OECD Compendium of Productivity Indicators 2025

6. Industry-level multifactor productivity growth

Context and key findings

Multifactor productivity (MFP) is an estimate of the overall efficiency with which inputs are used in the production process. It is calculated as a residual, using a growth accounting framework developed by Solow (1957[1]). This framework enables the decomposition of changes in economic output – in this case gross value added (GVA) – into contributions from changes in labour and capital inputs, and a residual known as MFP. More information on its derivation can be found in OECD (2025[2]).

Being a residual, changes in MFP reflect a mix of factors such as technological change, changes in capacity utilisation or economies of scale, or measurement errors. MFP is often associated with technological progress. Innovations, such as the use of Information and Communication Technology in production, alongside complementary assets like organisational and human capital, enables a more efficient production process (Bajgar, Criscuolo and Timmis, 2021[3]; André and Gal, 2024[4]). For instance, the adoption of computers and the expansion of the internet are considered as a key driver of MFP growth in OECD countries during the period from the mid-1990s to the early 2000s (Kretschmer, 2012[5]; Cette, Devillard and Spiezia, 2020[6]).

MFP estimates may also capture measurement errors and are particularly sensitive to how input and output volumes are measured. For instance, inaccuracies in measuring input volumes can arise when differences in input quality are not accounted for, or when various types of investment assets are not properly distinguished (Pionnier, Zinni and Baret, 2023[7]). Similarly, measuring output volumes presents challenges – especially when related to households, the non-market sector, and the informal economy – where gauging output values is often difficult (OECD/APO, 2022[8]).

MFP growth can also reflect cyclical fluctuations in the economy, and it has often been found to be pro-cyclical, i.e. declining during economic downturns and rising during expansions. This pattern is partly due to unmeasured variations in the utilisation of labour and capital over the business cycle (OECD, 2015[9]). Since adjusting inputs can be costly, firms often respond to short-term changes in demand by altering the intensity of use of existing resources – for instance, by hoarding labour during downturns or underutilising capital without immediately reducing capacity. Hence, adjusting MFP estimates to account for these cyclical effects enhances their usefulness by isolating the structural component of productivity growth, including technological progress (Fernald, 2014[10]; Martin and Jones, 2023[11]; Comin et al., 2025[12]) and Chapter 8 on cyclical adjustment of MFP).

Key findings

• Total economy MFP growth either stagnated or even turned negative in most countries with available data for 2023 (Figure 6.1). The largest declines were recorded in Austria and Luxembourg, where MFP growth fell by around -2%. In contrast, the highest MFP growth in 2023 was observed in the Slovak Republic and Slovenia, reaching approximately 4.5% and 2%, respectively. Most countries with available data in 2023 recorded a lower MFP growth than the long-run average MFP growth across OECD countries, with the exception of the Slovak Republic, Slovenia, and Japan. Against the backdrop of the introduction of ChatGPT and Gemini in early 2022, the subdued MFP performance suggests that any productivity gains from generative AI have yet to materialise (Filippucci et al., 2024[13]).

• However, the economy-wide average masks significant variation across industries. Mining and quarrying and financial services recorded negative MFP growth in most countries in 2023. The decline of MFP in these industries was mainly driven by a large fall in GVA coupled with limited adjustment in inputs - namely hours worked and capital services. Similar patterns were reported in Australia and Canada for these industries (Australian Bureau of Statistics, 2023[14]; Statistics Canada, 2025[15]). Manufacturing showed little to no MFP growth, except in the Slovak Republic and Luxembourg, which saw a notable increase. In contrast, the energy and information and communication industries experienced increases in MFP in more than half of the countries with data. Nevertheless, their positive contribution remained modest due to the relatively small weight of these industries in the total economy in most countries (Figure 6.2).

• Preliminary industry-level data for 2023 indicate that in most business economy industries, MFP growth was lower than prior to the COVID crisis (2010-2019). Notable exceptions were the information and communication industry, where MFP growth continued to outpace pre-COVID levels in more than half of the countries with data (Figure 6.3). The deceleration in MFP growth relative to the pre-COVID period was evident as early as 2022 across several business economy industries, with the trend intensifying in 2023. In mining and quarrying as well as manufacturing, the post-COVID rebound in 2021 proved short-lived, as MFP gains decelerated steadily over the following two consecutive years. Similarly, activities such as finance and professional activities experienced a loss of momentum, with MFP growth in 2023 falling below the relatively stronger performance seen in 2022. The only notable exception is the information and communication industry, where MFP growth in 2023 remained above pre-pandemic levels in half of the countries.

• MFP growth explained the overwhelming part of labour productivity growth in 2023 in most countries and industries. The dominance of MFP contribution over capital deepening has been evident in the past for long (OECD/APO, 2022[8]). For the G7 average, the sharpest declines in MFP growth were observed in mining and trade services (including wholesale and retail trade, transportation, and accommodation and food services). By contrast, the information and communication industry registered a positive MFP contribution to labour productivity growth of around 1.8 percentage points.

• The contribution of capital deepening, while still being negative, showed a slight improvement compared to its level in 2022 (Figure 6.4). The stagnation of capital deepening growth in 2023, in particular within the G7, was partly driven by the slowdown of capital services. Capital services, which reflect accumulated investment net of depreciation from wear, tear, and obsolescence, have stagnated in part due to weak investment trends in these countries (Chapter 4). In addition, their rate of depreciation has been increasing, in part because of the rising share of asset classes with shorter service lives, such as ICT and intangible assets (Sicsic, 2018[16]).

Figure 6.1. MFP growth for the total economy

2023, Per cent

Note: MFP estimates are produced for countries that provide detailed industry-level data on investment, capital stock, and hours worked, and that participated in the one-off OECD data collection on asset depreciation and retirement profiles. The dotted line represents the average MFP growth across OECD countries with available data over 2010-19.

Source: OECD Productivity database (2025).

Statlink

Figure 6.2. MFP growth across industries compared to the total economy

2023 or latest available year, Per cent

Note: Industry-level MFP estimates are produced for countries that provide detailed industry-level data on investment, capital stock, and hours worked, and that participated in the one-off OECD data collection on asset depreciation and retirement profiles. Countries for which data refers to 2022: Australia, Estonia, the United Kingdom, Croatia, Italy, Lithuania, Norway. Countries for which data refers to 2021: Spain, New Zealand and the United States.

Source: OECD Productivity database (2025).

Statlink

Figure 6.3. MFP growth in selected industries across G7 countries

Per cent

 

Note: 2023 figures do not include data for United Kingdom, Italy, and United States, whose data have not been released at the time of writing.

Source: OECD Productivity database (2025).

Statlink

Figure 6.4. MFP contribution to labour productivity in selected industries: G7 average

Percentage points

 

Note: 2023 figures do not contain data for United Kingdom, Italy, and United States, as data for these countries have not yet been released.

Source: OECD Productivity database (2025).

Statlink

Data sources

OECD Productivity Database: http://data-explorer.oecd.org/s/1v8.

References

[4] André, C. and P. Gal (2024), “Reviving productivity growth: A review of policies”, OECD Economics Department Working Papers 1822, https://doi.org/10.1787/61244acd-en.

[14] Australian Bureau of Statistics (2023), Estimates of Industry Multifactor Productivity.

[3] Bajgar, M., C. Criscuolo and J. Timmis (2021), “Intangibles and industry concentration: Supersize me”, OECD Science, Technology and Industry Working Papers 2021/12, https://doi.org/10.1787/ce813aa5-en.

[6] Cette, G., A. Devillard and V. Spiezia (2020), “Growth factors in developed countries: A 1960-2019 growth accounting decomposition”, Banque de France Working Paper.

[12] Comin, D. et al. (2025), “Revisiting Productivity Dynamics in Europe: A New Measure of Utilization-Adjusted TFP Growth”, Journal of the European Economic Association, https://doi.org/10.1093/jeea/jvaf003.

[10] Fernald, J. (2014), “A Quarterly, Utilization-Adjusted Series on Total Factor Productivity”, Federal Reserve Bank of San Francisco Working Paper 2012-19.

[13] Filippucci, F. et al. (2024), “The impact of Artificial Intelligence on productivity, distribution and growth: Key mechanisms, initial evidence and policy challenges”, OECD Artificial Intelligence Papers 15, https://doi.org/10.1787/8d900037-en.

[5] Kretschmer, T. (2012), “Information and Communication Technologies and Productivity Growth: A Survey of the Literature”, OECD Digital Economy Papers, Vol. 195, https://doi.org/10.1787/5k9bh3jllgs7-en.

[11] Martin, J. and K. Jones (2023), “An Occupation and Asset-Driven Approach to Capital Utilization Adjustment in Productivity Statistics”, NBER Chapters, in: Technology, Productivity, and Economic Growth, National Bureau of Economic Research, Inc.

[2] OECD (2025), The revamp of OECD Productivity Database, https://doi.org/10.1787/f07c55d4-en.

[9] OECD (2015), The Future of Productivity, OECD Publishing, Paris, https://doi.org/10.1787/9789264248533-en.

[8] OECD/APO (2022), Identifying the Main Drivers of Productivity Growth: A Literature Review, OECD Publishing, Paris, https://doi.org/10.1787/00435b80-en.

[7] Pionnier, P., B. Zinni and K. Baret (2023), “Measuring Capital and Multifactor Productivity: The Role of Asset Depreciation and Initial Capital Stock Estimates”, International Productivity Monitor, Vol. 45, pp. 155–177, https://ideas.repec.org/a/sls/ipmsls/v45y20236.html.

[16] Sicsic, P. (2018), A lower net investment rate than at the end of the 1990s, https://www.banque-france.fr/en/publications-and-statistics/publications/lower-net-investment-rate-end-1990s.

[1] Solow, R. (1957), “Technical Change and the Aggregate Production Function”, The Review of Economics and Statistics, Vol. 39/3, p. 312, https://doi.org/10.2307/1926047.

[15] Statistics Canada (2025), Multifactor productivity growth estimates and industry productivity database.