Policy Papers
Tall and Sprawl: The distribution of housing stock growth in the Greater Toronto Area, 2016 – 2021
Analyzing changes in housing stock in the GTA between 2016 and 2021 (the two most recent census years) the authors note that the vast majority of growth of housing stock is within a 3 km radius of Union Station, or on the urban fringe of the GTA. This “donut” pattern of growth has several policy implications.
Introduction
Ontario faces a severe shortage of housing, negatively impacting housing availability and affordability for homebuyers and renters alike. Though relatively recent in some regions, this phenomenon has long afflicted the Greater Toronto Area (GTA), Ontario’s (and Canada’s) largest metropolitan area, warranting closer investigation of the region’s growth patterns. The following analysis examines the GTA’s housing stock growth between 2016 and 2021, the two most recent census years.
The geographic distribution of housing stock growth over this period was highly concentrated in the GTA’s core—within a three-kilometre radius of Union Station—and on the urban fringe. While it is unsurprising that relatively more homebuilding is occurring in the regional core than in outlying areas, it is surprising that housing stock growth falls steeply in the neighbourhoods immediately surrounding the core, only to reappear in pockets farther away.
This “donut” pattern of growth has several policy implications as the region attempts to overcome a generational housing shortage. First, slow-growing central neighbourhoods and communities need to accommodate more housing, given the GTA’s severe shortage and demonstrated demand. Second, the steep decline in housing stock growth immediately outside the regional core (Downtown Toronto) is a strong indication of policy barriers to homebuilding in otherwise highly desirable neighbourhoods, rather than implying weaker demand. Third, faster-growing pockets in more distant communities are partially reflective of deliberate efforts to develop additional nodes of transport and commerce beyond the regional core, but could also be the result of insufficient homebuilding in desirable central neighbourhoods.
Faster housing stock growth in the regional core and in select outlying districts plays an important role in closing the gap between demand and supply, but given the magnitude of unmet need for housing in Canada’s largest urban region and primary port of entry, all governments with influence on housing development in the GTA have a role to play in enabling the construction of more homes across more neighbourhoods and communities.
Growth pressures and housing shortage in the GTA
Ontario is experiencing enormous growth pressures, centered around the Greater Toronto Area. Ontario added 775,448 residents between 2016 and 2021, 274,185 of whom were added to the Toronto census metropolitan area[1] (the GTA), and a further 88,899 of whom were added to the three census metropolitan areas immediately adjacent to the GTA (Hamilton, Oshawa and Barrie).
Further, the GTA-wide rental vacancy rate (a measure of rental unit availability) was 1.7% for purpose-built rental units and 1.1% for rented condominiums in 2022, well below Canada’s three-decade average of 3.2%.[2] Low rental vacancy rates are a primary indicator of housing shortages, as they result in upward pressure on rents.[3]
Indeed, mounting evidence shows that Ontario has a significant structural shortage of housing. Scotiabank Economics estimated in 2022 that Ontario would need to add 650,000 housing units to reach the same per capita level of housing as other provinces, or 1.2 million to reach the G7 average.[4] It further estimated that the GTA had the lowest number of housing units per capita among Canada’s 10 largest census metropolitan areas.[5]
Beyond Ontario’s current shortage, there is evidence that it is worsening with time. One housing unit was built for every one-to-two additional Ontarians during the 1970s. This grew to one unit being built for every two-to-four additional Ontarians during the 2010s, and has since grown to one unit being built for every six additional residents in 2022.[6]
In short, all of Ontario, but especially the GTA, has a clear shortage of housing. Further, this shortage has worsened over time, pushing home prices and rents higher.
Housing supply and distribution in the GTA, 2016 to 2021
Beyond its clear consequences for affordability, the GTA’s worsening housing shortage raises questions about the region’s growth patterns, including the geographic distribution of housing stock growth. If the GTA is not building enough homes, which districts are driving this tendency? Similarly, which districts are adding more homes, and why?
To answer these questions, this section examines the GTA’s housing stock growth at the census tract level. Census tracts (CTs) are neighbourhood-sized urban geographies, designated by Statistics Canada within census metropolitan areas and census agglomerations.[7]
Figure 1 features a map of the GTA, divided into 1,227 census tracts. Each tract’s colour reflects the numeric value of its housing stock change (i.e. how many homes were added or subtracted) between 2016 and 2021. Shades of red reflect negative housing stock growth, while shades of blue reflect positive housing stock growth. Deeper shades indicate greater changes in housing stock, and white tracts exhibited little-to-no growth, as summarized in the map’s legend.
Figure 1: Map of housing stock growth, by census tract in the Toronto CMA (2016 to 2021)
Sources: Statistics Canada (2022c; 2022d), authors’ calculations
Census tracts in Downtown Toronto—the region’s core, within three kilometres of Union Station (an area bounded by Trinity Bellwoods Park, Yonge Street, the Don Valley, and Lake Ontario)—appear in dark blue, indicating stronger housing stock growth. Farther away from the core, tracts turn to lighter shades of blue and red, indicating slower, or even negative housing stock growth. Beyond this ring of “inner” suburbs including much of Toronto, Mississauga, and Oakville, a second band of blue appears in surrounding “outer” suburbs, such as Vaughan, Markham, Brampton, Milton and Pickering (all beyond the 34-kilomtre mark). This is followed by two less obvious bands of slower, then faster housing stock growth: one including parts of King Township, Caledon, Whitchurch-Stouffville and Uxbridge; another including East Gwillimbury, Bradford West Gwillimbury, and New Tecumseth.
Figures 2a and 2b present the same data as in figure 1, but as bar charts. The horizontal axis in both figures represents distance, in one-kilometre increments, from Union Station (a proxy for the GTA’s core). The vertical axis in figure 2a shows the total 2016 to 2021 growth in housing units across all census tracts located within each one-kilometre band. In figure 2b, the vertical axis shows the average growth in housing units among census tracts in each one-kilometre band.
Figure 2a: Housing stock growth by distance from Union Station (2016 to 2021)
Sources: Statistics Canada (2022c; 2022d), authors’ calculations
Figure 2b: Average census tract-level housing stock growth, by distance from Union Station (2016 to 2021)
Sources: Statistics Canada (2022c; 2022d), authors’ calculations
As observed in figure 1, figures 2a and 2b show that housing stock growth between 2016 and 2021 was most pronounced in the core, then less pronounced in the areas immediately surrounding the core, only to reappear in increments farther away. Indeed, 34,495 housing units were added in census tracts located within three kilometres from Union Station—21.7% of all GTA housing stock growth between 2016 and 2021. The following three one-kilometre bands, comprising areas between three and six kilometres from Union Station, added 5,951 housing units over the same period, or 3.7% of GTA housing stock growth.
Not only did Toronto’s core grow faster than its immediate surroundings (and indeed any other segment of the GTA), but it did so within a significantly smaller land area. 47 census tracts are located within three kilometres of Union Station, covering a land area of 21 square kilometres. The following three-kilometre segment comprises 84 census tracts covering a land area of 48.4 square kilometres. In other words, central Toronto added almost six times as many units as its surrounding neighbourhoods within less than half the land area. In other words, most of the growth has been concentrated in the old City of Toronto, and newer parts of the City like Scarborough have seen almost no growth at all.
Other fast-growing segments of the GTA were located in the 22 to 23-kilometre band, the two bands between 34 and 36 kilometres, and the 38 to 39-kilometre band – all outside of the City of Toronto. The first (22 to 23 kilometres) added 9,066 housing units between 2016 and 2021—5.7% of total GTA housing stock growth—spread across 35 census tracts located in Mississauga, Vaughan, Richmond Hill, Markham, and Scarborough. Notable districts within this band include the rapidly developing neighbourhoods surrounding the Square One shopping centre in Mississauga, and Downtown Markham.
The two bands located between 34 and 36 kilometres from Union Station added 13,024 housing units between 2016 and 2021, representing 8.2% of the GTA’s housing stock growth. Most of this growth occurred in outlying census tracts located at the urban fringes of Oakville, Brampton, Richmond Hill, and Pickering. Similarly, the 38 to 39-kilometre band’s 8,933 units (5.6% of GTA growth) were primarily located in census tracts at the urban fringes of Brampton, Caledon, and Ajax.
On average, census tracts beyond the 34-kilometre mark grew more quickly than those between 3 and 34 kilometres from Union Station (both distances are represented in Figure 1). This is in part because the 34-kilometre mark roughly captures the extent of the GTA’s contiguous urban area, meaning that much of the growth observed beyond this line is occurring in new subdivisions built on previously undeveloped (also known as “greenfield”) land. Further, areas beyond the 34-kilometre mark include fast-growing communities along or near Highway 400, to the GTA’s north, such as Newmarket, Aurora, New Tecumseth, Bradford West Gwillimbury and East Gwillimbury. Milton and Whitchurch-Stouffville, to the west and northeast of the contiguous GTA, respectively, also featured fast-growing census tracts comprising primarily greenfield development.
In very broad terms, the census data explored in this section suggest that the GTA’s housing stock growth is driven primarily by development within Toronto’s core, as well as the creation of new neighbourhoods near the region’s urban fringe. In other words, there’s a donut pattern with fast growth in the center and on the fringe.
In fact, the 270 census tracts located either within three kilometres, or more than 34 kilometres away from Union Station were host to 51.6% of the GTA’s housing stock growth between 2016 and 2021. The remaining growth occurred in the 957 census tracts located between three and 34 kilometres from Union Station. In other words, slightly more than half of the GTA’s housing stock growth occurred in the fastest-growing fifth (22.0%) of GTA neighbourhoods, while the majority of neighbourhoods experienced slower growth or no growth at all.
Causes and consequences
The GTA’s uneven growth pattern is not a natural, market driven phenomenon. It was driven in large part by government policy decisions – most notably, zoning.
The GTA is a relatively new metropolitan area. Only one-third of housing units were built before the 1960s.[8] This is notable since the City of Toronto’s first zoning bylaw came into effect in 1952.[9] Zoning restricted (and continues to restrict) density and land uses throughout the region, ensuring that most of the land zoned for residential uses only allowed single-detached houses while small pockets allowed for multi-family housing. This is why wards that were primarily built out after the 1960s tend to have a “tall and sprawl” development pattern, with oceans of detached housing surrounding islands of high-rises, while older downtown-adjacent neighbourhoods such as Beaches-East York tend to have more diverse housing types, with fewer detached houses.[10]
Older North American cities that were largely built before zoning tend to have less jagged development patterns. New York City is a classic example. Lower Manhattan has extremely high density – much like the areas surrounding Toronto’s Union Station. But unlike in Toronto, density does not drop sharply beyond the downtown core. Indeed, more residential parts of Manhattan are characterized by walk-up apartments and “brownstone” row housing than single-detached homes. Density decreases beyond Lower Manhattan, but the outer boroughs (as well as adjacent New Jersey cities) are for the most part very dense by North American standards.[11] While it is not unusual for Toronto subway stations to be surrounded by single-detached homes, this is less the case in New York City.
There are two major consequences of this development pattern in the GTA. The first is that it has limited the supply of housing in the neighbourhoods between three and 34 kilometres from Union Station. This places more pressure on more distant communities to grow than would be the case if more homes could be built closer to the regional core. Toronto is constrained both by geography and by policy. The surface area of the city is largely developed, and the Greenbelt makes outward growth challenging.
The second consequence is that there is less housing near public transit stations. This not only means there are fewer potential transit riders for those stations, but that the price of housing near those stations is higher due to scarcity, and the types of dwellings. People who are reliant on public transportation (as opposed to people who make the lifestyle choice to live in transit-friendly neighbourhoods), are unlikely to be able to afford single-detached homes. That means the people who need transit the most are the least able to live by it. Further, the significant public expenditures involved in building rapid transit infrastructure serves less of the tax base than would be the case if more residents were allowed to live closer to transit stations.
Of course, these conclusions, and the data used for the analysis in this report, are backwards looking. That is, they analyse past growth patterns and policy consequences. Housing policy in the GTA and Ontario are in flux. Below we list some policy considerations for GTA governments moving forward.
Policy considerations
At least three policy considerations stem from the findings presented in the previous sections. First, slow-growing central neighbourhoods and communities can accommodate more housing, given the GTA’s severe shortage and demonstrated demand. Second, the steep decline in housing stock growth immediately outside the regional core (Downtown Toronto) is a strong indication of policy barriers to homebuilding in otherwise highly desirable neighbourhoods, rather than implying weaker demand. Third, faster-growing pockets in more distant communities are partially reflective of deliberate efforts to develop additional nodes of transport and commerce beyond the regional core, but could also be the result of insufficient homebuilding in desirable central neighbourhoods.
1.Further development is required in slow-growing neighbourhoods
The most obvious lesson stemming from the data presented in the previous section is that without a greater share of neighbourhoods shouldering more of the GTA’s housing stock growth, the region is unlikely to close the wide, growing gap between housing demand and housing supply. Indeed, slightly over half of the region’s growth occurred in the minority of census tracts located less than three kilometres or more than 34 kilometres from Union Station.
Policy efforts by all governments in the GTA should therefore focus efforts on increasing both the share and absolute amount of housing stock growth occurring in all neighbourhoods, but especially in the large majority (78.0%) of census tracts located between three and 34 kilometres from Union Station. Not only are these areas not growing as fast as areas within or far away from the core, but well over a third of census tracts (38.7%, or 370) located in this area exhibited a net loss in housing stock.
Several policies aimed at “unlocking” more housing in neighbourhoods beyond the regional core have been proposed or adopted in recent years. Notably, the government of Ontario’s Bill 23, the More Homes Built Faster Act, allows up to three units on virtually all residential lots without the need to change zoning bylaws,[12] while the City of Toronto approved a bylaw enabling the construction of up to four housing units in residential zones citywide.[13]
Both policies share the aim of removing some of the barriers facing residential development in high-demand, low-growth neighbourhoods such as those surrounding the city core. For all their promise however, they fall well short of the recommendations proposed by the Ontario Housing Affordability Task Force, whose report was published in early 2022. In particular, the Task Force recommended allowing six to 11 storeys with no minimum parking requirements on any streets utilized by public transit.[14] Rather than being presented as a fait accompli, the zoning policies implemented by the Ontario provincial government and the City of Toronto should serve as starting points for further efforts to relax land-use restrictions and accelerate planning processes.
2. Slow housing stock growth immediately outside the core suggests policy constraints, not market factors
To an extent, the findings summarized in the previous section are not entirely unexpected. Indeed, the broad literature on urban economics posits that demand for urban space falls alongside distance from a metropolitan core,[15] leading to denser urban development in the core and gradually falling density beyond. In this regard, it is unsurprising that many of the census tracts closest to the GTA’s core exhibited the most housing stock growth between 2016 and 2021, and that relatively less growth occurred outside the core.
This study’s most significant finding, therefore, is not that a disproportionate share of growth occurred in the GTA’s core, but rather that the rate of housing stock growth falls steeply in the areas immediately beyond the core. Instead of declining smoothly with each subsequent one-kilometre band, as traditional economic models would suggest,[16] housing stock growth falls sharply at the three-kilometre mark, and remains relatively low until the 22nd kilometre, and again until the 34th kilometre. When accompanied by demand indicators such as home prices and rents, what quickly becomes clear is that the neighbourhoods immediately outside the core are not just slow-growing, but highly desirable. In fact, several census tracts in neighbourhoods such as the Annex, Rosedale and Cabbagetown—all among the most expensive in the Canada—actually lost more housing than they gained between 2016 and 2021.
In other words, to explain these districts’ slow growth (or indeed negative growth) as market-driven is at best simplistic. The same can be argued regarding most of the districts lying between 3 and 34 kilometres from Union Station, which have also experienced important demand-side pressures in recent years. Rather, the sharp decline in housing stock growth in areas surrounding the core is likely indicative of policy barriers preventing the addition of housing.
3. Faster housing stock growth far from the regional core is driven by multiple factors, including a lack of housing closer to the core
The faster housing stock growth observed in census tracts beyond the 34-kilometre mark is driven by multiple factors. As noted in the previous section, the fastest growing tracts within this segment were located on or near the fringe of the contiguous urban GTA. That is, they include new neighbourhoods on the urban periphery, rather than the addition of homes in existing neighbourhoods.
Specifically, this segment’s fastest-growing tracts include several suburban neighbourhoods recently built along the northern edge of Oakville, the west and north of Brampton, and throughout York and Durham Regions. As a fast-growing metropolitan area, pressure for the GTA to expand its urban footprint is to be expected, as is demand for the relatively lower density housing types often built in new suburban communities.[17]
However, it is also likely that a portion of the growth occurring at the region’s urban fringe is accommodating households that would otherwise have preferred living closer to the core, had more housing been available in the neighbourhoods between three and 34 kilometres from Union Station. Indeed, research from 2016 measures the link between land-use regulation and housing supply, finding that many of the communities mentioned, notably in Brampton and York Region, would have grown more slowly between 2006 and 2011 had regulatory barriers such as lengthy, uncertain project approval timelines and local opposition been relaxed in communities closer to the regional core.[18]
Intuitively, it is likely that some percentage of households residing in outlying areas do so as a result of insufficient or unaffordable housing options closer to the regional core, which concentrates a higher proportion of the GTA’s employment, educational, commercial and entertainment features. This lens offers important considerations for local and provincial policymakers, who are tasked with anticipating a fast-growing population while also advocating for denser forms of urban development. Reconciling these two realities will require significant policy interventions to enable the rapid, feasible addition of housing units across all GTA communities.
Conclusion
As local and provincial governments consider ways of closing the large, growing gap between housing demand and housing supply in Ontario, they would benefit from greater understanding of current and past housing stock growth patterns. This report examined such patterns, comparing 2016 to 2021 housing stock growth among the GTA’s 1,227 census tracts.
The main finding is that a minority of the region’s census tracts are driving the majority of its growth. In particular, census tracts located within three kilometres of Union Station, or more than 34 kilometres from Union Station (that is, at the urban fringe) led housing stock growth in the GTA. The large majority of census tracts located in between added far less housing, with more than a third of such census tracts actually posting a net decline in housing stock between 2016 and 2021.
The three main policy considerations stemming from these findings are first, that the bulk of slow-growing neighbourhoods offer a significant opportunity to widen the number of areas contributing to housing supply; second, that policy barriers are the likelier reason for a marked fall in housing stock growth beyond Toronto’s Downtown core; and third, that slow growth just outside the core has likely contributed to demand for housing at the urban fringe.
Josef Filipowicz is a policy specialist focusing on urban and regional policy issues, including land-use regulations, housing affordability, property taxation, and municipal finance. He previously held roles at CMHC and the Fraser Institute, and holds an M.A. in Political Science from Wilfrid Laurier University and a Bachelor of Urban and Regional Planning from Ryerson University. His work has been featured in numerous news outlets including the Wall Street Journal, Globe and Mail, Toronto Star, Maclean’s, Detroit News, and Financial Post.
Steve LaFleur is a public policy analyst with over a decade of experience working for Canadian think tanks. He is a former Senior Policy Analyst at the Fraser Institute. He holds an M.A. from Wilfrid Laurier University. His work has appeared in most major Canadian media outlets including the Globe and Mail, the National Post, and the Toronto Star.
For more information about Ontario 360 and its objectives contact:
Sean Speer
Project Co-Director
sean.speer@utoronto.ca
Drew Fagan
Project Co-Director
drew.fagan@utoronto.ca
on360.ca