Ontarians on the Move, 2021 Edition #16 — Housing demand and the growth in the number of households.

Sixteenth in (what I hope) will be a series on population growth, migration, and what’s going on with Ontario’s housing market. Previous piece: Ontarians on the Move, 2021 Edition #15 — New Ministry of Finance Population Projections

TL;DR version: While population growth is important, ultimately it is the growth in the number of households that ultimately determines the demand for housing. My estimate is that over the past four years we should have built an additional 100,000 homes to keep up with new household formation.

Much of the Ontarians on the Move series has looked at the relationship between population growth, the housing supply and how the twin forces of drive until you qualify and the musical chairs effect have reshaped Ontario in ways that are poorly understood.

Occasionally, someone on Twitter will comment that the demand for homes isn’t driven by population growth, it’s driven by the growth in the number of families. So population growth tells, at best, an incomplete story.

My response: You are correct.

I’ve been using population growth as a proxy for the growth in the number of households. We have good, relatively real-time data on Ontario’s population, we don’t have good relatively real-time data on the number of households.

The two are likely to be strongly correlated, making population growth a good proxy for the growth in the number of families. Though this need not necessarily be the case. Imagine a hypothetical scenario where, all of a sudden, the number of twins that were born rose dramatically. So families that were expecting one baby got two. A doubling of the number of babies is certainly population growth, but the number of households really hasn’t changed any. The size of each household has increased, but the number of households haven’t.

What would our analysis look like if we switched our focus from population growth to growth in the number of households?

We don’t have good real-time data on the number of households, so we’ll need to estimate it. Best place to start is at the CMHC, which has several series examining the number of households.

Let’s start with the series: Number of Households by Household Type: 1976–2036, which contains household data, broken down by age of the head of the household, for 1976–2011, along with projections from 2016–2036. Here’s the number of households in Ontario up to 2011:

Source: Number of Households by Household Type: 1976–2036. I’m not sure how they’re defining “head of the household” — I suspect it’s the age of the oldest member of the household.

I’m not sure how CMHC is defining household here, and different data sources will yield somewhat higher figures. Looking at the numbers, the numbers for 15–24 look quite low, which leads me to believe it’s counting university and college students as still being parts of their parents household, not forming their own. But I can’t seem to find a definition anywhere, so that might be completely off.

If we examine compare this to Ontario’s population in each year for each cohort, we see the household formation rate for the 15–24 and 25–34 cohorts is low… and falling. Despite this, a relatively constant 45% of people above the age of 14 are considered “head of the household” as the falling rates for younger people is offset by an aging population.

Source: Number of Households by Household Type: 1976–2036

A Parliamentary Budget Office study found similar numbers and trends when examining Canada as a whole:

Source: Household Formation and the Housing Stock

In short, if you want to calculate the number of households in Ontario, you could simply multiply the population of 15+ year olds by 45–46%, as this ratio has been near-constant since 1991. A more sophisticated approach would be to use the age ratios from 2011 and apply these consistently going forward. Not surprisingly, we see a similar trajectory in the estimated number of net new households as we do with population growth:

Source: Number of Households by Household Type: 1976–2036 and Population estimates, July 1, by economic region, 2016 boundaries

Comparing Housing Completions to the Growth in the Number of Households

Every household needs a place to call home, so we should expect that growth in the number of households should track closely to the growth in the number of homes. A December 2016 PBO report, Household Formation and the Housing Stock finds that historically, across Canada, this holds true:

Source: Household Formation and the Housing Stock

We should recognize that housing completions and household formation are linked in three ways:

  1. Household formation causes housing completions: Communities build homes to meet the growing number of families.
  2. Housing completion causes household formation: People need somewhere to live if they are to start a household. A lack of homes causes (forces) younger people to live with their parents or several roomates, suppressing household formation.
  3. Housing completion causes household attraction and retention: All else being equal, a limited supply of housing will lead to households leaving (and being forced elsewhere), and a growing supply of homes will lead to households moving in from other jurisdictions.

Using CMHC housing completions data for Ontario and matching it with the CMHC data on household formation, I generated an Ontario-specific version of the PBO graph. Not surprisingly, it looks quite similar:

Source: Number of Households by Household Type: 1976–2036, Population estimates, July 1, by economic region, 2016 boundaries and CMHC housing completions data. For housing completions, the yearly totals were taken from July 1 to June 30 for any given year, to match the time periods for population growth.

In the average year between 1976 and 2011, there were 67,000 households formed and 65,000 home completions. Not identical, but pretty close, and the two series move in tandem relatively well.

That has not held true, however, since the population started growing in 2015–16:

Source: Number of Households by Household Type: 1976–2036, Population estimates, July 1, by economic region, 2016 boundaries and CMHC housing completions data. For housing completions, the yearly totals were taken from July 1 to June 30 for any given year, to match the time periods for population growth.

Another way to represent this data is to add the 5-year period 2011–16, and the 4-year period 2016–20 to the previous chart:

Source: Number of Households by Household Type: 1976–2036, Population estimates, July 1, by economic region, 2016 boundaries and CMHC housing completions data. For housing completions, the yearly totals were taken from July 1 to June 30 for any given year, to match the time periods for population growth.

Over the last four years, on average, there have been 27,000 fewer homes built, each year, then family formation estimates would predict. Historically, that gap has only been 2,000. I need to write the next part in bold:

Over the past four years, Ontario built 100,000 fewer homes than it should have, given our population growth.

When the 2021 Census data is released sometime next year, I think we’ll find that the number of households formed in Ontario between 2016–2020 will be well below the 95,000 estimate from this method, which will close this gap. The reason is simple:

Housing completion causes household formation: People need somewhere to live if they are to start a household. A lack of homes causes (forces) younger people to live with their parents or several roomates, suppressing household formation.

That’s one of the bigger tragedies in all of this. While a shortage of housing in Ontario did cause all of the following:

  • Prices to rise in the GTA.
  • Families to move out of the GTA, thanks to drive until you qualify.
  • Prices to rise in other communities thanks to an influx of people who drove until they qualified.
  • Families to move out of those home communities to even further afield communities due to the musical chairs effect.
  • The environmental impact of families having to move to a different community than the one they work in.
  • Young families having less disposable income, due to higher monthly housing costs and the need to save up large downpayments.

We should not forget about all the households that were not formed, simply because young people did not have the money. Young people having to put their plans on pause, delaying moving out, getting married (including living common-law), and having kids, simply because they can’t afford a place to call home.

It’s inexcusable what we’re doing to our young, simply because we refuse to build more housing.

Senior Director, Smart Prosperity. Assistant Prof, Ivey Business School. Exhausted but happy Dad of 2 wonderful kids with autism. I used to do other stuff.