Does Immigration Drive Up Canadian Housing Prices?

The Free Press recently published a fascinating article claiming that immigration is not a significant cause of housing cost increases in the U.S. I’m not sure I’m completely convinced by their arguments, but the piece immediately got me wondering about Canada. Is there a meaningful relationship between the massive waves of immigration we’ve experienced and our own skyrocketing housing costs?

Fortunately, there are two Statistics Canada datasets that will be particularly helpful here: the shelter costs component of the Consumer Price Index and immigration numbers from the components of population change table. Unlike the U.S., we do have government data tracking where immigrants settle.

Those two tables provide overlapping data for 16 large municipal areas. Since most immigrants (92 percent of them in 2022/2023) settle in census metropolitan areas, those are obviously the regions on which we should focus.

One problem with the immigration data in its original form is that it gives us raw immigrant numbers for each city. This could potentially distort our results if, for example, a massive city like Toronto attracted only twice as many immigrants as the relatively tiny Charlottetown. To overcome that, I converted the immigration numbers to a proportional amount (per 1,000 total residents).

I then calculated the statistical relationship between immigration numbers and shelter costs using the Pearson correlation coefficient.

The Pearson correlation coefficient is a number between -1 and 1 that tells us how strongly two things are connected. 1 means a perfect positive relationship: when one thing goes up, the other always goes up too. -1 means a perfect negative relationship: when one thing goes up, the other always goes down. 0 means no relationship: the two things don't seem to affect each other at all. A higher number (either closer to 1 or -1) means they move more predictably with each other.

Here are the year-by-year results I got:

Bearing in mind that our data sample is smaller than we’d prefer, those do suggest a fairly strong correlation. Of course, that doesn’t prove that higher immigration levels are necessarily causing price increases. There may be other structural inputs at work here. But it is in important piece of evidence. And I think we can be confident ruling out the reverse: immigrants certainly aren’t actively attracted to cities by their higher costs!

I also considered the possibility that the full impact of immigration might not be felt right away. Perhaps it takes a few months for housing markets to adapt to new demographic conditions. So I ran the Pearson calculations again, this time comparing, say, 2016 immigration numbers with 2017 shelter costs. The results showed a slightly stronger correlation:

Here are those numbers mapped over time: