Labour productivity growth remained subdued across most OECD countries in 2024, extending a slowdown observed since the mid-2000s. GDP per hour worked grew by 1.2 % in the OECD weighted average, but with a median of only 0.4 % across member countries, well below the 1.8% median recorded during 2001-07.
The United States continued to outperform most OECD countries, with labour productivity growth reaching 2.2% in 2024 while it was below 1 % in over half of OECD countries.
Whereas virtually all OECD countries were more productive in 2024 than two decades ago, the pace of improvement has roughly halved since the early 2000s. If pre-2007 trends had continued, productivity levels would be an estimated 10-20% higher today.
Economic growth in 2024 relied on increases in labour inputs rather than improvements in efficiency. Employment expansion acted as the primary driver of GDP growth in 18 countries out of the 21 countries with growth-accounting data. Multi-factor productivity (MFP) – the efficiency with which labour and capital are combined – contributed negatively to GDP growth in 14 countries.
OECD Compendium of Productivity Indicators 2026
Labour productivity lies at the heart of long-term improvements in material well-being. When productivity advances, countries can raise incomes without relying solely on longer working hours or larger workforces. When it stalls, those ambitions become harder to achieve.
This chapter examines labour productivity, measured as value added per hour worked, across OECD countries from two complementary perspectives. First, cross-country comparisons of productivity levels provide a snapshot of where countries stood in 2024 and how far they have moved in the most recent decade. Second, growth-accounting decompositions isolate the proximate sources of GDP growth and labour productivity growth, contrasting the most recent patterns with the 2000-19 average.
Labour productivity levels between the least and the most productive OECD countries differed by roughly a factor of eight in 2024
A cross-country comparison of labour productivity levels reveals wide differences across the OECD (Figure 2.1). Measured in 2020 constant price Purchasing Power Parities (PPP) dollars, GDP per hour worked in 2024 ranged from USD 18.7 in Colombia to USD 135.7 in Ireland, with the United States at USD 84.1. Excluding Ireland and Luxembourg (whose headline figures are inflated by multinational activities), Norway (USD 99.7) and Denmark (USD 92.2) led the ranking. Between 2014 and 2024, virtually all countries saw their productivity levels rise.
GDP per hour worked, USD 2020 PPP constant prices
Note: Bars show GDP per hour worked in constant 2020 USD PPPs for 2024. Countries are ordered by their 2024 GDP per hour level. Grey diamond markers show GNI per hour worked. OECD and EU-27 aggregates are total hours worked weighted averages.
Source: Authors’ calculations based on OECD Productivity Database and OECD National Accounts.
However, GDP per hour worked can overstate domestic productive capacity in economies where the activities of multinational enterprises create a large gap between GDP and Gross National Income (see Box 2.1). Ireland is the most prominent case: its GDP per hour worked ranks among the highest in the OECD, but the GNI-based measure, which strips out income repatriated abroad, is considerably lower. This caveat applies to varying degrees, to several other small open economies. GNI-adjusted figures therefore offer a useful complement for cross-country comparisons. Using GNI rather than GDP per hour worked also narrows the measured dispersion across OECD countries, reducing the gap between the highest and lowest productivity levels from a factor of around eight to around six.
Most OECD countries recorded single-digit labour productivity growth in 2024, with wide dispersion
Labour productivity growth remained modest across most OECD economies in 2024 (Figure 2.2). While the OECD aggregate - calculated as a weighted average using hours worked as weights – recorded growth of 1.3%, underlying country-level performance was in many cases much weaker: the median country posted growth of only 0.4%, and more than half of OECD countries recorded gains below 1%. The United States, at 2.2% in 2024, remained among the stronger performers, while a number of European economies clustered close to zero.
Annual growth in GDP per hour worked, percentage
Note: Bars show the year-on-year growth rate of GDP per hour worked in 2024. Countries are ordered from highest to lowest growth. OECD and EU-27 aggregates are total hours worked weighted averages.
Source: OECD Productivity Database.
Two features stand out. First, dispersion across countries remains sizeable even in a generally subdued year, indicating that national factors still shape short-run productivity outcomes. Second, observations cluster overwhelmingly in the low-single-digit range. This concentration motivates the longer-run perspective developed in the next section: to understand whether 2024 reflects temporary headwinds or a continuation of a deeper slowdown, the cross-section needs to be placed in the context of productivity growth in the early 2000s.
Conventional productivity measures referring to the total economy rely on GDP as the numerator for output. GDP captures the total gross value added produced within a country’s borders (plus taxes less subsidies on products), regardless of who owns the factors of production. In most OECD economies, the distinction between domestic production and domestically owned production is minor. In a handful of countries, however, the activities of multinational enterprises (MNEs) create a large wedge between the two.
Ireland is the most prominent case. Many of the world’s largest technology and pharmaceutical firms have established their European headquarters in Ireland. Key explanations for the strong presence of MNEs in Ireland include access to the EU single market, a favourable corporate tax environment, and a highly educated workforce. These firms book substantial revenues and profits in Ireland and, critically, hold large stocks of intellectual property (IP) on their Irish balance sheets. The associated depreciation of these IP assets inflates Irish GDP without a corresponding increase in domestic incomes or employment.
According to Ireland’s Central Statistics Office (CSO), modified gross national income (GNI*) (as opposed to the traditional GNI) which adjusts for the factor income of re-domiciled companies, the depreciation on imported R&D services and traded IP, and the depreciation on aircraft held by leasing companies, stood at just 57% of GDP in 2024 (EUR 321 billion versus EUR 562 billion) (CSO, 2025[1]). Luxembourg, where the financial sector generates large cross-border income flows, displays a similar though less extreme pattern.
The practical consequence for productivity measurement is that GDP per hour worked can substantially overstate the productive capacity of the domestic economy in affected countries. Measured in 2020 constant price PPP dollars, Ireland’s GDP per hour worked of USD 135.7 in 2024 ranks far above any other OECD country, but a GNI-based measure would place it considerably lower, closer to the levels observed in other high-income European economies (Figure 2.1). Profit shifting driven by differences in corporate tax regimes can inflate measured output and multifactor productivity in low-tax jurisdictions without any change in real production (Bricongne, Delpeuch and Lopez-Forero, 2023[2]).
GNI-based productivity measures offer a useful complement to GDP-based ones. GNI equals GDP minus primary income paid to the rest of the world (principally investment income repatriated by foreign-owned firms) and plus the primary income received from the rest of the world. It thus provides a closer approximation of the income accruing to domestic residents.
With GNI per hour worked series not routinely available on a harmonised basis for all OECD countries, GDP-based measures remain the standard in the OECD Productivity Database. Readers should bear these caveats in mind when interpreting cross-country comparisons, particularly for countries near the top of the GDP-per-hour ranking.
The OECD Compendium of Productivity Indicators (2025[3]) and the companion documentation on the revamped OECD Productivity Database (OECD, 2025[4]) discuss these measurement challenges in further detail.
The pace of improvement has halved since the early 2000s, and the United States has pulled further ahead
The evolution of GDP per hour worked in constant price PPP from the late 1990s to 2024 reveals three key patterns (Figure 2.3).
First, the United States has steadily pulled further ahead of other major economies. Measured in 2020 constant PPP dollars, output per hour rose from USD 50.8 in 1995 to USD 84.1 in 2024, representing a compound annual growth of 1.8%. The EU-27 and Japan also advanced over the same period, but at a slower pace: 1.1% per year in both cases. Over three decades, even modest growth-rate differentials compound into a substantial gap. The EU-27 moved from roughly 90% of the US level in 2000 to 75% in 2024; Japan fell from 73% to 62%.
GDP per hour worked, USD 2020 PPP constant prices
Note: GDP per hour worked is expressed in constant 2020 USD PPPs. OECD and EU-27 series are total hours worked weighted averages. Ireland and Luxembourg are excluded from the line plot, as their GDP per hour worked is inflated by multinational profit shifting and cross-border income flows respectively (see Box 2.1).
Source: OECD Productivity Database.
However, it is worth highlighting that cross-country productivity comparisons are sensitive to the choice of PPP benchmark year (Inklaar et al., 2022[5]; Bournot, Koechlin and Schreyer, 2011[6]). While the ranking of the United States above the EU-27 and Japan is robust across PPP vintages, the precise magnitude of the gap should be interpreted with caution. For a broader discussion, see Krugman (2026[7]) and Ackerman (2026[8]) on the distinction between current price and constant price PPP comparisons.
Second, the pace of gains has decelerated across the OECD. Median annual labour productivity growth across OECD countries was 2.0% during 2001-07, declined to 1.0% during 2008-19, and fell further to 0.8% during 2020-24. This median masks a pandemic-related spike in 2020 in countries where hours worked contracted more sharply than output, which temporarily inflated measured productivity. Different explanations for this deceleration have been explored in the literature and, while the jury is still out, there is broad agreement that the slowdown is not merely a statistical artefact. Syverson (2017[9]) provides an overview of the “productivity J-curve” debate, the idea that general-purpose technologies can initially depress measured productivity during periods of costly adoption and learning before eventually raising it. Cette et al. (2016[10]) examine the role of declining business dynamism. Gordon (2016[11]) and Bloom et al. (2020[12]) offer complementary perspectives on whether the slowdown reflects secular stagnation or delayed technology diffusion.
The cumulative effect of this deceleration is substantial. Based on OECD Productivity Database labour productivity levels, had the pre-2007 pace been sustained, GDP per hour worked in 2024 would have been approximately 10-12% higher across the major OECD economies – a gap of roughly USD 9-11 per hour worked at 2015 PPP prices. This underscores the extent to which the slowdown is a first-order constraint on living standards.
Third, behind the headline aggregates lies considerable diversity across countries. Korea nearly quadrupled its productivity level from USD 13.9 to USD 52.0 per hour worked, in constant 2020 PPPs, between 1995 and 2024. Boosted by the activities of multinational enterprises (see Box 2.1), Ireland saw an increase from USD 42.5 to USD 135.7 over this period. Poland’s trajectory illustrates the broader post-transition catch-up pattern in Central and Eastern Europe (Timmer et al., 2010[13]): labour productivity rose from USD 19.6 to USD 52.1, narrowing the gap with mid-ranking Western European economies. At the other end, Colombia (USD 18.7) and Mexico (USD 33.6) remain well below the OECD average. The dispersion of country trajectories illustrates both the scope for convergence and the persistence of structural productivity gaps.
These longer-run productivity developments should also be seen in the context of broader structural changes in OECD economies, including shifts in the sectoral composition of value added and labour input. Box 2.2 illustrates how the OECD Structural Analysis Database (STAN) can help analyse these patterns.
STAN is a harmonised industry-level database for long-run, cross-country analysis of productivity and structural change. It provides annual data from 1970 onwards for OECD countries on output, value added, labour input, investment and capital stock, based on annual national accounts.
Measuring structural change and comparing long-term trends
STAN supports analysis of long-term developments in key indicators such as labour productivity growth. It also allows for the analysis of shifts in the sectoral composition of value added and in the allocation of labour and capital across industries. One example is the rising share of information and communication and business services (ISIC Rev.4 I-M) in total value added (Figure 2.4). This reflects trends such as digitalisation, deeper global integration and the growing use of business services as production inputs. Across the OECD and the G7, these activities have followed a clear upward trend.
Value added of ISIC Revision 4 industries I-M as a percentage of total economy value added
Note: The OECD aggregate is computed as unweighted average. Due to data limitations, the OECD group used includes the following countries: Austria, Belgium, Canada, Czechia, Denmark, Spain, Finland, France, the United Kingdom, Italy, Japan, South Korea, the Netherlands, Norway, New Zealand, Portugal, Sweden and the United States.
Source: OECD, Annual National Accounts Database, Structural Analysis (STAN) Database, http://oe.cd/stan; and national sources.
Limitations and measurement challenges
Despite its value, STAN has important limitations. Extending series backwards often requires linking data across changes in industrial classifications, accounting standards, and benchmark revisions. In addition, while STAN mainly draws on countries’ annual national accounts by activity, more detailed industry estimates rely on supplementary sources whose quality and availability vary. As a result, gaps remain where no reliable estimates are possible, and industry detail is not fully comparable across countries. Country notes and metadata should be consulted when making cross-country comparisons. For further methodological detail, see Horvát and Webb (2020[14]).
Recent economic growth increasingly relied on additional labour inputs, rather than efficiency gains
Yet, labour productivity gains do not translate one-for-one into GDP growth: changes in employment, hours worked, and the working-age share of the population also shape the GDP trajectory. To understand what has been driving recent GDP growth and why it has not evolved in step with productivity, it is useful to turn to a structural decomposition.
Percentage points, 2024* versus 2000-19 average
Note: Contributions from total hours worked, capital services, and multifactor productivity (MFP) to GDP growth. Countries are ordered by total GDP growth from highest (left) to lowest (right). Solid bars show 2024, or the latest available year where 2024 is unavailable; countries using pre-2024 latest data are marked with an asterisk. Faded bars show the 2000-19 average. OECD and EU-27 aggregates are total hours worked weighted averages.
Source: OECD Productivity Database.
A standard growth-accounting framework decomposes real GDP growth into the contributions of three components: total hours worked (labour input); capital services, which measure the productive contribution of the capital stock weighted by asset type; and multifactor productivity (MFP), which captures the overall efficiency with which labour and capital are combined, including the effects of technological progress, organisational improvements, and resource reallocation across firms and sectors, but also measurement residuals and specification error. For readers less familiar with the framework, MFP can be thought of as the “residual” part of growth that cannot be attributed to simply using more labour or more capital (see also OECD/APO (2022[15]) for a discussion of this interpretation). Industry-level labour productivity and MFP developments are examined in Chapter 4.
Across the 21 countries with data in both periods, median GDP growth was 2% in the 2000-19 average and 1.2% in the latest year. More importantly, the composition of growth has shifted (Figure 2.5). During 2000-19, contributions from labour, capital, and MFP were typically more evenly shared than in the most recent period. While labour input growth was the primary driver of GDP growth in 18 out of 21 countries in 2024, MFP made a negative contribution in 14 countries.
This pattern carries clear economic implications. Expanding hours worked can support output in the near term, particularly during recoveries or in the context of tight labour markets. Economies that rely primarily on adding more inputs rather than using them more efficiently, however, face constraints as labour supply tightens, a concern of growing relevance in the context of population ageing, though one that can be partly offset through migration, rising labour force participation, and investment in human capital (OECD, 2024[16]).
The cross-country variation in the pattern of contributions is considerable. The United States retains positive contributions from all three components, around 1.1 p.p. for labour input and MFP and 0.7 p.p. for capital in 2023, the latest available year for this economy. Japan also performs comparatively well, with a strong MFP contribution of 1.1 p.p. By contrast, several European economies display a more fragile pattern: Estonia and Austria saw negative MFP sharply reduce GDP growth or pull it into contraction. One potential, yet only partial, explanation is offered by McGowan et al. (2017[17]), who show that weakening business dynamism and the persistence of low-productivity firms have contributed to MFP stagnation across OECD economies. Other factors, including weak investment in intangibles and slower diffusion of frontier technologies, have also been documented in the literature. Changes in work organisation, including the wider use of teleworking and hybrid work, may also affect productivity performance, with effects depending on job characteristics, management practices, digital tools and the extent to which remote work supports flexibility without weakening collaboration and knowledge diffusion (Criscuolo et al., 2021[18]).
The broader implication is that the post-pandemic period has not restored the earlier, more balanced growth model. In many economies, output expansion has come to depend on adding hours rather than on raising the productivity of those hours. Reactivating drivers of growth beyond labour inputs will therefore be important going forward (OECD, 2026[19]).
References
[8] Ackerman, S. (2026), Europe’s productivity keeps outpacing the US, https://sethackerman.substack.com/p/europes-productivity-keeps-outpacing.
[17] Adalet McGowan, M., D. Andrews and V. Millot (2017), “The Walking Dead? Zombie firms and productivity performance in OECD countries”, OECD Economics Department Working Papers, No. 1372, OECD Publishing, Paris, https://doi.org/10.1787/180d80ad-en.
[12] Bloom, N. et al. (2020), “Are ideas getting harder to find?”, American Economic Review, Vol. 110/4, pp. 1104-1144, https://doi.org/10.1257/aer.20180338.
[6] Bournot, S., F. Koechlin and P. Schreyer (2011), “2008 Benchmark PPPs measurement and use”, OECD Statistics Briefs, Vol. 17, https://www.oecd.org/content/dam/oecd/en/data/insights/data-explainers/2024/6/Statistics-Brief-2008%20Benchmark-PPPs-measurement-and-use-March%202008.pdf.
[2] Bricongne, J., S. Delpeuch and F. Lopez-Forero (2023), “Productivity slowdown and tax havens: Where is measured value creation?”, Journal of International Economics, Vol. 143, p. 103757, https://doi.org/10.1016/j.jinteco.2023.103757.
[10] Cette, G., J. Fernald and B. Mojon (2016), “The pre-Great Recession slowdown in productivity”, European Economic Review, Vol. 88, pp. 3-20, https://doi.org/10.1016/j.euroecorev.2016.03.012.
[18] Criscuolo, C. et al. (2021), “The role of telework for productivity during and post-COVID-19: Results from an OECD survey among managers and workers”, OECD Productivity Working Papers, No. 31, OECD Publishing, Paris, https://doi.org/10.1787/7fe47de2-en.
[1] CSO (2025), Annual National Accounts 2024, Central Statistics Office, https://www.cso.ie/en/releasesandpublications/ep/p-ana/annualnationalaccounts2024/.
[11] Gordon, R. (2016), The Rise and Fall of American Growth, Princeton University Press, https://doi.org/10.1515/9781400873302.
[14] Horvát, P. and C. Webb (2020), “The OECD STAN Database for industrial analysis: Sources and methods”, OECD Science, Technology and Industry Working Papers, No. 2020/10, OECD Publishing, Paris, https://doi.org/10.1787/ece98fd3-en.
[5] Inklaar, R. et al. (2022), Inconsistencies in Cross-Country Price Comparisons over Time: Patterns and Facts, Springer, https://doi.org/10.1007/978-981-19-2023-3_13.
[7] Krugman, P. (2026), Europe v America: Who’s Really Winning?, https://paulkrugman.substack.com/p/europe-v-america-whos-really-winning.
[19] OECD (2026), Foundations for Growth and Competitiveness 2026, OECD Publishing, Paris, https://doi.org/10.1787/40a7532f-en.
[3] OECD (2025), OECD Compendium of Productivity Indicators 2025, OECD Publishing, Paris, https://doi.org/10.1787/b024d9e1-en.
[4] OECD (2025), The revamp of the OECD Productivity Database, OECD Publishing, Paris, https://doi.org/10.1787/f07c55d4-en.
[16] OECD (2024), OECD Digital Economy Outlook 2024 (Volume 1): Embracing the Technology Frontier, OECD Publishing, Paris, https://doi.org/10.1787/a1689dc5-en.
[15] OECD/APO (2022), Identifying the Main Drivers of Productivity Growth: A Literature Review, OECD Publishing, Paris, https://doi.org/10.1787/00435b80-en.
[9] Syverson, C. (2017), “Challenges to mismeasurement explanations for the US productivity slowdown”, Journal of Economic Perspectives, Vol. 31/2, pp. 165-186, https://doi.org/10.1257/jep.31.2.165.
[13] Timmer, M. et al. (2010), Economic Growth in Europe: A Comparative Industry Perspective, Cambridge University Press, https://doi.org/10.1017/cbo9780511762703.