Tuesday, 4 December 2007

On housing prices in the Greater Toronto Area

As my mortgage payments were slowly but steadily growing along with interest-rate hikes during the last 2.5 years (see my previous post), the market value of our residential property in a quiet suburb north of Toronto was persistently climbing as well. Of course, it is personally very comforting to see a value of your house to go up by 30% over a couple of years. Periodically, a couple of real estate agents are calling to deliver us the exciting news about new records set by prices on similar houses in our neighborhood. I feel happy and express my sincere excitement every time I talk to them, but politely decline an offer to put our house on sale and find something bigger, better, and (of course) "just a bit" pricier ("A bit" is "only" S50-100K). No thanks no. I will pay off this one first and then we will see. Moreover, I see some clouds on the horizon of the Canadian housing market, which are about to bring a refreshing rain of falling housing prices (speaking metaphorically, the thunderstorm is already raging across the border).

Do I want housing prices to go down? Yes. Am I nuts telling such things? I do not think so. It looks like we are about to see the long-term cycle taking its long overdue turn towards the sizable price correction, and I think that this correction is absolutely necessary because, as an economist, I dislike pricing bubbles.

Market bubbles arise mostly due to speculative and often artificially infused demand; are almost always created thorugh the excessive use of leverage; contribute to the overall inflation rate; distort and limit consumption and savings choices of many households; increase personal debt burden in the economy and thus eventually hurt the overall economic growth; and stand behind many cases of personal bankruptcies. Honestly, don't you think that prices of houses in Toronto and its suburbs look a bit too high lately? I definitely do not think that these prices are justified by fundamentals. I bet that those who are still shopping around for a house can really tell that.

I was recently taking part in a project, which analyzed and compared price elasticities of local taxes on residential and commercial property across 25 GTA municipalities. The charts below plot some of the GTA and Toronto data related to our project. They show time series of equalized residential assessment per capita (which basically is used as a measure of the average property value) and the corresponding tax rates on residential property (the upper and lower tier rates only, the education portion not included).

I read somewhere that historically a full long-term cycle of residential property prices in Canada lasts approximately 18 years. By looking at the GTA residential property data, we can see that the new cycle, which took off in 1989-1990 (I arbitrarily count from the year during which the prices peaked and reversed), is about to begin in 2008-2009 if the long-term patterns are to be respected. In fact, we are kinda a bit late already in Canada with a start of a pricing correction. The U.S. housing market has already taken a dive, and I do not see any compelling reason why we in Canada should enjoy our own housing bubble however.

By looking at the assessment trend line, we can observe how the market accelerated in 1986-1989 before sharply reversing in 1990. The slump continued for some time, before the market picked up in 1997 and again has shown strong double-digit growth in most GTA municipalities since 2003.

Also, looking at the charts, we can see a strong, statistically significant negative relationship between property values and property tax rates. One of the measures that governments use in order to determine what property tax rates to set is price elasticity, which measures the response in property values to a marginal increase in the tax rate (dB/dt*T/B, where B is the tax base and t is the tax rate). The government is not interested in or sincerely cares about changes in property values per se; rather, their concern is how their total property tax revenue (tax base*effective tax rate, or B(t)*t) will be affected by changes in the tax rate. A Laffer curve, or a revenue hills curve, which can be contructed using price elasticity estimates, is used to show the relationship between tax revenues and corresponding tax rates, including a revenue-maximizing tax rate.

Speaking in terms of price elasticities of property values, we can define the revenue-maximizing tax rate as the one under which the price elasticity is equal to -1 (1 in absolute terms). If it is below -1 (that is, we have low elasticity), then the tax revenue can be increased by a further tax hike. If it is above -1 (that is, we have high elasticity), further tax increases will drive down property values too much and as a result the overall government property-tax revenues will fall. For obvious reasons, governments prefer finding themselves on a low-elasticity portion of the Laffer curve where they can still increase a tax rate a bit more to squeeze extra revenue dollars from tax payers.

Our analysis on the GTA residential property data has indicated that the price elasticity estimates among GTA municipalities and regions are currently way below -1 for pretty much every municipality in the Greater Toronto Area and have been staying low at least over the last ten years. From the elasticity formula above, we can figure out that elasticity estimates can be low when residential tax rates are too low and when tax-base values are overinflated by a price bubble. Indeed, we can observe from the charts that most GTA municipalities were decreasing their effective tax rates or at least kept them flat during the last several years when the speculative growth in housing prices accelerated noticeably and when the buy-buy-buy-a-house chant was all over the place. Without a doubt, such tax policy further contributed to the excessive growth in housing prices.

Am I advocating for higher property taxes? No (as a home owner). Should the local governments decrease taxes when the real estate market is already overheated? Defintely no since they fuel the speculative demand and contribute to a bubble's development. Should the residential real-estate prices be periodically corrected by the market? Yes, absolutely. Is there a role for local governments in control of property prices? Maybe yes in theory, but I doubt that it can be justified in reality for a variety of objective reasons.

Is there any other way to keep housing prices in check? The minimum-leverage and minimum-credit-rating requirements in mortgage financing can be one of the effective natural controls. The U.S. government are learning these simple things the hard way right now, or are they? Looks like they do not care that much. The Canadian government should if they do not want to be on the same page with the socially indifferent U.S. "civil" servants.


Anonymous said...

Very good post!

By chance do you have any data on the the non residential class? Toronto seems to be in an interesting position with regards to its 905 neighbours. Some vague data that I came across demonstrates that Toronto's non residential tax rates have been fully capitalized in to values. As such Toronto's non res. tax rate through twice as high, produces no additional revenue. Clearly suggesting that it is to the far right on the Laffer curve.

BEIT said...


The available data on the commercial/industrial class do not give unambiguous conclusions, but the analysis does suggest that the Toronto non-residential property rates may lie above the revenue maximizing point while rates set by the other GTA municipalities are still on the unelastic portion of the Laffer curve.

To make statistically reliable conclusions from the residential and non-residential data, however, it would be nice to have historical numbers collected over at least 3 long-term cycles (each long-term cycle is approximately 18 years), which is not quite possible at this point.