Assessing global potential output growth: April 2025

Global potential output growth

In this note, we present the annual update of the Bank of Canada staff estimates for growth of global potential output. These estimates served as key inputs to the April 2025 Monetary Policy Report (MPR).

Economic recovery since the COVID-19 shock

By 2024, global potential output growth had almost fully recovered from the shock caused by the COVID-19 pandemic, rising from 2.2% in 2020 to 3.1% (Chart 1). Global potential output growth is nevertheless slower than it was before the pandemic, partly reflecting the negative effects from population aging. The US economy has been an outlier, with exceptionally strong growth since the pandemic.

• US economic growth has been stronger than that of other economies. This strength reflects a significant increase in growth of trend total factor productivity (TFP), which is currently estimated to be more than 50% higher than its 2015–19 average. Strong performance of the information and the professional services sectors, robust firm creation and efficient reallocation of labour after the pandemic have contributed to the US exceptionalism.
• Potential output growth in China has continued to slow gradually due to population aging, but strong investment in its manufacturing capacity has partially offset this contraction.
• In the euro area, potential output growth has been constrained by structural competitiveness issues in its manufacturing sector, which have been exacerbated by relatively higher energy prices since Russia’s invasion of Ukraine.
• The recovery for oil-importing emerging-market economies (EMEs) has also been gradual, reflecting scarring from the COVID-19 shock and, more recently, a slowdown in inflows of foreign direct investment.

Two scenarios for potential output given elevated US trade policy uncertainty

The United States has threatened, imposed and suspended tariffs on its trading partners, including Canada. Many tariffs remain in place, and what will happen next is unclear. Accordingly, the Bank presented two illustrative scenarios in the April 2025 MPR, each with different assumptions about the future course of US trade policy. They were designed such that, when taken together, they encompass a range of trade policies and associated paths for inflation and growth of gross domestic product (GDP) for the global economy. These scenarios were presented instead of a base-case economic projection.

We therefore provide two estimates of potential output that correspond with the two scenarios presented in the April MPR.1

• Scenario 1 assumes that most tariffs imposed since the trade conflict began are negotiated away, but the process is unpredictable. Uncertainty about US trade policy continues, and businesses and households around the globe remain cautious until the end of 2026.
• Scenario 2 assumes that the uncertainty and limited tariffs in Scenario 1 persist and other US tariffs are imposed. A long-lasting global trade war unfolds.

Scenario 1

Global potential output growth moderates marginally to average 3% between 2025 and 2028. The slowdown mainly reflects weakening trend labour input (TLI) growth amid the aging of the global population (Chart 2). Moreover, as threats of a widespread trade conflict loom, businesses typically postpone investment out of cautiousness. We estimate this will remove about 0.1% from the level of global potential output by 2027 through reduced capital accumulation. The rise in total factor productivity—including a boost from the adoption of artificial intelligence (AI)—partially offsets these negative forces.

Compared with the 2024 assessment, the outlook for global potential output growth is marginally stronger (Chart 3).2

The outlook for the United States is revised up because the boost to productivity from AI more than offsets the drag on labour input from tighter immigration policies.

Potential output growth for China is also more robust because of China’s:

• stronger-than-expected capital accumulation in the manufacturing sector
• recent changes to raise the mandatory age for retirement

These positive revisions are partially offset by a weaker outlook for the euro area and EMEs.

Scenario 2

A more severe US-driven trade war slows global potential output growth to around 2.9% through 2027. Productivity growth declines significantly because the tariffs distort the allocation of resources away from optimal levels implied by their comparative advantage. The cyclical slowdown in investment also weakens capital stock accumulation over time.

Relative to Scenario 1, the additional tariffs assumed under Scenario 2 slow global potential output growth by an average of 0.1 percentage points through 2027. The level of global potential output is therefore about 0.4% lower in 2027.

Table 1: Projection for potential output growth

* Shares of gross domestic product (GDP) are based on International Monetary Fund (IMF) estimates of the purchasing-power-parity valuation of country GDPs for 2023 from the IMF’s October 2024 World Economic Outlook. The individual shares may not add up to 100 due to rounding.
‡ Croatia joined the euro area on January 1, 2023. The current projection and historical data include the change in membership.
§ The oil-importing emerging-market economies (EMEs) grouping excludes China. It is composed of large EMEs from Asia, Latin America, the Middle East, Europe and Africa (such as India, Brazil and South Africa) as well as newly industrialized economies (such as South Korea).
◊ “Rest of the world” is a grouping of other economies not included in the first five regions. It is composed of oil-exporting EMEs (such as Russia, Nigeria and Saudi Arabia) and other advanced economies (such as Canada, the United Kingdom and Australia).

United States

After reaching 2.7% in 2024, potential output growth slows from 2025 to 2028 due to declining trend population growth. Stricter immigration policies and increased deportations reduce trend population growth from 0.9% in 2024 to an average of 0.7% from 2025 to 2028. However, trend labour productivity growth remains strong over 2025–28, averaging 1.9%. This growth is supported by AI-related capital spending, anticipated efficiency gains from increased adoption of AI across the economy and continued strength in information and the professional services sectors.

Scenario 1

Potential output growth is 2.4% in 2025 and then averages 2.2% through 2028 due to the slowdown in TLI growth (Chart 4). Over the same period, trade-related uncertainty removes on average 0.1 percentage points from the growth of capital accumulation.

Compared with the April 2024 assessment, potential output growth is revised up by about 0.2 percentage points, on average, from 2025 to 2027 (Chart 5). The level of potential output is higher by 0.9%.3

• The upward revision over the scenario horizon stems from stronger capital accumulation of around 0.6 percentage points and trend TFP growth of around 0.1 percentage points, on average, over 2025–27. This partly reflects AI-related investment spending, which boosts capital accumulation by around 0.1 percentage points, on average, over 2025–27 (Box 1). Similarly, the widespread adoption of AI in the economy contributes modest efficiency gains to annual trend TFP growth of 0.1 percentage points, on average.
• These effects are partly offset by weaker TLI over the same period due to slower population growth. Population growth is 0.3 percentage points weaker, on average, over 2025–27 because of a lower level of projected net immigration. The flow of undocumented immigrants slows, while the number of deportations ramps up to an annual pace of 400,000.

Scenario 2

Potential output growth slows more sharply than in Scenario 1, falling to 1.7% in 2026 before recovering to 2.2% in 2028 (Chart 4). Trend labour productivity growth moderates to an average of 1.6% over this period. Overall, this leaves the level of potential output 1% weaker by 2027 than in Scenario 1.

Two factors are behind this:

• Capital accumulation growth is revised down by 0.3 percentage points, on average, from 2025 through 2028 as businesses delay or forgo new investments due to increased barriers to trade and more expensive capital goods.
• Trend TFP is also revised down by 0.1 percentage points, on average, over the same period. This reflects the effects of tariffs, which limit competition and negatively affect the effective allocation of resources in the economy.

Euro area

The combination of the COVID-19 pandemic and persistently high energy prices since Russia’s invasion of Ukraine severely affected potential output in the euro area. Recovery from these shocks has been slow given the presence of other existing structural challenges. These include a high regulatory burden, a lack of infrastructure investment and a lack of policy coordination across countries in the euro area. These factors continue to constrain potential output growth, with AI adoption and increased fiscal investment providing only a partial offset.

Scenario 1

Potential output grows by 1.0% on average over 2025–28 (Chart 6). Population aging weighs on TLI over 2025–28 as a result of reduced immigration and low participation rates for people 65 years old and older. However, recently announced additional infrastructure and defence spending partially offsets this by raising the level of capital stock and trend TFP.

Compared with the 2024 assessment, potential output growth is revised down by around 0.1 percentage points per year from 2025 to 2027 because of larger drags from increased competition from China, labour market frictions and rigid regulations. This factor is partially offset by an upward revision to TLI due to higher estimates of the working-age population from 2023 to 2026.

Scenario 2

The level of potential output is 0.2% lower than in Scenario 1 by 2027. This is mainly because of inefficient resource allocation and weaker capital deepening caused by the imposition of tariffs from the United States.

China

Potential output growth in China has been on a steady decline since the early 2010s because the size of its labour force has been shrinking due to population aging. The slow adoption of structural reforms has also limited productivity growth. Instead, potential output growth continues to be supported by strong investment, which has been more intently focused on expanding capacity in the manufacturing sector in recent years. Going forward, potential output growth continues to moderate as the pace of capital deepening slows, while TLI continues to contract and trend TFP growth remains modest without meaningful reforms.

Scenario 1

Potential output growth in China slows to an average of 4.1% over 2025–28 (Chart 7). The contraction in the labour force reflects population aging. Additionally, capital deepening moderates because high levels of public debt limit capacity for further policy-fuelled investments. Lastly, the lack of structural reform, geopolitical tensions with advanced economies and demographic shifts weigh on trend TFP growth, but the gradual adoption of AI offsets this downward pressure.

Compared with the 2024 assessment, potential output growth is revised up by 0.3 percentage points, on average, in 2025–27. This leaves the level of potential output higher by 0.9% by 2027.4 The upward revisions over the scenario horizon mainly reflect labour market reforms and stronger-than-expected investment growth.

• Capital deepening is revised up due to policy-fuelled support for non-residential investment, particularly in sectors such as electric vehicles and batteries.
• TLI growth has been revised up due to reforms to the retirement age. Government officials announced the mandatory retirement age would be raised from 55 for women and 60 for men to 58 for women and 63 for men.
• Trend TFP growth is broadly unchanged because the steady adoption of AI offsets the loss of efficiency from the tariffs that have already been implemented. The gradual adoption of AI through the economy is anticipated to yield some productivity gains. However, the imposition of a 10% tariff from the United States should result in some inefficient reallocation of production, lowering productivity prospects.

Scenario 2

With the full 25% tariffs, the level of potential output in China is 0.1% lower by 2027. This is less than in Scenario 1 mainly because of:

• lower investment from the higher cost of imported capital and lower global demand
• weaker productivity growth from shifting production to less efficient sectors

The marginal effect of Scenario 2 for China is lower than for other regions due to the relatively small share of trade with the United States in China’s economy (estimated at 3% of GDP) and the possibility of diverting trade to other countries to lessen the impact of the tariffs. Furthermore, the marginal tariff increase in Scenario 2 is lower in China because Scenario 1 already embeds a 10% tariff.

Oil-importing emerging-market economies

Potential output growth in oil-importing EMEs has gradually recovered from the pandemic-induced collapse in 2020 (Chart 8). Relative to most other regions, the post-pandemic recovery in oil-importing EMEs has been slow due to deep scarring effects in the labour market because the labour force participation rate dropped significantly during the pandemic and has only just recovered. A weakening in global foreign direct investment inflows in recent years also weighed on growth. With the recovery mostly complete, potential output growth should remain roughly stable over 2025–28.

Scenario 1

After growing 3.8% in 2024, potential output stabilizes at around 4% over 2025–28. TLI growth eases as the participation rate fully recovers from the COVID-19 pandemic. In contrast, investment growth strengthens as domestic financial conditions generally ease. Meanwhile, trend TFP growth improves as the region continues to catch up to frontier technologies. Some progress in implementing structural reforms also helps improve productivity.

Compared with the 2024 assessment, potential output growth over 2025–27 is 0.1 percentage points lower per year.

• Part of the downward revision reflects a weaker pace of inflows from global foreign direct investment, which weighs on both productivity growth and capital accumulation across EMEs. This trend has been already observed in recent years, and the environment of heightened trade uncertainty embedded in the first scenario will likely exacerbate this.
• Country-specific developments, particularly a deep financial crisis in Egypt, also contribute to the downward revision.

Scenario 2

Potential output across EMEs grows by 3.9% on average over 2025–28. Potential output growth will be particularly affected in 2026 and 2027, falling from about 4% to 3.8% and 3.9%, respectively. The level of potential output will be 0.3% lower than it is under Scenario 1 by the end of 2027.

In particular, Mexico undergoes significant adjustments to its economy given the deep integration of the North American market and supply chain. While Mexico faces less of an increase in tariff rates than other countries, its economy is among the hardest hit by US tariffs, with the level of its potential output about 1% weaker than in Scenario 1.5

Other regions

In Japan, potential output growth remains stable at around 0.6% over 2025–28 under Scenario 1. While the participation rate keeps rising, population aging increasingly puts downward pressure on TLI. Compared with the 2024 assessment, potential output growth in Japan is revised down modestly as recent and expected tightening of monetary policy weigh on investment. Under Scenario 2, the imposition of tariffs further lowers the level of potential output by 0.2% by 2027.

In the rest-of-the-world grouping, potential output growth in Scenario 1 averages 2.1% over 2025–28, roughly in line with its pre-pandemic average. The contribution from capital deepening picks up as several oil producers invest to diversify into non-oil sectors. Compared with the 2024 assessment, potential output is marginally stronger, reflecting robust investment trends in Russia since in recent years. Under Scenario 2, with tariffs imposed, potential output growth instead averages 2% through 2027 and the level of potential is lower by 0.3%.

Uncertainties around the outlook for global potential GDP

Trade-related uncertainties are clearly the biggest risks the global economy currently faces. The two scenarios presented in this note are illustrative scenarios, but the full range of possibilities is extensive given the potentially unprecedented shifts in US policy. Moreover, risks around the potential impacts of these tariffs are also a source of uncertainty (See the Risk section of the April 2025 MPR for more details).

Aside from these considerations, prominent risks specific to the outlook for global potential output are also worth noting:

• In the United States, the administration’s wide range of other policy announcements present both upside and downside risks. These include announcements about cuts to corporate and personal income taxes, broad-based deregulation and lower immigration. These policies could stimulate certain sectors of the economy, boosting productivity and growth, but they also carry the risk of increased costs and inefficiencies for businesses.
• In China, potential output growth could be enhanced due to a faster-than-expected implementation of structural reforms, such as reducing barriers to entrepreneurship and increasing the depth of financial markets. However, trend TFP growth could be much weaker if state interventions were to hinder innovation and productivity gains, as argued in last year’s assessment of global potential output growth (Benmoussa et al. 2024).
• In the rest-of-the-world grouping, tight global monetary conditions could worsen sovereign debt risks across several EMEs, resulting in increased economic uncertainty and reduced investment. In contrast, certain countries with fiscal space could adopt a more expansionary fiscal stance to help offset negative impacts of heighted global uncertainty and a trade war. This could potentially lead to higher government spending and investment in infrastructure.

Appendix: Methodology for estimating potential output

We estimate potential output growth for every region using a growth accounting framework centred on a Cobb-Douglas aggregate production function. This framework assumes the following relationship between a country’s aggregate output and each factor of production (where \(\%∆x\) denotes the percentage change in variable \(x\)):

\(\displaystyle\,\%∆Y_t\) \(\displaystyle=\,\alpha\%∆\left(\frac{K_t}{L_t}\right )\) \(\displaystyle+\,\%∆L_t\) \(\displaystyle+\,\%∆TFP_t\) \(\displaystyle,\) \(\displaystyle (1)\)

where \(Y\) is real GDP, \(\frac{K}{L}\) is real capital stock per worker, \(L\) is labour input, \(TFP\) is total factor productivity, and \(\alpha\) is the share of capital income in output.

We construct country-level capital stocks using the perpetual inventory method based on:

• either national accounts data on gross fixed capital formation or detailed asset-level investment data; and
• data on average depreciation rates and prices of various asset types.6

Potential output is evaluated based on actual capital stocks because it determines the limits on an economy’s productive capacity today. Labour input—the total number of hours worked in the economy—is calculated as the product of four components:

• the working-age population
• the labour force participation rate
• 1 minus the unemployment rate
• average work hours per person employed

Finally, we calculate TFP growth as the Solow residual in equation (1) using national accounts data on real GDP growth. Thus, TFP growth captures contributions to productivity from many factors, including global improvements in technology, efficiency gains resulting from domestic innovation, structural reforms, terms-of-trade shocks, financial and geopolitical crises, and human capital accumulation.7

To abstract from the business cycle, we use trend levels of labour input and TFP to construct potential GDP growth as the sum of the respective input contributions according to the decomposition in equation (1). This notion of potential output coincides with production at full capacity—that is, the level consistent with full employment and long-run TFP.

References

Abraham, S., D. Brouillette, A. Chernoff, C. Hajzler, S. Houle, M. Kim and T. Taskin. 2025. “Potential Output in Canada: 2025 Assessment.” Bank of Canada Staff Analytical Note No. 2025-14.

Acemoglu, D. 2024. “The Simple Macroeconomics of Artificial Intelligence.” National Bureau of Economic Research Working Paper No. 32487.

Acemoglu, D. and P. Restrepo 2018. “The Race Between Man and Machine: Implications of Technology for Growth, Factor Shares, and Employment.” American Economic Review 108 (6): 1,488–1,542.

Acemoglu, D. and P. Restrepo. 2019. “Artificial Intelligence, Automation, and Work.” In The Economics of Artificial Intelligence: An Agenda, edited by A. Agrawal, J. Gans and A. Goldfarb, 197–236. University of Chicago Press.

Acemoglu, D. and P. Restrepo. 2022. “Tasks, Automation and the Rise in US Wage Inequality.” Econometrica: Journal of the Econometric Society 90 (5): 1,973–2,016.

Aghion, P. and S. Bunel. 2024. “AI and Growth: Where Do We Stand?” Federal Reserve Bank of San Francisco Working Paper.

Baily, M., E. Brynjolfsson and A. Korinek. 2023. “Machines of Mind: How Generative AI will Power the Coming Productivity Boom.” Brookings Institute, May 5.

Bank of Canada. 2025. Monetary Policy Report (April).

Benmoussa, A., R, Dastagir, E. Ekanayake, J. Guénette, H. Lao, J. Rolland-Mills, A. Spencer and L. Xiang. 2024. “Assessing Global Potential Output Growth: April 2024.” Bank of Canada Staff Analytical Note No. 2024-10

Briggs, J. and D. Kodnani. 2023. “The Potentially Large Effects of Artificial Intelligence on Economic Growth.” Global Economics Analyst, Goldman Sachs Economics Research, March 26.

Filippucci, F., P. Gal and M. Schief. 2024. “Miracle or Myth? Assessing the Macroeconomic Productivity Gains from Artificial Intelligence.” Organisation for Economic Co-operation and Development Artificial Intelligence Papers, No. 29.

McKinsey. 2023. “The Economic Potential of Generative AI.” McKinsey Report, June 14.

Endnotes

Acknowledgements

We would like to thank Eric Santor, Subrata Sarker, Fotios Raptis, Steven Gagnon, Justin-Damien Guénette and Abeer Reza for their useful comments and suggestions. In addition, we thank and Colette Stoeber and Nicole van de Wolfshaar for excellent editorial assistance and Guylaine Létourneau and Philippe Audet-Cayer for their help translating this note into French.