Accuracy of Canada’s housing data under scrutiny
Canada’s surprisingly strong real estate market is leading to heightened scrutiny of the data used to assess sales.
The main source of information about residential sales and prices is the Canadian Real Estate Association, which represents more than 100,000 real estate agents across the country. On or around the 15th day of each month, after obtaining and working with numbers from local real estate boards, CREA releases information about the previous month’s sales levels.
The numbers are getting more attention amid debate about just how inflated the market may be. Nearly all economists agreed that house prices were too high when Finance Minister Jim Flaherty stepped in and tightened mortgage insurance rules in July, 2012. Sales dropped, diminishing talk of any potential bubble as the market slumped for nearly a year. In recent months, it has bounced back, and topped economists’ expectations.
It appears that the number of houses changing hands this year will exceed last year’s total. But most experts say the strength is unlikely to last long, and the market should cool a bit as higher mortgage rates trickle through the system.
As people try to make sense of the market’s twists and turns, some are raising questions about the accuracy of CREA’s statistics. One of the main critics is former Member of Parliament and blogger Garth Turner, who has recently been calling CREA’s numbers into question on his blog. He alleged, among other things, that the numbers are being distorted by real estate agents listing one house with two or three local real estate boards (so more buyers will see the listing).
Mr. Turner’s skepticism caught the attention of other analysts. “I must confess that it was a blog by Garth Turner that kind of raised that issue,” said Douglas Porter, Bank of Montreal’s chief economist. “I found it fascinating.”
“You do have to look at [the numbers] through the lens of where they’re coming from but just like any other economic data, I don’t think any single number should be treated as the final gospel,” Mr. Porter said. “There is a margin for error around every economic statistic, and we take it with a grain of salt, but I think the CREA numbers do a good job of capturing the trend.”
CREA chief economist Gregory Klump said that just because a listing appears more than once online, does not mean it is included more than once in the statistics. “The fact is, Realtor.ca is not the basis on which we do our statistics,” he said, referring to CREA’s website. “Just because something is duplicately listed on Realtor.ca doesn’t mean that it’s duplicately counted in sales and in active listings.”
Mr. Porter noted that unless dual listings were a new or growing issue, there would not likely be a great impact on comparisons, such as how this September’s house sales compared with last September’s. “We’re just trying to get a sense of whether the market is strengthening or weakening or stable,” he said.
Questions have also been raised about revisions local real estate boards make to their numbers, which are not immediately captured in CREA’s national statistics. For example, the Toronto Real Estate Board counts a sale when an unconditional offer is accepted or when non-title conditions are removed, rather than when the sale closes, Mr. Klump said in an e-mail. If a deal later falls through, the board revises its numbers for the month in question.
CREA does not revise its numbers each month, but rather on a schedule that it says minimizes distortions to comparisons. “It may be argued that TREB’s methodology produces a more accurate picture of history once all revisions have been incorporated, and it may also be argued that CREA’s methodology produces more historically comparable statistics for the current month being reported on,” Mr. Klump said.
Confusion has also arisen from seasonal adjustments CREA makes to its monthly data, which Mr. Klump noted it does using methods from Statistics Canada.
Source: The Globe and Mail