Canada’s Poor Productivity a Key Driver of Higher Home Prices

HIGHLIGHTS

• Canada’s housing market has been on a roller coaster ride since the pandemic as reflected by the profile for real private investment in residential structures, housing starts, sales and prices over this period.

• Housing affordability worsened significantly over this period with house prices reaching historical highs and mortgage rates increasing with the tightening in monetary policy since early 2022. Indeed, over this 5-year period home ownership affordability pressures have reached degrees like those witnessed in the early 1980s.

• Using Scotiabank Economics’ macro-econometric model of the Canadian and U.S. economies, we estimate that tightening supply constraints in construction from 2019Q3 to 2024Q4—reflecting weakening productivity and rising construction material costs—and above-normal population growth since 2022 each contributed to raise the benchmark MLS Home Price Index (HPI) a bit more than $50,000 over that 5-year period.

• This implies that if supply constraints had not tightened and population growth had stayed near its long-term average, the benchmark MLS HPI would have been slightly below $616,000 instead of the near $719,500 posted for 2024Q4.

• Our assessment and results strongly press the need to work on improving productivity to achieve housing affordability. Indeed, reducing bureaucratic burdens will also make housing supply more responsive to demand, thereby reducing price increases for a given rise in demand in the future.

• From 2024 to 2026, weaker population growth and uncertainty about trade barriers and their economic impact will reduce demand for homeownership. We forecast housing resale activity will slow in 2025 and 2026, declining from near 483,000 in 2024 to about 459,000 in 2026.

• Tight supply constraints will contribute to raise house prices especially in 2026 and mitigate progress on affordability from the past decline in interest rates and robust growth in real income. We expect the MLS House Price Index to rise by 0.4% in 2025 and 7% in 2026 with still-elevated supply constraints and pressure from the existing dwelling shortage.

• Of course, this expected profile for housing sales, starts and prices would be weaker if additional tariffs announced by the U.S. turn out more important than assumed in this forecast.

CONTEXT AND OBJECTIVES

Housing affordability remains a key concern of Canadians. Rightly so. There has been a well documented and systematic underbuilding of dwellings to satisfy housing demand that well pre-date the pandemic. Government policies that have frustrated the response of supply to demand have contributed to this sluggish supply response. The market for housing in Canada is not a particularly well functioning one. Governments have recently turned their attention to the factors limiting supply growth, but efforts are haphazard and have yet to bear fruit in a meaningful way.¹ In addition, demand-enhancing policy measures have been implemented over the years—most recently in the Fall of 2024 with measures tailored for first-time home buyers—which have generated headwinds for reducing the demand-supply imbalance, which is key for improving affordability.

Housing market affordability relates to the ability for a household to sustain its monthly shelter payments with its income. To identify the various factors at play in the Canadian housing market, particularly as they relate to home prices and affordability, we developed an econometric approach to identify the drivers of a series of housing-related indicators and incorporated those in Scotiabank’s macro-econometric Model. This work allows us to decompose the drivers of key housing indicators in a precise manner, and very importantly allows us to quantify the impacts of population growth and supply constraints on home prices and affordability.

The results are sobering. Supply constraints, as proxied by real unit labour costs in the construction sector, account for close to 27% of the increase in the national MLS home price (MLS HPI) over the last 5 years. This reflects to a large degree the nearly continuous decline in construction sector productivity since 2018, rising financing rates for developers and builders, and increased costs for construction material after the pandemic. The decline in productivity has been so sharp that construction sector productivity is now well below where it was in 1997, the year data were first available. This awful performance is the result of bottlenecks in the construction process coming largely from municipal regulation and red tape. We know that construction firms have deployed capital to be more productive yet the gains from those technologies and innovations have not been able to offset the impact of an increasingly challenging regulatory environment for builders. Annex 1 shows the estimated impact from a 1% rise in supply constraints (i.e. unit labour costs) and a 1% rise in real disposable income for all households. From these estimates—and on the bright side—they suggest that reducing unit labour costs, presumably through increased productivity, could have significant payoffs in terms of reducing house prices over time. Home prices would fall by about 1.2% for each 1% decline in unit labour costs. This is the key policy result flowing from our work.

Other factors are of course important. Population growth has also played an important role in driving home prices up, as was the post-pandemic demand for more spacious homes. Our results suggest that also about 27% of the increase in prices from 2019Q3 to 2024Q4 can be attributed to above-average population growth witnessed since mid-2022. This channel is less of a concern going forward as it seems clear population growth will slow in coming years. In fact, we predict less than 1% population growth this year and next, in sharp contrast to the 3%+ growth observed in the last three years. Interest rates have an important impact on prices, as expected, but this impact is dwarfed by that of supply constraints and population growth.

On affordability, our results suggest population growth has had a more important impact on the deterioration in affordability than supply constraints. This may seem counterintuitive given the above results, but it accounts for the fact that stronger population growth is putting upward pressure on economic activity and mortgage rates—through increased demand for both housing and non-housing consumption and government services—thereby raising inflation pressures and interest rates while those from tighter supply conditions are acting solely through rising house prices.

The key takeaway from our work is that supply constraints remain a considerable obstacle to affordability. It is simply impossible for home prices, and affordability, to return to pre-pandemic levels without a substantial increase in productivity. Policymakers should not kid themselves: we cannot generate a substantial improvement in affordability without much greater focus on productivity. Absent very significant action on that front, Canadians should expect only modest improvements in affordability at best. Ultimately, increasing productivity growth in residential construction—by reducing stringency of regulation and bureaucratic burdens—will make housing supply more responsive, which is the best option to sustainably reduce house price growth and improve affordability by better aligning house price growth with that of households’ income. Slower population growth going forward will also ease affordability pressures.

The remainder of this note lays out developments in key housing and affordability metrics since the pandemic, breaks down the impact of various drivers on these metrics, and then lays out a forecast for some of these variables in 2025–2026.

KEY DEVELOPMENTS SINCE THE PANDEMIC BEGAN

Chart 1 shows the profile of real business investment in residential structures since 2015. This indicator is a component of real Gross Domestic Product (GDP) and includes expenditures for new construction, renovation and ownership transfer costs. We see it increased sharply in the early weeks of the pandemic. And from 2019Q3 to 2022Q1, with this period ending just before the Bank of Canada (and other central banks around this time) started hiking its policy rate, the average contribution to real GDP growth from residential investment was about 5 times that of its pre-pandemic decade average.

In the early months of the pandemic, strong demand pushed housing activity to elevated levels. This is visible from the strong rise in housing sales and starts to their historical highs in early 2021 (chart 2). The impact of rising policy interest rates, which started in early 2022 in Canada and several major economies, subsequently contributed to cool housing conditions. Real residential investment declined significantly and is still below its level just before the pandemic started. Housing sales and starts largely contributed to this decline, but they were still above their pre-pandemic decade average at the end of 2024. 

Over the entire period since the pandemic began, housing prices have largely mimicked the dynamic of residential investment and housing sales and starts (chart 3). Both the national average MLS price and the MLS House Price Index (HPI) rose by more than 50% from the beginning of the pandemic to early 2022 with the strong rise in demand bumping into rigid supply. They both started trending down thereafter with economic and housing market conditions softening but they were still more than 30% above their end of 2019 level at the end of 2024.

Chart 4 shows annual growth in population since 2015. Prior to the pandemic, population growth was trending up slightly but declined during the worst of the pandemic period due to the impact on immigration from border restrictions imposed by the federal government. It then recovered after with these restrictions being lifted and from the increase in immigration targets by the government. 

Chart 5 is showing an indicator of supply conditions, which is real (or inflation adjusted) unit labour costs in the construction sector. This indicator is based on economic theory and is widely used by economic modellers as a proxy for the cost of producing an extra unit of a given product, including a dwelling in the case of our study. From the chart, we see that supply constraints tightened significantly from 2019Q3 to 2024Q4 which has made supply less reactive over this period. This tightening in supply conditions reflects declining productivity, rising costs for construction material and financing rates for developers and builders. 

Chart 6 shows 1-, 3- and 5-year benchmark mortgage rates since 2015. These were relatively stable until about the Bank of Canada started hiking its policy rate, at which point they increased significantly.

Finally, chart 7 shows the Bank of Canada Housing Affordability Index. This index captures the monthly income burden for a representative household from ownership-related costs, including mortgage payments and utility fees. It therefore reflects developments for each determinant mentioned previously. From 2019Q3 to 2024Q4, this index rose from 35.6% to 47%, an 11.4 percentage point increase, implying a heavier burden of home ownership costs in a household’s income. Indeed, in the third quarter of 2023, this index was at its highest since the early 1980s, a period when the conventional 5-year mortgage rate was above 20%. It declined since with both mortgage rates and house prices declining but was still near historical highs at the end of the historical period.

No stranger to the trend deterioration in affordability conditions reflected by this index is the upward trend for the share of apartments in total housing starts since late 1990s. Increased efforts by governments to densify housing helped raise supply for dwellings in larger structures, including condos for ownership. These structures allow developers and builders to reduce production costs per dwelling—including by reducing the size of available units—hence providing cheaper options for potential buyers. The trend deterioration in affordability conditions since the early 2000s—which accelerated with the start of the pandemic—has helped to increase demand for these smaller less expansive units. However, these structures take more time to plan and build, hence it is reasonable to assume that their profitability and feasibility is more affected by uncertainty and local regulations. With an increasing share of such units in new construction, any tightening in regulation for new residential construction—including for project review and approval delays and requirements—is likely to affect the construction costs of the average dwelling by more than in the past.

CONTRIBUTION OF HOUSING DRIVERS TO DYNAMIC OF RESIDENTIAL INVESTMENT AND PRICES: OVERVIEW

Table 1 shows the contribution of key drivers of housing demand along with our indicator of supply conditions over the period 2019Q3 to 2024Q4. From our estimates, the national housing market was relatively balanced in 2019Q3 with levels of residential investment and the MLS benchmark HPI consistent with housing market fundamentals. Hence we use this quarter as the base period. 

A first takeaway from table 1 is the impact the tightening in supply constraints since 2019 has had on the reported housing indicators, notably the MLS HPI and affordability conditions. This tightening lifted the national MLS HPI by 9.9% from 2019Q3 to 2024Q4, meaning that if supply constraints had stayed unchanged over this period, the benchmark MLS HPI in 2024Q4 would have been close to $667,500 instead of the about $719,400 posted for that quarter, a $52,000 difference. And these tighter supply conditions contributed to deteriorate affordability by raising the index by 2.8 percentage points over the same period, hence responsible for close to one-third of the deterioration in this index.

Another highlight from table 1 is the contribution from above-normal population growth—especially since 2022. Over this period, it lifted the MLS HPI by 9.8%, or about $51,500. Now accounting for the impact of population growth on the MLS HPI and mortgage rates—the latter from the channel described earlier—we estimate that rising population contributed to more than 80% of the deterioration in affordability conditions since 2019 (or 9.5 percentage points of the observed 11.4 increase).

The rise in mortgage rates from 2019Q3 to 2024Q4 helped slow growth in the MLS HPI but contributed modestly to the worsening in affordability conditions by raising monthly mortgage payments.

Growth in real disposable income per capita contributed positively to the dynamic of the reported housing indicators but this impact was much less important than for tightening supply constraints and rising population. This is not surprising given the growth performance of this driver has been relatively anaemic over this period.

OUTLOOK FOR HOUSING INDICATORS

Table 2 shows the forecast profiles for selected housing related indicators over the period 2024–2026. We expect housing sales activity to ease over this period due to the ongoing uncertainty about trade tariffs and a significant slowing in population growth. Housing sales are expected to decline from near 483,000 units in 2024 to just below 459,000 in 2026. Housing starts will also decline from 2024 to 2025 as the past and expected further tightening in supply conditions (mostly from weak productivity performance) weighs on new construction activity. Housing starts will then recover and reach about 250K units in 2026 as these supply constraints stabilize in 2026 and government policies aimed at increasing housing supply continue to support housing construction activity.

The MLS HPI is expected to recover over 2025–2026, following a -1.6% decline in 2024. We forecast that growth in the MLS HPI will reach 0.4% in 2025 and 7% in 2026. The combination of the expected tightening in supply conditions from 2024 to 2026 and the existing significant dwelling stock shortage will push house prices higher over this period, especially in 2026.

We forecast growth in real business residential investment to firm up from a -0.9% decline in 2024 to 5.4% and 2.2% (positive) growth rates in 2025 and 2026 respectively. The affordability index is expected to improve slightly from 2024 to 2026, with the burden of home ownership costs in households’ income declining down from 49.4% in 2024 to 46.3% in 2025 as the expected increase in per capita income dominates the modest expected rise in house prices. Affordability conditions deteriorate somewhat from 2025 to 2026 as the strong pick-up in house price growth dominates expected growth in per capita disposable income. Consequently, the housing affordability index rises to 47.1% in 2026, but still a weaker level than for 2024.

¹ The recent focus on supply by governments occurred at a time of rising interest rates, which are holding back supply growth despite efforts by governments to improve outcomes.

ANNEX—MODELLING FRAMEWORK AND MODEL PROPERTIES

Our econometric framework for housing indicators discussed in this report is based on an error-correction principle. These indicators evolve simultaneously with other economic and financial indicators within our large macroeconomic model of the Canadian and U.S. economies.

In the long term, the level of the real business residential investment and other housing market indicators are based on estimated long-term economic relationships with the level of relevant fundamental determinants, or drivers. These long-term relationships essentially act as anchors to which these housing indicators revert to when submitted to unexpected developments (or shocks).

The error correction equations assume the modeled indicator converges to this long-term fundamental or desired level, but they also capture the effect of various rigidities that are slowing the adjustment of the modeled housing indicators towards their long-term fundamental level. Examples of such rigidities are mortgage or price contracts for construction material, and a less responsive supply curve in the short- than in the longer term. This framework also allows to incorporate the prospective nature of housing-related decisions by capturing the effect of expectations of future economic and housing market conditions (assuming that these adjustments are costly under a Polynomial Adjustment Cost framework).

Below is the list of the housing indicators reported in table 2, along with their drivers. These could enter both in level—defining the long-term level of the housing indicator—or with lagged changes (either in growth rate or percentage point difference) to capture the shorter-term movements in the indicator. In the case of business residential investment, expectations of future values for the level of the fundamental determinants are also included.

Real Business Residential Investment

Long-term level determinants

• Real (inflation adjusted) disposable income for all households
• Real unit labour costs in construction sector
• 5-year conventional real mortgage rate
• Labour force participation rate

Short-term dynamic determinants

• Previous period’s deviation level of residential investment from its long-term desired level
• Expected profile for this long-term desired level (therefore assuming a Polynomial Adjustment Cost framework for residential investment)
• Growth in stock market (TSX, inflation adjusted)
• Deviation of real disposable income growth from potential GDP growth

MLS HPI

Long-term level determinants

• Real business residential investment relative to long run trend population 
• Unit labour costs in construction sector (inflation adjusted)

Short-term dynamic determinants

• Previous period’s deviation of the MLS HPI from its long-term level
• Lagged growth rates for the MLS HPI
• Growth in stock market (inflation adjusted, TSX)
• Deviation of real residential investment growth from potential GDP growth
• Growth in real unit labour costs in construction

Housing Sales

Long-term level determinants

• Population
• Real 5-year conventional mortgage rate
• Labour force participation rate
• Real unit labour costs in construction

Short-term dynamic determinants

• Previous period’s deviation of housing sales from their long-term level
• Change (in percentage points) in the real 5-year conventional mortgage rate in the current and previous periods
• Current and previous period growth in stock market (inflation adjusted, TSX)

Housing Starts

Long-term level determinants

• Population
• Real unit labour costs in construction
• Real business residential investment as a ratio to potential GDP
• Unemployment rate

Short-term dynamic determinants

• Previous period’s deviation of housing starts from their long-term level
• Previous period’s growth in housing starts
• Deviation of real residential investment growth from potential GDP growth
• Change (in p.p.) in the unemployment rate from previous period

We now illustrate the channels through which developments in real unit labour costs in construction—or supply constraints—and real disposable income for all households affect key housing indicators.

The left chart below shows the response of real business residential investment, the MLS HPI and the BoC Housing Affordability index following a temporary 1% increase in unit labour costs. This shock reduces housing supply, thereby raising the level of the MLS HPI with a peak impact of 1.6%. Higher house price deteriorates affordability for home ownership by raising mortgage payments, and the maximum impact on this index is 0.6 percentage points. This deterioration in affordability conditions reduces housing market activity, which translates into a trough decline of 0.6% for real business residential investment. 

The chart on the right shows the impact on the same indicators as above from a temporary 1% increase in all-households disposable income. This rise can originate from a combination of a rise in income per capita and rise in working-age population. More income raises demand for housing, hence housing activity, thereby increasing both house prices and business residential investment. Their respective peak impact is +0.94% and +0.84%. However, this shock is worsening affordability conditions with the index rising by 0.5 percentage points at its peak as both the house price and mortgage rates are increasing. The latter is due to the impact from higher household consumption (housing and non-housing) on economic conditions, which is lifting economic activity above its potential. Consequently, inflation pressures are rising, which requires the central bank to raise its policy rate to keep inflation near its targeted rate, hence the rise in mortgage rates.