“Ethnicity” doesn’t seem to explain why more 25–34-year-olds are living with their parents in Ontario… but not in other provinces.

Playing around, and getting some help, with some census data.

TL;DR: The percentage of 25–34-year-olds living with their parents in Ontario communities like London, Guelph, and Barrie rose substantially from 2001 to 2016. Yet the proportion stayed relatively steady in other Canadian cities like Montréal, Edmonton, St. John’s, Regina, and Calgary. Changes in the “ethnic” make-up of these communities does not appear to explain why this happened.

I wrote a piece earlier this week titled More 25–34-year-olds are living with their parents in cities in Ontario… but not in Alberta and Quebec. The TL;DR for the piece was as follows:

The percentage of 25–34-year-olds living with their parents in Ontario communities like London, Guelph, and Barrie rose substantially from 2001 to 2016. Yet the proportion stayed relatively steady in other Canadian cities like Montréal, Edmonton, St. John’s, Regina, and Calgary. We ought to be asking why.

I received a lot of fantastic feedback and questions on the piece. Common responses were along the lines of “One thing you might want to use as a control here is ethnicity” and “Culture playing a part?”

It’s a great question. Fortunately for us, University of Waterloo economist Mikal Skuterud also thought it was an interesting question, so he decided to take a look. He was gracious enough to e-mail his findings to me.

First, I’ll start with my summary of his findings. Then, I will post his methodology and findings, in his words, in full. I’m so grateful he’s permitting me to share this with you.

What Prof. Skuterud did is take 2001 and 2016 Census data for individuals aged 25–34 living in one of Canada’s 20 biggest metros (CMAs). Ten of those are in Ontario (Ottawa, Oshawa, Toronto, Hamilton, St. Catharines-Niagara, Kitchener-Cambridge-Waterloo, London, Windsor, Sudbury, and Thunder Bay) and ten are not (Halifax, Quebec City, Montreal, Winnipeg, Regina, Saskatoon, Calgary, Edmonton, Vancouver, and Victoria).

Here is the data, before controlling for culture or ethnicity:

Proportion of 25–34-Year-Olds Living with a Parent in Ontario and Non-Ontario Metros

In 2001, 13.3% of 25–34-year olds in our non-Ontario metros lived with a parent; 15 years later this had increased to 15.8%, a 2.5 percentage point increase.

In Ontario, on the other hand, the proportion went from 19.1% to 25.9%; a 6.8 percentage point increase.

That’s absolutely massive. The gap between Ontario and the rest of Canada was 5.8 percentage points in 2001; by 2016 it had risen to 11.1 percentage points.

But we haven’t applied any controls for ethnicity or culture. To examine that question, Prof. Sketerud examined whether or not the individual self-reported as being of “non-European” ethnicity.

As it turns out, it does not explain why Ontario’s proportion of 25–34-year-olds living with their parents rose much faster than in the rest of the country.

Here are the results:

Proportion of 25–34-Year-Olds Living with a Parent in Ontario and Non-Ontario Metros by Self-Reported Ethnicity

For our two non-Ontario groups, the “living with a parent” proportion grew by 1.9 percentage points from 2001 to 2016. For our two Ontario groups, it rose by 6.3 percentage points!

Here is Prof. Skuterud’s explanation of what the results show:

Finally, if we control for the higher share of the Ontario population that identifies as non-European ethnicity, the 2001-to-2016 increase in the Ontario “advantage” if anything increases slightly (it is now 4.4 percentage points instead of 4.3 percentage points). How is this possible? The answer is simply that while the share of Ontario’s population that is non-European is higher, that difference hasn’t become bigger over time. In 2001, 25% of Ontario youth (25–34) in the selected CMAs were ethnic compared to 17% outside Ontario, an 8 percentage-point difference. But in 2016, those numbers stood at 33% in Ontario and 25%, still an 8 percentage-point difference.

Applying different controls for “culture” or “ethnicity” may yield slightly different results, but overall there does not seem to be much-supporting evidence supporting the theory why the percentage of 25–34-year-old Ontarians living with a parent rose sharply from 2001 to 2016 whereas the changes in other provinces were more modest.

Instead, we should consider that it may be caused by a lack of housing.

As promised, here is Prof. Skuterud’s write-up of the methodology and results:

— -

Here are the results I promised.

I pooled the 2001 and 2016 Census PUMFs and restricted the sample to individuals aged 25–34 living in one of Canada’s 20 biggest CMAs (HALIFAX, QUEBEC, MONTREAL, OTTAWA, OSHAWA, TORONTO, HAMILTON, ST.CATHARINES, KITCHENER-WATERLOO, LONDON, WINDSOR, SUDBURY, THUNDER BAY, WINNIPEG, REGINA, SASKATOON, CALGARY, EDMONTON, VANCOUVER, VICTORIA).

I then generated the following dummy variables:

1. withfolks=1 if the individual was living with their parent(s)
2. yr2016=1 if the data point is from 2016 (instead of 2001)
3. ontario=1 if the individual was living in Ontario
4. ontario2016=1 if [yr2016=1 and ontario=1]
5. ethnic=1 if the self-reported ethnicity of the individual was non-European

Here’s the first regression, which tells us that Canadian youth (25–34) were 4.2 percentage points more likely to be living with their folks in 2016 than in 2001.

This regression tells us that Ontario youth have always been more likely to live with their parents than youth outside Ontario, but this Ontario “advantage” was 5.8 percentage points in 2001 but increased to 10.1 percentage points (5.8 + 4.3) in 2016. This is consistent with what you found in the charts you posted yesterday.

Finally, if we control for the higher share of the Ontario population that identifies as non-European ethnicity, the 2001-to-2016 increase in the Ontario “advantage” if anything increases slightly (it is now 4.4 percentage points instead of 4.3 percentage points). How is this possible? The answer is simply that while the share of Ontario’s population that is non-European is higher, that difference hasn’t become bigger over time. In 2001, 25% of Ontario youth (25–34) in the selected CMAs were ethnic compared to 17% outside Ontario, an 8 percentage-point difference. But in 2016, those numbers stood at 33% in Ontario and 25%, still an 8 percentage-point difference.

So it doesn’t look like the result you found can be explained by the relative ethnicity of Ontario’s cities.