The main housing component of the CPI, the so-called owned accommodation component, has been pulling inflation down since the Canadian economy started to turn the corner in the fall of 2009. A surprising result given that housing prices have been growing at about 6 per cent over the same period.Galit Rodan/The Globe and Mail
The dramatic rise in housing prices over the last decade has been a major concern for Canadians. And with good reasons: according to any of the usual metrics – housing prices relative to household income for instance – the housing market appears out of whack. The federal government has taken notice too and has been tweaking rules about mortgage insurance to tighten lending standards, and dampen the housing market.
But what about the Bank of Canada, which after all, all but sets the price of mortgages? It has been mostly sitting on the sideline, keeping its short-term interest rate at 1 per cent for the last two years. The Bank has many good reasons for not raising rates: there are still some very dark clouds over the world economy. And raising rates also risks bringing the Canadian dollar to new highs, which is already hurting our exports.
Most importantly, inflation appears under control, and is currently below the Bank's target of 2 per cent as measured by the Consumer Price Index (CPI). But many people might be surprised to learn that housing prices don't affect the CPI much. In fact, the main housing component of the CPI, the so-called owned accommodation component, has been pulling inflation down since the Canadian economy started to turn the corner in the fall of 2009. A surprising result, to say the least, given that housing prices have been growing at about 6 per cent over the same period.
So, is there something wrong with the CPI? The short answer is no. Rather, this surprising result can be explained by the fact that the CPI cannot perfectly meet the needs of all users. While it can be used as a general measure of consumer price changes it must also meet the needs of those seeking a cost-of-living indicator – to adjust the value of government transfers such as Old Age Security for instance. Consider this: with interest rates at historically low levels, the cost of owning a house has grown more slowly than actual housing prices would suggest. That's an important reason why the housing component of the CPI has been going down in recent years, instead of up.
The CPI was never designed explicitly as an inflation indicator to be used for conducting monetary policy, even though it has officially served that purpose since 1991. Take the fact that mortgage rates are low and how that might impact the stance of monetary policy: the Bank of Canada is not raising rates partly because the CPI is low; and the CPI is low partly because rates are low. This sort of circular thinking clearly shows one of the drawbacks of using the CPI for conducting monetary policy.
What we need is a new inflation indicator that is designed solely for that purpose, while keeping the current CPI for other purposes that it is well designed to accomplish, like to adjust government transfers. In a recent study, I recommend an inflation indicator based on a so-called Net Purchases approach for owner-occupied housing. This approach has been adopted by countries such as Australia and New Zealand, and has been gaining traction elsewhere.
This approach measures changes in market prices – actual transaction prices – for owned accommodation; it treats the purchase of a house exactly like any other purchase and doesn't consider how the purchase of the house is financed and therefore does not include an element related to mortgage rates. Adopting this approach would have an important impact on inflation – with differences in inflation readings of up to a full percentage point – and more importantly on monetary policy: it would prompt the Bank of Canada to tighten interest rates sooner and to a greater extent when housing prices are rising quickly, and to loosen them when housing prices are dropping.
The main argument against my proposal is that housing is not a consumption good, but an asset, some argue. But the inflation indicator I propose is only concerned with houses that are occupied by their owners. The decision to buy a house in which to live is primarily a consumption decision; in fact, buying a house is the most significant consumption decision most households will ever make. There may be an expectation on the part of the buyer that the home will gain in value, or not, but such considerations are secondary. Assets on the other hand can be described as deferred consumption: you give up consumption now by investing in an asset, in the hope of having more consumption later. Owner-occupied housing does not fit the bill, and deserves to be treated like other goods in an inflation indicator.
Constructing an inflation indicator that better reflects changes in housing prices, and making it an official inflation indicator, would help ensure monetary policy is appropriately responsive to housing prices, which could help improve Canada's overall economic prospects.
Philippe Bergevin is a Senior Policy Analyst at the C.D. Howe Institute and author of a recent report titledHousing Bubbles and the Consumer Price Index: A Proposal for a Better Inflation Indicator.