Supply Constraints and Ontario Housing Prices

A key feature of housing markets in Canada over the last decade is the sustained increase in prices particularly in larger urban centers such as Vancouver and Toronto. Data from Canada Mortgage and Housing (CMHC) on average MLS housing prices for Ontario as a whole shows that between 1990 and 2015 the increase was from $171,979 to $465,441 – a rise of 171 percent. In Toronto, the average MLS housing price was $254,890 in 1990 and $622,046 in 2015 and is expected to range from $$694,000 to $742,000 in 2017.

Of course, the drivers of housing prices in Ontario seem fairly straightforward to most of us. There has been an increase in Ontario's population since 1990 of nearly 36 percent, which should serve to boost housing demand. More importantly, interest rates have dropped and remain at historic lows, which make financing homes much easier even with the price increases. If economic factors are driving demand up faster than supply, then one would expect prices to increase.

One other factor may be responsible for rising prices in Ontario – and no, this is not a digression on foreign buyers. It has been suggested that over the last decade, the supply of new housing coming onto the market has become more constrained as a result of the regulation of residential development.   Some of these regulatory factors in Canadian housing prices rising include land-use regulation—approval timelines, timeline uncertainty, municipal council and community impacts, costs and fees, and the prevalence of rezoning. In the GTA, the Green Belt policies designed to combat urban sprawl have been blamed for rising housing with at least one estimate that the “sprawl-busting policies are responsible for a quarter to a third of the run-up in real-estate prices across the region.” Of course, it has also been countered that there are still ample supplies of serviced land and that developers are sitting on these supplies to drive prices up.

A recent story in the New York Times referred to a paper on US housing markets by Ed Glaeser and Joe Gyourko that tries to answer the question of how much housing prices might come down if land-use regulations were reduced. They conclude for example that the actual cost of a home in a place like San Francisco should cost about $281,000 to build but the actual price is closer to $800,000 with most of the difference being caused by regulation and environmental reviews that ultimately suppress housing supply. Needless to say, you should take a look at their paper to get a full appreciation for what they have done in constructing their estimates. Of course, one also has to look at what the benefits of any environmental or land use regulation may be to quality of life to get a fuller picture if any such price differences caused by such effects are worth it. That price difference may include the cost of fresher air, better traffic flows, and improvements in the quality of urban life from reducing sprawl and fostering density.

Still, this is interesting. If you look at housing starts in Ontario over the period 1990 to 2015 (see the Figures) annual freehold home starts went up from 28,104 in 1990 to peak at 62,305 in 2002 and then dropped quite steeply – even before the 2008-09 recession reaching a low of 28,460 in 2009 and then essentially flattening out. In 2015, they were at 33,729. Condo starts, on the other hand have seen ebbs and flows but generally have grown from 11,435 in 1990 and peaked at 34,633 then  dropping to 20,045 in 2014 and then rising to 27,911 in 2015. If one looks at total starts in Ontario, they go from 53,341 in 1990 to peak at 80,933 in 2003 then fall to 47,939 by 2009 and stand at 68,292 in 2015.

The picture for Toronto shows similar trends to Ontario as a whole but then the GTA is half of Ontario’s population, economy and employment. Freehold starts in Toronto peak in 2002 at 31,490 and by 2015 were half that number. Meanwhile, condo starts peak in Toronto in 2012 at 28,435 and have since moderated but were 23,862 in 2015.

 

Overall, the total annual supply of new housing coming onto the Ontario market has fallen since peaks reached in the first few years of the 21^st century. However, that fall is driven largely by the reduction in freehold homes and rental unit starts.  In Ontario in 2015, freehold home starts still outnumber condos but condo starts now outnumber freehold homes in Toronto. In both Toronto and Ontario, rental starts are down from the early 1990s. In both Ontario and Toronto, the drop in freehold home starts begins after 2002. It should be noted that the Ontario Greenbelt legislation for the Golden horseshoe area was proposed in the November 2003 Throne speech and was created by legislation that was passed in 2005. Correlation is not causation and many factors affect housing markets. Yet, one has to ask how much of the Glaeser/Gyourko story applies to Ontario when it comes to the effect of policies such as the Green Belt legislation.

Quick Update:

Well, Frank (see comments) had a good suggestion.  I should probably have plotted total starts also. So here is the chart.  Two points.  Additions to new supply have basically flattened out in Toronto since the peak in the early 2000s.  Ontario overall is down.  Much of the volatility seems to be driven by Toronto though deciding that by just eye-balling the graph is a pretty crude way to do it.

 

 

Comments

[Frank]

Hi Livio, would it help if the total housing starts is also plotted? A rough calculation of the total housing starts based on your plots appears to suggest a relatively flat trend of it. A document from CMHC also confirms the relatively flat trend of the total housing starts in Toronto (https://www.cmhc-schl.gc.ca/odpub/esub/64319/64319_2016_B02.pdf).

If these data patterns are right, then the implication would be that, the total housing starts have been roughly flat in the recent decade, while a larger portion of it is shifted from freehold to condo. This would reduce the supply issue to two questions: why the total housing starts does not go up? why builders shift from freehold to condo. These would be the specific context to think about the role of Greenbelt legislation. My guess at this point would be that, the Greenbelt legislation might increase the cost of freehold and hence be a factor for the shift from freehold to condo. But does the Greenbelt legislation also contribute to the flat trend of total housing starts in Toronto?

[Livio Di Matteo]

Thanks for the suggestion Frank. Have plotted total starts.

[Nick Rowe]

Livio: interesting. Two things especially surprise me:

1. The peak in 2001/2. Why then? Interest rates are lower now than then.
2. Just how volatile housing starts have been since 2006. I could see a big drop then rise immediately following the financial crisis, but the volatility has continued.

There's no obvious macro demand-side explanation for those facts. Supply-side regulation??

[Bob Smith]

If I were a political partisan - and I am - I might suggest that clearly electing the Harris Tories was the best thing that ever happened to the Ontario construction industry/new home market. :)

I don't know to what extent it explains the trend, but you would probably want to consider developments in the mortgage market both in terms of product development and regulatory change. I would be surprised, for example, if the big drop off in hew home starts in 2013 didn't have something to do with the elimination of the 40 year mortgage in the summer of 2012 (do housing starts have a seasonal component - if they happened in the spring, a change in June wouldn't show up until the next year). Similarly, the rise of the 40 year mortgage probably goes a long way to explain the run up in new home starts in Toronto from 2010 to 2013 (at which time, every one and their dog was buying condos - looking like nothing so much as the US in 2006). Those sorts of changes are going to have more dramatic effects on "affordability" (which probably drives buying decisions more than more macro factors like interest rates).

[Bob Smith]

Just as a follow up on the importance of mortgage regulations, I just saw this story about the Toronto rToronto al estate market : https://www.thestar.com/business/2017/02/15/toronto-housing-called-bubble-as-new-listings-fall-rapidly.html

Much of it is industry bumpf that I take with a grain of salt, but the comment that changes to mortgage regulations last year has driven people to the suburbs rings true. I've been tying to figure out why houses in my little bedroom community on the outskirts of Toronto has seen property prices skyrocket in the last year (e.g. 20-25%)' despite being a good hour or more from downtown by train or car. To give some sense, for what my fairly average detached home in a nice neighbourhood with a nice lot might have gone for two years, you can now buy a smaller townhouse, in a cramped neighbourhood, squished between a mall and a highway. This would explain that otherwise ridiculous result.

[Majromax]

@Nick Rowe:

It looks like the volatility of starts has been almost entirely in condominium starts. Toronto condominiums are predominantly high-rise (2011 NHS data), and new development there will have the strongest impacts from regulation. We could be seeing the impacts of either single large approvals (how many units are in a typical high-rise?) or of periodic zoning changes to allow for high-rise development.

[rsj]

@livio,

I was going to denounce your characterization of San Francisco's housing costs, but am forced to admit that the linked reference is a really interesting and well thought out paper.

The reason, initially, why I was going to argue against the 500K regulatory figure is that the fees are nowhere near that on a per unit basis. However, the larger regularly constraints, such as building limits, also have a "cost". This cost is in some sense impossible to estimate except by modeling supply and demand curves which the paper does. For example, what would a house cost if there was no zoning restrictions at all, and no EIR requirements? This is good stuff -- thanks!