How Canada’s Housing Inflation Numbers & Policy Ignores New Homeowners

Increases in real housing prices across Canada have led the G7 for the past twenty years and have created a major issue around affordability for new homeowners and renters in Canada’s largest cities. However, between 2000 and 2021, Canada’s inflation rate hovered around the Bank of Canada’s 2% target, primarily led by imported disinflation in durable and semi-durable goods. While shelter costs reported under Canadian CPI are up 1.78x since 2002, the aggregate MLS Housing Price Index for Canada is up by 3.03x since 2005 and average rental costs since 2002 reported by the CMHC are up 2.1x. This substantial disparity between cost increases of new rentals and housing and CPI shelter costs brings two factors into question: 1) Does our current method of our inflation measurement actually inaccurately reflect housing costs, and 2) Should the role of central banks partially shift to target housing costs in light of failed government action?

First and foremost, the role of central banks like the Bank of Canada is to manage monetary policy to achieve low inflation stability. However, the reality is if we do not report inflation effectively for the largest cost component for most Canadians, central banks cannot actually undertake their primary function. When it comes to measuring housing costs, there is high complexity and varying methods that central banks use which results in inaccuracy in reflecting actual cost increases and basket weights, as seen above. While discussed below, for more detailed information on how housing CPI is calculated in Canada, please refer to Statistics Canada. Reporting inflation properly consists of identifying each component’s appropriate weighting or “basket size” in the CPI index and accurately measuring pricing changes over time periods. As each individual consumer spends different proportions of their post-tax or net income on different components of the index and weights of spending across different components change over time, the CPI is updated to reflect different weightings every few years, with the current weighting for shelter at 28.34% of the index based on Canada’s 2022 basket. Even this weighting can be considered controversial as it includes consumer expenditure samples from households with mortgage-free properties (34% of Canadian households) and households with fixed shelter costs through rent control or lower purchase prices and interest rates from lower-cost years since 1999. The CPI is also designed to account for costs of the average person or household, which skews away from the reality for Canadians with new leasing or homeownership costs.

To understand how these implications of CPI measurement are not accurately reflecting actual increases in housing costs, we have to look at how housing inflation is measured in Canada, which differs from methods used in other countries. The shelter component of the Canadian CPI consists of three sub-categories: 1) Rented Accommodation; 2) Owned Accommodation; and 3) Water, Fuel and Electricity. The primary issue of accuracy with Canada’s housing CPI lies within rented and owned accommodation.

Firstly, rented accommodation is measured through 8,000 rented sample households annually and uses statistical modelling that factors in the age of dwelling, number of bedrooms, type of building, and services (rent includes furniture, household appliances, utilities, etc.) to obtain an observed rent price. While this method eliminates qualitative differences impacting pricing in rental accommodations, the issue in Canada’s CPI rent lies in it factoring in both new and existing rental leases, with rent-controlled units skewing CPI rental costs downwards from the advertised price. Statistics Canada acknowledges this by stating that other data sources do not take into account quality differences as the CPI does above, and that advertised prices overestimate average rental costs for all renters because of large price increases when a new tenant takes over a unit. This brings the first instance of CPI housing reflecting housing costs for the average person rather than under market conditions.

Secondly, the component of owned accommodation is the largest measurement challenge to any central bank they seek to eliminate the implicit “investment value” of a house from continued increases in housing costs over economic cycles. The issue with this is that the “investment value” of a house reflects the buying condition for new and prospective homeowners. The way that Canada measures the CPI index for owned accommodation is through six components: 1) Mortgage Interest Cost; 2) Homeowner’s Replacement Cost; 3) Property Taxes and Other Special Charges; 4) Homeowner’s Home and Mortgage Insurance; 5) Homeowner’s Maintenance and Repairs; and 6) Other Owned Accommodation Expenses. Again, the issue of accuracy lies mainly within mortgage interest costs and homeowner’s replacement costs.

When it comes to mortgage interest costs, it applies complex methodologies to factor in mortgage rates and housing costs weighted over the past 25 years, the standard mortgage amortization period. As Canadians have mortgage rates that are typically renewed every five years or less, unlike the United States where 30-year fixed rates are available, Canada’s CPI component has greater exposure to prevailing interest rates that implicitly creates a double-edged inflation sword. If rates are high to combat inflation, mortgage servicing costs will typically rise even with the associated decrease in housing prices (higher sensitivity to rate changes than market value changes), while if rates are low, mortgage servicing costs will decrease. As a result, a low Bank of Canada policy rate creates lower mortgage servicing cost inflation, and vice versa.

When considering the homeowner’s replacement cost, this is supposed to reflect the costs of using owned accommodation that would traditionally use the cost of purchasing a home. However, Statistics Canada instead measures a theoretical “replacement cost” or depreciation component, the hypothetical amount needed to replace their housing based on expected home value and external housing prices. This is then multiplied by a “house / property” ratio to obtain a house value estimate to which a 1.5% depreciation rate is applied. Statistics Canada has acknowledged that owned accommodation is controversial and that it is under-estimated in the CPI, and states that the owned accommodation is, “not designed to be an indicator of housing price inflation nor housing affordability,” but rather to, “quantify the rise in price of owning and occupying a dwelling.”

Taking a step back to our first original question, after looking at the way we measure housing CPI in Canada, our current method of inflation measurement might accurately reflect costs for the average Canadian but not for the Canadian renting a new property or buying a new home today. This creates a lag in reflecting today’s market costs into inflation data, misleading Canadians about the true change of living costs. The Bank of Canada then uses this inflation data to make policy decisions, but these policy decisions impact prospective and new renting or homeowning Canadians who are affected by today’s actual housing costs, not the Canadians who have rent-controlled leases, mortgage-free homes, or locked-in mortgage rates. While certainly not the only contributing factor, the understated increase in housing costs for Canadians in the CPI led to a lack of tighter monetary policy in the 21st century that would have reduced the rate of housing cost increases.  In the future, this will actually lead to an overstated increase in housing costs that might amount to overaggressive monetary policy. If the Bank of Canada’s goal is to keep inflation low and stable, with the ultimate goal of protecting Canadians’ purchasing power, it has undoubtedly failed in this aspect by being a contributor to housing unaffordability. To restore monetary policy decision-making in Canada to truly serve the best interests of Canadians and our economy, the Bank of Canada might not need a radical reform of measuring housing CPI, but ultimately should focus on new homeownership costs over average homeownership costs when making policy decisions.