By bringing together workers, capital, businesses, and ideas in a compact geographic market, cities promote improved productivity performance, and thereby faster economic growth and higher real incomes for workers. The affordability of housing in a city can aid or hinder this synthesis. As demand for labour grows in the most productive metropolitan areas, workers’ willingness to relocate in pursuit of better opportunities in the job market becomes central not only to their own fortunes, but also to overall economic efficiency and industrial competitiveness. An important factor conditioning labour mobility is the affordability of housing, which varies greatly among regions and cities. In particular, if the expected productivity and wage gains associated with the geographical mobility of labour are blunted—partially or fully—by higher housing costs, overall standards of living can suffer.

This publication provides a better understanding of how housing affordability has evolved in metropolitan areas, comparing how shelter costs, as a share of income, have changed between 2006 and 2016 for 396 Canadian and American metropolitan areas. Specifically, we identify the growth in the share of median gross incomes dedicated to shelter costs, per bedroom, in each metropolitan area, for those two years.

According to this measure of changes in affordability, shelter costs fell as a share of median incomes in the vast majority (312) of the full sample of Canadian and US metropolitan areas over this period. The overall decline averaged 7.3% (8.7% when weighted by population). However, most of the urban regions that showed improved housing affordability were in the United States, with all but three Canadian locations (out of 52) exhibiting increases in per-bedroom shelter costs as a share of incomes. In fact, shelter costs as a share of income increased by 7.2% across the full sample of Canadian cities over this period (7.6% when weighted by population).

We also explore the relationship between changes in affordability and population growth. We find that a majority of metropolitan areas was able to combine improved affordability with a growing population. Here too, however, Canada differs from the United States, with 46 of 52 Canadian cities combining growing populations with reduced affordability; this includes the country’s six metropolitan areas with more than one million inhabitants. However, for both the American and Canadian samples, the statistical relationship between the percentage change in population and the percentage change in the share of income dedicated to shelter between 2006 and 2016 is weak; this suggests that decreasing housing affordability is not an inevitable consequence of an increase in housing demand generated by a growing population.

We finish by decomposing the data in our measure of affordability to separate out changes in median incomes and shelter costs and glean preliminary insights into what is driving the divergence between Canadian and American metropolitan areas. We find that nominal median household income actually grew faster in our sample of Canadian cities than in US cities. However, shelter costs per bedroom grew significantly faster in Canadian locations than in the US locations in our sample. Hence, the decline in housing affordability in Canada relative to the United States reflects a more rapid increase in shelter costs—rather than a slower growth of median incomes—in Canadian cities compared to the majority of their American counterparts.

Our findings suggest several areas for future research. One is the identification of the factors that contributed to the much faster growth of shelter costs in Canadian cities relative to US cities. Our study identifies candidate regions for case studies examining the factors influencing changes in shelter costs. A second is the extent to which the eroding affordability of housing in Canadian cities affected the geographic mobility of Canadian workers.