourhoods with very few owners may be underrepresented.

What the study does establish clearly is that one-size-fits-all walkability policy is inadequate. The authors point to tools like inclusionary zoning — requirements that new developments include a share of affordable units — and mixed-income housing near high-walkability, transit-adjacent areas as mechanisms to preserve equitable access. They cite Canada's National Housing Strategy and the Rapid Housing Initiative as existing policy frameworks that align with this kind of mixed-tenure, transit-adjacent thinking.

The argument isn't that walkability is bad. It's that, in large cities especially, the health benefits of walkable design need to be deliberately decoupled from the price pressures that make those neighbourhoods inaccessible to the people who might benefit most.

The broader takeaway is both nuanced and actionable. Walkability's relationship with housing costs is not a fixed property of urban design — it's contingent on the scale of the city and the structure of its housing stock. In Canada's smaller cities, walkable neighbourhoods appear financially accessible.

In its largest, they are becoming premium products. Understanding that distinction is the first step toward building cities where the ability to walk to work, to school, or to the grocery store doesn't depend on your income.