In the past 60 years, inflation-adjusted GDP growth in Canada has trended lower, except for a period between the mid-1990s and the mid-2000s (chart 1). Given annual average growth in the most recent business cycle (2009-2019) clocked in at 1.75%, it's no surprise, with an election just around the corner, that proposed policies to generate higher economic growth have taken center stage. Post election, we believe it's likely that budgets will be premised on projections that the country will be able to raise its long-term GDP growth rate--perhaps above 2%, or even, ambitiously, above the 3% of the mid-1990s to mid-2000s. With a faster growth rate, the budget deficit would be smaller, taking pressure off deficit reduction and possibly even opening the door to tax cuts or spending increases.

Chart 1

S&P Global Ratings believes economic growth in Canada on an annual average real GDP growth basis for the next 10 years is likely going to be close to what was experienced the 10 years before the pandemic. At 2.2% average annual growth (our central estimate), this would still be a tad higher than the 1.8% of the recent past (2015-2019)--a calculation artifact in estimation due to higher labor productivity growth in longer history compared with 2015-2019. It's hard to foresee change in annual growth average by more than a few tenths of a percentage point--either direction--in the coming decade.

To assess Canada's growth over the next decade, we applied a Monte Carlo simulation to generate a probability distribution of a range of outcomes. We estimated the likely range of future potential GDP growth by taking random draws from 38 years' (1981-2019) labor productivity, combined with median growth in the employment rate of the working-age population during the last business cycle, and Statistics Canada's (M1 medium) growth projected for the working-age population (1). In 1 million simulations, the estimated median annual real GDP growth rate over the next 10 years is 2.18%, with the interquartile range 1.44%-2.92%. (See table 1 for distribution of outcomes that vary with historical experience in labor productivity.) The odds of the growth rate averaging 4% or above (a mark surpassed six times since 1981) during the 10-year span are only 5%, which would essentially require the economy to repeat some of the fastest labor productivity growth in the past (a 90% confidence interval ran from 0.4% to 4%). By drawing on history, in a sense, this analysis effectively includes not only a range of policy outcomes but also imbedded uncertainty about the economy.

Adding Them Up: A Modified Growth Accounting Framework

We use a modified version of a simple growth accounting framework to assess the impact of changing labor productivity, employment rate, and working-age population growth on the economy's long-term performance.

We separate real GDP into its two primary contributions: GDP per employed person and labor inputs:

• GDP = (GDP/hours worked)*(hours worked/no. of employed)*(no. of employed/working-age population 15+)*working-age population 15+

Simplifying the above equation gives:

• GDP = GDP per no. of employed (i.e. labor productivity) * employment rate * working-age population

The multiplicative factor is transformed into an additive relationship by using the natural log of both sides, thus giving us the growth rate contribution of each component (in percentage points) to GDP growth.

Beyond the cyclical (demand-driven) boost to the economy, growth in the long term (once the unemployment rate is at or near "full employment," which is about 6% in Canada) is essentially determined by growth (not level) in output per labor hours (labor productivity) and growth in total labor hours worked (labor quantity). Changes in both these factors can bring about a range of plausible outcomes for growth in the next decade, especially past 2022 when our short-term forecast has the unemployment rate reaching its full employment level, and any future cyclical boost coming from additional demand from fiscal policy is offset by the Bank of Canada, given its mandate to maintain price stability.

Of the two sources of potential growth, assumption on growth in labor quantity is straightforward: Via Statistics Canada, we already know the working-age population growth in the coming years (determinants: fertility choices were made decades ago, mortality rates are relatively stable, and immigration is the only factor varying with policy). Combine this with growth in the employment rate (the fraction of population working)--admittedly, a greater unknown but a smaller contributor in the grand scheme of things (employment growth averaged 0.11% in 1981-2019, 0.08% in 2010-2019)--and we get the contribution to growth from the labor quantity channel. It is likely that average annual employment rate growth is going to stay close to what was experienced in 2010-2019. (See table 2 for distribution of outcomes that vary with historical experience in employment rate growth as well in addition to variation in labor force productivity.) Given the current distribution of the population and immigration targets, Canada is basically on a trajectory to maintain its working-age population growth rate within 1%-1.5%, which is meaningfully above the growth rate projected for the U.S. (chart 2).

Chart 2

Chart 3

On the other hand, the other source of potential growth--growth in labor productivity--is trickier to project and tends to swing average growth forecasts one way or the other. Growth in labor productivity comes from three broad factors: increases in the amount of capital per hour worked (also known as "capital deepening" or "capital intensity"); improvements in the quality of labor; and growth in "multifactor productivity" (also referred to as total factor productivity, or TFP), which captures improvements in the way firms use their capital and labor but also embeds any errors in the estimated contributions from capital deepening and labor quality--commonly interpreted as a measure of technical progress.

Chart 4

As a matter of "adding up," any softening impulse from working-age population and employment rate to supply-side growth must be offset by increasing labor productivity growth to maintain the same level of GDP growth over the horizon.

The Shrinking Case Of Productivity

Labor productivity growth has slowed--in accounting terms--in Canada since the early 1970s. Input growth went up but increase in output was slower--the economy absorbed more and more workers and more and more capital, but it didn't translate into more and more output growth. This trend has been true for other G7 nations as well, the reasons behind which are beyond this article's scope. (For a detailed discussion on productivity growth, see "The Strange Case Of Shrinking U.S. Productivity Growth: Myth, Mismeasurement, Or Multiyear Phase?" published May 5, 2016, on RatingsDirect.) In Canada, one hallmark stands out, and goes back to the 1970s: the consistent low contribution from TFP in the business sector (see chart 5).

Chart 5

Lower TFP growth in Canada (and other countries more recently) has raised the issue of how estimated TFP growth is interpreted, especially with large downswings in Canada in the mining and oil and gas sector (8% of GDP), which suggests--oddly enough--that no technical progress in combining labor and capital inputs has been made in the sector since the 1960s (see chart 6). Researchers at Statistics Canada have concluded that as the measurement system improved, the unexplained part of labor productivity growth (the residual) has diminished (2). Labor productivity growth, and ensuing economic growth, has come mostly from tangible and intangible business investments. With goods sectors such as construction and mining, which are both large in their size as a share of GDP (see chart 7), business investment in these sectors will continue to play an important role in breaking growth numbers one way or the other. On the other hand, the service sector--which has grown in time to now make up a little more than 70% of the economy (versus 57% in 1961)--is where productivity growth is much harder to achieve than in goods. (Note: For our empirical exercise, we are agnostic about contribution to labor force productivity between capital intensity and TFP.)

Chart 6

Chart 7

As the need for COVID-19 emergency measures diminishes, whichever party forms the new government, its structural policies could have a small effect on the future growth rate; for instance, regulatory or tax reforms could raise the growth rate from the current long-term consensus of 1.8% and restrictions on trade or immigration could lower it. For example, non-tariff barriers arising from interprovincial differences in both product and labor markets have long been flagged as a headwind to total factor productivity (and hence labor productivity) of the business sector. The Bank of Canada estimated that expediting the removal of such non-tariff barriers could add two-tenths of a percentage-point increase in long-run GDP growth. The OECD estimates five-tenths of a percentage-point increase, on average, in annual GDP to 2030 if a more ambitious reform that lessens restrictive product-market regulation on other fronts, including telecommunications (3). Such liberalization, in the process, would reallocate employment towards provinces that experience large productivity gains from trade. Either way, it is hard to foresee any policies that could change the growth rate by more than a few tenths of a percentage point over the next decade nationally.

We applied a similar estimation method to derive 10-year growth distribution at the provincial level (chart 8).

Breaking Down Projected Growth At The Provincial Level

It isn't surprising that provincial growth projections converge toward the national level (see chart 8). Nevertheless, when we dig deeper into the key factors that contribute to real GDP growth--labor productivity, labor force participation, and working-age population--it becomes clear that the national median tells us very little about what is happening at the regional level (see chart 9). Although labor productivity appears to have the least amount of variability among the provinces, it does not explain this rank ordering, which is based on our modeling of real GDP growth projections for each province. Gains in employment and working-age population are the keys to the differences between provincial growth projections. Because provinces with similar economic structures generally tend to follow the same growth trends, we have compared the composition of real GDP growth by three groups: large, diversified economies, the Atlantic provinces, and economies concentrated in oil and gas and mining.

Chart 8

Chart 9

Compared with peers, British Columbia, Manitoba, Ontario, and Quebec are the most diversified economies. Their real GDP growth rates have been historically more stable than those of other provinces. The cyclical growth of the Canadian economy, as a whole, is somewhat influenced by resource and commodity prices. As a result, we expect that real GDP growth for British Columbia, Ontario, and Quebec will increase at a slightly faster pace than the national average in the next 10 years. This is largely because more diversified economies are poised to ride out any weakness in energy and mining better than more concentrated ones are, in particular where it relates to their overall participation rates. Although Manitoba's high participation rate will bolster economic stability, it will also hinder real GDP growth in the medium term compared with its direct peers and the national average.

Canada is facing the hindrances of an aging population. While the trend is national, it is most notable in the Atlantic provinces, which have recorded years of outmigration by young people. In the next 10 years, our modeling predicts that New Brunswick, Nova Scotia, and Prince Edward Island will record a lower contribution from the working-age population to average real GDP growth compared with the national median. Nevertheless, investments in machinery and equipment should help to improve labor productivity and somewhat offset the impact of an aging workforce.

On a relative basis, the economies of Alberta, Newfoundland and Labrador, and Saskatchewan have a higher exposure to commodities than other provinces. Our modelling projects that economies more concentrated in the resource sector will have more volatility in real GDP growth forecasts in low- and high-growth scenarios (at the 25th and 75th percentiles). Provinces whose economies are significantly influenced by resource prices show less correlation between national and provincial growth. Historically, when commodity prices were high, a shortage of skilled workers and the demand to fill high-paying jobs led to increased intraprovincial migration to Alberta and Saskatchewan. As a result, the largest projected contribution of working-age population to real GDP growth is in these two provinces. In line with its Atlantic peers, Newfoundland and Labrador is contending with an aging population. Unlike the other resource-based economies, its growth forecast is bolstered by an increasing participation rate and relatively higher contribution from labor productivity. In its Fiscal Sustainability Report for 2021, the Office of the Parliamentary Budget Officer (PBO) also projects that Newfoundland and Labrador will have the highest labor productivity growth of the provinces, on average, between 2026-2095.

What This Means For Provincial Creditworthiness

We believe that government policies on the labor market matter and can influence the long-term fiscal outcomes for Canadian provinces, affecting fiscal sustainability. According to the PBO, population aging will contribute to slower growth in total hours worked and as a result, projects that economic growth in New Brunswick, Newfoundland and Labrador, and Nova Scotia will face material hurdles from population aging (3). Measures to encourage investment and increase labor productivity will help to counteract demographic shifts and bolster real GDP growth. Manitoba, Nova Scotia, and Prince Edward Island have successively used Provincial Nominee Programs to attract newcomers to increase the working age and rate of labor participation in their economies. Notably, Quebec's employment rate has materially improved as a result of policies enacted in the recent past to encourage people to remain in or to return to the workforce.

Structural issues in a province's economy are factored into our assessment of a regional government's economy. Our economic assessment is underpinned by wealth and income levels, measured by GDP per capita. While national GDP per capita is used as a starting point, we apply qualitative adjustments in our analysis for each province, when relevant, for economic growth prospects, economic concentration, and/or volatility and socioeconomic profile.

In addition, real GDP growth forms the assumptions that are the foundation of our fiscal forecasts, which filter into our assessments for budgetary performance and debt. While the provincial share of national corporate income tax revenue is correlated to performance across the country, a province's revenue-generating ability depends largely on its underlying tax base. The PBO also notes that real GDP per capita at the provincial level is an important contributor to fiscal capacity that determines a province's eligibility for equalization payments. Expenditures are primarily based on growth assumptions for population and inflation. A province's operating and capital needs are largely influenced by socioeconomic trends. Health care costs related to an aging population could outpace revenue growth, without commensurate increases to federal transfers or increases in provincial revenue that are driven by real GDP growth. All these components spur budgetary performance and borrowing needs and ultimately, whether a province will achieve fiscal sustainability.

Table 1

Notes

1) Furman (2017) does similar exercise for the US using hours worked, labor force participation rate, labor productivity, and working age population. They find US median potential growth rate to average 1.7% annually between 2016-2026, with 1.3% to 2.1% interquartile range. https://www.piie.com/blogs/realtime-economic-issues-watch/what-ten-million-simulations-tell-us-about-president-trumps

2) For a full description of the multifactor productivity program at Statistics Canada, see https://www150.statcan.gc.ca/n1/pub/15-206-x/2013031/part-partie1-eng.htm

3) For a survey on literature on regulatory burdens in Canada and impact of deregulation, see https://www.oecd.org/economy/surveys/Canada-2021-OECD-economic-survey-overview.pdf

Appendix

Summary statistics: Monte Carlo Simulation (n= 1,000,000)

Table 2

Table 3