Solving puzzles in the Canadian housing market: foreign ownership and de-coupling in Toronto and Vancouver

Abstract

Using new data from the Canadian Housing Statistics Program (CHSP), this paper provides a basis for an integrated account of the Canadian housing market in the last two decades. It shows how the housing markets in Vancouver and, to a lesser extent, Toronto have become de-coupled from local incomes due to significant flows of foreign capital. Once this dynamic is appreciated, certain puzzling elements of the Canadian market become intelligible. The analysis points to possible policy solutions to intense housing affordability challenges. It also provides evidence of widespread tax avoidance in certain urban areas. Furthermore, it documents a methodology that researchers in other national contexts, who may lack government generated data on non-resident or foreign ownership, may adopt to infer the presence of de-coupling through foreign ownership.

Keywords:

• Canadian housing market
• foreign ownership
• de-coupling
• housing affordability

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Indeed, as can be seen visually in Figure 2, if Toronto and Vancouver and their ‘spill-over’ markets (Victoria and Hamilton) are removed from the scatterplot, the correlation between median incomes and house prices strengthens dramatically (R² = 0.46).

2 That is, they selected the alpha that most accurately predicted house prices in that two-year window, and then extrapolated that alpha to the entire period (2001–2019). See the discussion below about the possible problems with this approach.

3 It should be noted that even if prices align with the ‘attainable’ house price arrived at through the SBC approach, this does not mean that housing prices are in a sustainable long-term position. Given that the SBC approach implies that most borrowers are borrowing to their limit, and have taken on larger mortgages as interest rates have fallen, then these borrowers are in significant danger if interest rates rise. For a discussion of the implications of falling interest rates, see e.g. Schembri, 2015; Andrle & Plasil, 2019.

4 This was calculated using median renter incomes by CMA (2017) relative to the CMHC’s index of rental rates for two-bedroom apartments in each CMA (2018). The rate of pre-tax income that would need to be devoted to (annualized) rent ranged from around 27 percent in Calgary and Edmonton to 40 percent in Vancouver and Victoria and 53 percent in Toronto. This range of 0.26 for rental markets compares to a range of 0.52 for the alphas in the ownership markets. Other metrics for rent prices would lead to slightly different, but likely consistent, figures (ordering-wise).

5 Granted, misrepresentation of income can occur widely among those earning domestically. The point here, however, is that it is unlikely that it will be much more prominent in some parts of a metropolitan area than others in the absence of concentrated foreign ownership. High rates of ‘satellite families’ may show up in aggregate statistics, by contrast.

6 See Jumana Al-Tawil, ‘Homeownership, income, and residential property values’, 2019, Statistics Canada.

7 This analysis excludes very small municipalities in Metro Vancouver, such as Belcarra and Lions Bay, which do not have enough data to support the kind of analysis conducted in section 5 (for privacy reasons, Statistics Canada has repressed certain data). UBC, or Metro Vancouver A, is also excluded, even though it is large enough to be included in some of the analysis in section 5. The data from UBC is consistent with the analysis below, even though the degree of de-coupling observed in UBC dwarfs that found in the other municipalities.

8 For example, in London, Ontario (CMA), if we compare the median assessed value of SFD properties owned by working-age households to the rate of ‘low income ownership’ among that same segment across the eight sub-regions (or CSDs), the correlation is strongly negative (r = −0.86): i.e. as house prices go up, the rate of ‘low income ownership’ declines steadily. In Vancouver, the correlation is strongly positive (r = 0.75).

9 Data collected through the Speculation and Vacancy Tax in B.C. reinforces this point. Under this legislation, homeowners are asked to declare whether they are satellite families (more than half of household income is earned and taxed abroad) or foreign owners (non-permanent residents or non-citizens). The share of self-declared satellite families correlates at r = 0.93 with the (CHSP) rate of non-resident ownership across the large municipalities.