In 2024, labour productivity growth varied widely across industries in OECD economies, with mining and agriculture showing the largest cross-country dispersion. Aggregate productivity growth in manufacturing reflected the influence of key industries, with Denmark recording outsized gains driven by pharmaceutical value-added growth.
Between 2019 and 2024, information and communication services and professional and administrative services each gained around 0.8 percentage points in their share of total hours worked across 32 OECD countries. Manufacturing and agriculture each saw a decline of a similar magnitude, continuing a long-term structural shift toward knowledge-intensive services.
Overall labour productivity growth in 2023-24 was mainly driven by labour productivity gains within industries. A smaller, though positive, contribution came from shifts in the composition of total hours worked across industries.
Within industry labour productivity gains in 2023-24 largely reflected Multifactor Productivity (MFP) growth, capturing improvements in the overall efficiency with which labour and capital inputs are combined in the production process. Considerable cross country variation in MFP performance translated into differing labour productivity outcomes.
MFP growth in services generally exceeded the total economy average in 2023-24, whereas weaker performance in goods-producing industries weighed on overall productivity outcomes.
MFP growth has continued the downward trend observed since the 2000s, alongside increasing dispersion across countries. These differences in MFP growth are most pronounced in digital-intensive industries, possibly reflecting uneven cross-country adoption of advanced technologies.
OECD Compendium of Productivity Indicators 2026
Aggregate labour productivity growth is shaped both by the performance in individual industries and by the reallocation of resources between them. Since the mid-2000s, productivity growth has slowed in most OECD economies, but the slowdown has been unevenly distributed across industries. Understanding how productivity evolves at the industry level is therefore essential for designing policies that promote innovation, technology adoption and resource reallocation.
This chapter examines recent productivity patterns using data from the OECD Productivity Database, primarily using the 21-industry breakdown available for a broad set of countries and, where possible, a more detailed 38-industry breakdown. This allows the chapter to examine not only broad differences across industries, but also the role of specific industries, such as pharmaceuticals, in shaping aggregate outcomes. The most recent year with sufficiently broad coverage for industry‑level productivity analysis is 2024.
This chapter is organised in three parts. The first examines recent labour productivity developments across industries, highlighting cross-country dispersion in 2024, the key role of specific activities, and measurement issues in services. The second part decomposes aggregate labour productivity growth into two components: productivity changes within industries and effects of shifts in the composition of total hours worked across industries. The final section turns to industry‑level Multifactor Productivity (MFP). MFP captures the component of labour productivity growth that is not explained by capital deepening, that is, by increases in the amount of capital available per hour worked. The discussion explores differences in MFP growth across industries and the contribution of MFP to overall labour productivity growth.
From mining to services: A tale of divergent productivity distributions
Labour productivity growth varied substantially across industries and countries in 2024 (Figure 4.1). Labour productivity growth was most similar across countries in trade, transport and accommodation and in professional services. One possible explanation is that these are relatively large and diversified industries, which may mechanically reduce the likelihood of extreme annual productivity movements. In professional services, firms’ continued need for services that support core business operations may also help stabilise activity over short horizons (Ariu, 2016[1]).
Manufacturing occupied an intermediate position, with most countries recording modest growth, while Denmark stood out as a positive outlier linked to pharmaceuticals (discussed below). Dispersion was higher in information and communication services, possibly reflecting differences in the composition of investment (see Chapter 3) and regulatory frameworks (Bourlès et al., 2013[2]) and country-specific labour market adjustment patterns, including labour hoarding in Luxembourg (STATEC, 2025[3]).
Mining and quarrying and agriculture showed the widest dispersion. Regarding mining and quarrying, Hungary and Slovenia posted extreme gains while Spain, where this industry accounts for a small share of the economy, recorded a sharp contraction. In the case of Agriculture, Luxembourg and Estonia saw improvements in labour productivity. In contrast, labour productivity in this industry declined in France in 2024, reflecting exceptionally poor harvests driven by adverse weather (INSEE, 2025[4]).
Annual percentage change in Gross Value Added (GVA) per hour worked
Note: The figure displays the interquartile range and median of labour productivity growth (GVA per hour worked) across available OECD countries for each industry. Each box represents the 25th to 75th percentile range; the line inside the box marks the median. Whiskers extend to the 10th and 90th percentiles. For each industry, the country with the highest and lowest labour productivity growth is labelled. The x-axis is truncated at −35% to +55% for readability; where a country’s value falls outside this range, the dot is pinned to the axis edge and the actual value shown in brackets. Coverage: 25 OECD countries (plus Bulgaria, Croatia and Romania). OECD countries not included due to data unavailability: Australia, Chile, Colombia, Costa Rica, Iceland, Israel, Japan, Korea, New Zealand, Switzerland, Türkiye, United Kingdom, United States. Industry groupings follow ISIC Rev. 4. “Trade, transport and accommodation” covers wholesale and retail trade (G), transportation and storage (H), accommodation and food services (I), arts, entertainment and recreation (R), and other service activities (S). “Professional services” covers professional, scientific and technical activities (M) and administrative and support services (N). “Utilities” covers electricity, gas and water supply (D and E).
Source: OECD Productivity Database.
Pharmaceutical value added in Denmark and Ireland drove outsized manufacturing productivity gains
Pharmaceutical productivity surged in Denmark in 2023 and 2024, driven by GLP-1 therapies (Figure 4.2). This single industry pushed overall Danish manufacturing productivity growth well above the OECD median. Disregarding pharmaceuticals would bring Denmark’s manufacturing labour productivity growth rate closer to the European Union mean, underscoring the importance of considering the role of individual industries in shaping aggregate productivity dynamics.
Yet it is worth noting that this industry is particularly sensitive to measurement choices due to the central role of large pharmaceutical MNEs in organising production across borders. Depending on how national statistical offices classify these global production arrangements, value added can be attributed differently across industries and countries (Kuchler et al., 2025[5]; Office for National Statistics (ONS), 2025[6]).
Annual percentage change in real GVA per hour worked
Note: Labour productivity is measured as gross value added per hour worked at constant prices. The pharmaceutical industry corresponds to ISIC Rev. 4 Division 21 (Manufacture of basic pharmaceutical products and pharmaceutical preparations). The OECD median is computed across 24 countries with available data.
Source: OECD Productivity Database.
Hours worked continued to shift toward knowledge-intensive services at the expense of manufacturing and agriculture
Beyond year-to-year productivity changes within industries, the distribution of hours worked across industries has also shifted notably over the past five years, reflecting broader structural changes (Figure 4.3).
Percentage point change in share of total hours worked (panel A) and in share of total employment (panel B)
Note: Industry shares are computed as each industry’s share of total hours worked in the economy. Industries are classified according to ISIC Rev. 4 at the A10 level. Coverage: 29 OECD countries (plus Bulgaria, Croatia and Romania). OECD countries not included: Colombia, Costa Rica, Iceland, Japan, Korea, New Zealand, Switzerland, Türkiye, United Kingdom. Industry groupings follow ISIC Rev. 4. “Trade, transport and accommodation” covers wholesale and retail trade (G), transportation and storage (H), accommodation and food services (I), arts, entertainment and recreation (R), and other service activities (S). “Professional services” covers professional, scientific and technical activities (M) and administrative and support services (N). “Utilities” covers electricity, gas and water supply (D and E).
Source: OECD National Accounts; OECD Productivity Database.
On average across 32 OECD countries with available data, professional services and information and communication services have expanded their share of total hours worked, while manufacturing and agriculture have continued a long-term decline in their shares, as documented in earlier editions of the Compendium (OECD, 2024[7]; OECD, 2025[8]). These structural shifts shape the between-industry component of aggregate labour productivity growth, which captures productivity effects from the reallocation of labour inputs across industries with differing productivity performance.
Across industries, changes in hours worked and employment generally moved in the same direction over 2019-2024, although the magnitude of these changes was not always the same (Figure 4.3). In the case of trade, transport and accommodation, the employment share fell by around 0.2 percentage points while the hours-worked share fell by nearly 0.6 percentage points. This points to an industry-wide decline of approximately 29 hours per worker per year, the largest of any industry. This pattern is consistent with the post-pandemic expansion of part-time and flexible contracts in hospitality and retail, as well as the growing role of platform-based work in transport and delivery (OECD, 2024[7]).
In contrast, professional services as well as information and communication recorded larger gains in their share of hours worked than in their employment share, reflecting labour shortages that led firms to extend the working time of existing staff rather than hire proportionally (OECD, 2024[9]).
As the economic weight of services continues to increase across OECD countries, service-sector labour productivity needs to be interpreted with care. Productivity measures in services are particularly sensitive to measurement challenges, including changes in quality, digital delivery models and the measurement of labour inputs (Box 4.1).
Labour productivity is commonly measured as real gross value added per hour worked. While simple in principle, the indicator must be interpreted with sector‑specific measurement challenges in mind. In services, the link between production, prices, and real value added is often less direct: many services are intangible, customised, and difficult to quality-adjust, and a growing share is also highly exposed to digitalisation and rapid technological change (OECD, 2001[10]).
Prices of non-market services are not observed, and output is typically proxied by inputs, which by construction limits measured productivity growth. Regarding health, education, care and personal services, measurement is particularly challenging because service quality is difficult to define, especially when preventive measures or complementary innovations (such as pharmaceuticals) influence outcomes independently of the service itself.
In the case of market services, value is increasingly generated through quality improvements, faster logistics, better advice, more convenient platforms, and more effective business support, which may not be fully captured in measured real value added. Financial services pose a further problem: a substantial share of output is measured indirectly through implicit financial services on loans and deposits. As a result, measured value added can be sensitive to interest‑rate movements and may display counterintuitive cyclical patterns.
Digitalisation is reshaping how services are produced, delivered and priced, with direct implications for measurement. A growing concern is the treatment of quasi-free or zero-price digital services such as search, social media and communication platforms, which deliver welfare to users but generate revenue indirectly and are therefore imperfectly reflected in GDP (Byrne, 2022[11]; Coyle and Nguyen, 2023[12]). Cloud computing changes the boundary between investment and intermediate consumption (OECD, 2019[13]). Moreover, platform-mediated and occasional self-employment is difficult to capture in conventional statistics (OECD/ILO/European Union, 2023[14]) and constructing appropriate deflators for fast-changing services such as telecommunications remains a challenge (Abdirahman et al., 2022[15]). Together, these issues mean that measured productivity in digitally intensive services may understate true gains.
This section explores the contributions of different industries to overall labour productivity growth. Overall labour productivity growth in this section refers to the total economy, excluding real estate, public administration and defence, education, and health activities. While the previous section focusses solely on within-industry productivity developments, this section complements it by considering the role of labour reallocation across industries. Further details on the decomposition of industry-level labour productivity growth are provided in the methodology document.1
Industry-level labour productivity growth can be decomposed into within-industry and between-industry contributions. The within-industry component captures productivity developments occurring inside individual industries, such as technological progress, efficiency gains, or the exit of less productive firms. In contrast, the between-industry component captures changes in the distribution of total hours worked across different industries with varying productivity levels and growth rates. Aggregate productivity growth can weaken when a rising share of total hours worked concentrates in relatively low‑productivity industries such as accommodation and food services. Such a shift can stem from both movements of workers into these activities and changes in hours worked by workers who stay within the same industry. Likewise, a rising share of hours worked in industries experiencing weak or negative productivity growth will weigh on overall performance.
Within-industry effects, rather than structural shifts across industries, have been the primary driver of productivity growth in the period following the COVID-19 pandemic (OECD, 2025[8]). This section examines the relative importance of these channels in explaining recent productivity performance across countries. The analysis covers the period from 1995 up to 2024 for countries where sufficiently granular data are available, and up to 2023 for the remaining countries.
Productivity growth continues to be driven by within-industry improvements
Over the period 2023-24, the within-industry effect, which captures improvements in productivity within industries, acted as the main driver of aggregate labour productivity growth – in line with the pattern observed in recent years. This pattern is visible in most countries, with the exception of Australia, Bulgaria, and Croatia, where the between-industry effect plays a more prominent role in shaping aggregate productivity outcomes (Figure 4.4).
Costa Rica, Ireland and the Slovak Republic saw strong gains in within-industry performance over 2023-24, contributing to significant increases in overall labour productivity. On the contrary, declines in within-industry productivity were most pronounced in Estonia, Luxembourg, and Austria, weighing on aggregate productivity performance. This considerable heterogeneity across countries highlights the importance of examining the industry composition of productivity developments, which is explored in the next section.
Percentage points
Note: Labour productivity growth is measured based on hours worked for the total economy, excluding real estate, public administration and defence, education, and health activities. The decomposition is conducted at the A38 industry level for most countries, with the exception of Australia, Germany, Spain, France, Latvia, Poland, Portugal, Romania, and Sweden, for which it is performed at the more aggregated A21 level. Figures for Japan and Korea correspond to employment-based labour productivity due to the lack of industry-level hours worked data. Data refer to 2023-24 average for most countries, except for Australia, Colombia, Costa Rica, Japan, Peru, and the United Kingdom where data refers to 2023 and Korea to 2022.
Source: Authors’ calculations based on OECD Productivity Database.
Uneven contributions of the within-industry effect across countries
A decomposition of the within-industry effect over the 2023-24 period reveals substantial cross-country diversity in the set of industries driving labour productivity growth, with no clear common pattern. Notably, industries that underpinned productivity gains in some countries contributed negatively in others. These differences likely reflected differences in countries’ industrial composition, firm dynamics, broader business environment conditions, as well as countries’ positions in the business cycle (Figure 4.5).
Percentage points
Note: The category “Finance” encompasses finance and insurance services, corresponding to section K of ISIC Rev 4. “Trade, transport, accommodation” covers wholesale and retail trade (Section G), transportation (Section H), as well as accommodation and food services (Section I). Manufacturing corresponds to section C, while information and communication correspond to section J of ISIC Rev. 4. “Others” refers to all remaining industries. Figures are employment-based for Japan and Korea and hours worked-based for other countries. Data for most countries are reported as the 2023-24 average. In the absence of 2024 data, figures correspond to 2023 for Australia, Colombia, Costa Rica, Japan, Peru and the United Kingdom, and to 2022 for Korea.
Source: Authors’ calculations based on OECD Productivity Database.
Manufacturing
The within-industry effect of manufacturing played an important role in driving overall productivity growth over the 2023-24 period in a number of countries (Figure 4.6). Denmark stands out, with a contribution about twice as large as in other countries, mainly driven by strong performance in the pharmaceutical industry (Figure 4.2). In OECD economies, these activities are typically associated with high productivity, as they create significant value added for domestic firms while relying on production located abroad and limited domestic labour input (Kuchler et al., 2025[5]; Ishi, 2024[16]).
By contrast, manufacturing made negative contributions to labour productivity growth in several European Union countries, with the largest declines observed in Austria and Germany. This weakening likely reflects both structural and cyclical factors, including the shifts in consumer preferences, lingering energy cost pressures, and subdued demand over 2023-24 (Reinstaller and Sellner, 2024[17]; Weyerstraß et al., 2025[18]). In particular, demand for traditional manufacturing products – especially in automotive-related activities – has softened amid changing global demand patterns and the transition towards electric vehicles, weighing on production in countries such as Germany, where these industries play a central role. At the same time, persistently high energy prices have continued to strain manufacturing across Europe. Over 2023-24, electricity prices remained about twice their pre-COVID-19 levels and significantly higher than in the United States, increasing production costs for energy-intensive industries, undermining competitiveness, and reducing households’ purchasing power, thereby further dampening demand (Heussaff, 2024[19]; OECD, 2025[20]).
Percentage points
Note: Manufacturing corresponds to section C of ISIC Rev 4. Data refer to 2023 for Australia, Colombia, Costa Rica, Peru, and the United Kingdom.
Source: OECD Productivity Database.
Information and communication services
Over the 2023-24 period, the contribution of information and communication services to overall labour productivity growth was mixed, with more than half of OECD countries recording positive contributions while the remaining countries experienced negative impacts. Information and communication services contributed most strongly to aggregate labour productivity growth in the United States and the United Kingdom, adding around 0.7 percentage points. For the European Union, the average contribution of this industry remained broadly flat, reflecting offsetting developments across member states. For example, Denmark recorded a positive contribution of around 0.5 percentage points, whereas the impact was negative in Croatia and Poland, at approximately 0.7 percentage points (Figure 4.7).
Percentage points
Note: Information and communication correspond to section J of ISIC Rev 4. Data refer to 2023 for Australia, Colombia, Costa Rica, Peru, and the United Kingdom.
Source: OECD Productivity Database.
The recent acceleration in AI invites comparisons with the dot-com boom of the late 1990s and raises the question whether it is already translating into labour productivity gains, particularly in the Information and Communication Technology (ICT) industries, which encompass both information and communication services and ICT manufacturing. The positive productivity developments within information and communication services over the period 2023-24 might offer early, albeit tentative, indications of AI’s productivity-enhancing potential. Yet, the scale of these productivity gains during the 2023-24 period remained within historical ranges and stayed below the peaks recorded during the dot-com period (1995-2000). According to the OECD Productivity database, labour productivity growth in these services averaged 1.6% across the OECD over 2023-24, compared with 4.6% during 1995-2000.
Nevertheless, recent labour productivity performance in information and communication services outpaced pre-COVID trends in around half of OECD countries. For example, in Germany, labour growth in the information and communication services reached 3.5% in 2023-2024, compared with 3.2% in the 2010-2019, while in the United Kingdom it stood at 7.8%, higher than the 6% recorded over the previous period. An optimistic reading of the recent productivity performance in this industry across many OECD countries, together with rising investment in ICT assets (see Chapter 3), provides encouraging signs that the adoption of new digital technologies may be supporting productivity improvements.
The negative development of labour productivity growth in information and communication services in a few countries, such as Croatia, Poland, and Luxembourg during the 2023-24 period, resulted from stronger growth in hours worked than in output. This increase in hours worked was accompanied by strong employment growth in these activities, suggesting that labour input grew primarily through the hiring of additional workers rather than longer working hours per employee (Figure 4.3). The negative labour productivity growth in information and communication in these countries, despite strong employment growth, may be linked to constraints in the supply of high-skilled labour. This, in turn, tends to limit firms’ ability to fully utilise new hires and capitalise on digital technologies (OECD, 2025[21]; MPF, 2024[22]). At the same time, investment in ICT assets has remained subdued in these countries over the last few years (see Chapter 3). The combination of rising employment and weak investment has led to limited capital deepening, further weighing on ICT labour productivity growth in these countries.
In contrast, productivity gains in information and communication services were sizable in the United Kingdom over the same period. This reflected strong growth in the value added produced by this industry, which has outpaced overall economic growth in the same period (OECD, 2024[7]). This performance appears to be supported by the role of digital activities in the economy, along with the continued uptake of digital technologies across firms (El-Rayyes et al., 2024[23]; Benz, Jaax and van Lieshout, 2024[24]).
The between-industry effect was mainly positive but modest
Over the period 2023-24, estimates indicate a positive, yet limited contribution of the between-industry component – defined as productivity effects from labour reallocation across industries – to overall labour productivity growth in around three quarters of OECD countries, generally ranging between -1 and 1 percentage point (Figure 4.8). Denmark, Costa Rica and Peru stood out as exceptions, recording a comparatively strong positive effect of about 1.6 percentage points, while Lithuania experienced a pronounced negative contribution of around -2.3 percentage points.
The modest between-industry effect is expected under normal economic conditions and is consistent with patterns observed in recent years (OECD, 2025[8]; OECD, 2026[25]).
Percentage points
Note: The category “Finance” encompasses finance and insurance services, corresponding to section K of ISIC Rev 4. “Trade, transport, accommodation” covers wholesale and retail trade (Section G), transportation (Section H), as well as accommodation and food services (Section I). Manufacturing corresponds to section C, while information and communication correspond to section J of ISIC Rev. 4. “Others” refers to all remaining industries. Figures for Japan and Korea are computed based on employment based while it is hours worked based for other countries. Data for most countries are reported as the 2023-2024 average. In the absence of 2024 data, figures correspond to 2023 for Australia, Colombia, Costa Rica, Japan, Peru and the United Kingdom, and to 2022 for Korea.
Source: Authors’ calculation based on OECD Productivity Database.
Manufacturing
In most countries, manufacturing recorded a positive between-industry effect over the same period. However, the overall impact was generally modest. Denmark stands out as a notable exception, with the contribution of manufacturing to the between-industry effect more than three times larger than that of the next highest country. This reflects a strong increase in manufacturing hours in Denmark, particularly in pharmaceuticals, where hours worked rose by about 17% in 2024 (Figure 4.9).
Percentage points
Note: Manufacturing corresponds to section C of ISIC Rev 4. Data refer to 2023 for Australia, Colombia, Costa Rica, Peru, and the United Kingdom.
Source: OECD Productivity Database.
Information and communication services
Over the period 2023-24, information and communication services generated a positive between-industry effect in more than half of the countries with available data, driven by a relative increase in the share of hours worked in these highly productive activities (Figure 4.10). In contrast, in countries that observed negative contributions, this resulted from a reduction in hours worked in this industry relative to the total economy. The strongest positive effect was recorded in Croatia, whereas the United Kingdom and Greece saw the largest negative contributions.
Percentage points
Note: Information and communication correspond to section J of ISIC Rev 4. Data refer to 2023 for Australia, Colombia, Costa Rica, Peru, and the United Kingdom.
Source: OECD Productivity Database.
In Croatia, hours worked in these services rose by around 20% over the same period. The growing share of hours worked in these digital-intensive activities in this country reflects partly policy-driven digital transformation, supported by substantial public investment aimed at upgrading infrastructure and strengthening the integration of digital technologies across sectors, including education (OECD, 2023[26]).
Despite this increase in labour input in Croatia, the output growth in the same industry remained relatively modest over the same period. Although it is premature to draw firm conclusions about these uneven growth trajectories, the pattern could potentially reflect a J-curve pattern, where the benefits of technology adoption materialise only gradually during the initial phase of diffusion (Brynjolfsson, Rock and Syverson, 2021[27]).
However, this analysis of the role of labour reallocation across industries in shaping aggregate labour productivity performance captures only part of the picture. It does not reveal the extent to which productivity changes are driven by underlying efficiency gains, e.g. through technological progress, better management, or gains in workers’ skills and experience. The following section will address this gap by examining developments in multifactor productivity at the level of the total economy and across industries.
MFP is widely regarded as a key driver of economic growth and has received considerable attention in the growth literature. It is measured using the growth accounting framework developed by Solow (1957[28]). It allows for the decomposition of changes in economic output into contributions from changes in labour and capital inputs, and a residual known as MFP. As such, MFP captures changes in economic performance that cannot be explained by changes in the use of these two inputs. Increases in MFP imply that more output can be produced with a given set of inputs, and it is therefore often interpreted as an indicator of technological progress and efficiency gains.
The same framework can be used to decompose labour productivity growth into contributions from capital deepening and MFP growth, thereby distinguishing productivity gains associated with higher capital intensity from those reflecting efficiency improvements.
Figure 4.11 illustrates this decomposition of labour productivity growth into MFP and capital deepening, which can be implemented at either the industry level or for the economy as a whole. MFP growth may capture a range of factors reflecting disembodied technological change, including improved management practices and learning-by-doing effects. In addition, since the labour input measure is not adjusted for labour quality, improvements in workers’ skills and experience may also be partly reflected in MFP growth. By contrast, increases in capital deepening reflect embodied technological change, through improvements in physical capital, ICT investment, and digital infrastructure.
Note: For further discussion of the components captured by MFP, see OECD (OECD, 2001[10]).
Despite its widespread use, measuring MFP remains challenging, with important implications for cross-country comparability (Box 4.2).
MFP is a central concept for analysing the sources of economic growth, as it captures how efficiently an economy combines labour and capital to produce output. Unlike labour productivity, which relates output to a single input, MFP compares output to a bundle of inputs and is typically derived within a growth accounting framework. In this framework, output growth is decomposed into contributions from labour, capital and a residual component, MFP, which captures the part of output growth not explained by measured factor inputs. As such, MFP encompasses technological progress, organisational improvements, economies of scale and other forms of innovation that enable greater output to be generated from a given set of inputs.
New estimates of MFP have been available in the OECD Productivity Database since 2025, incorporating several key improvements over previous editions. Data coverage has been expanded, with MFP now estimated for 21 industry groups. In addition, the revised approach no longer requires extending gross fixed capital formation series or relying on strong assumptions to estimate initial capital stocks. Further details on the methodology are provided in OECD (OECD, 2025[29]).
Despite these advances, MFP remains challenging to measure and interpret. As a residual, it is sensitive to measurement errors and omitted variables arising from inaccuracies in both output and input data. Measuring output is particularly difficult in areas such as household production, non-market activities, and the informal economy, where values are not directly observed (OECD/APO, 2022[30]). On the input side, estimates may be affected by insufficient adjustment for differences in labour quality (Ward and Zinni, 2024[31]). Similarly, incomplete or imprecise measurement of capital inputs – particularly regarding asset coverage, depreciation rates and capital stocks – can bias MFP estimates (Pionnier, Zinni and Baret, 2023[32]).
More broadly, the standard productivity framework provides an incomplete picture, as it does not fully account for factors such as environmental inputs and sustainability considerations (see Chapter 6).
Cyclical factors, such as variations in capacity utilisation, labour hoarding, or adjustment costs during economic upturns and downturns also affect estimates of MFP as well as labour productivity. For example, during periods of weak demand, firms may retain workers or underutilise capital, leading to a temporary decline in measured MFP. Conversely, in recoveries, a more intensive use of existing inputs can mechanically boost MFP (Martin and Jones, 2023[33]; Comin et al., 2025[34]). These cyclical effects can obscure structural trends and make short-term movements in MFP more difficult to interpret.
Heterogeneous multifactor productivity developments across countries, with services leading overall growth
MFP growth for the total economy over the 2023-24 period was mixed, with roughly half of countries recording positive growth and the remainder negative growth (Figure 4.12). The largest declines were observed in Estonia and Lithuania, where MFP fell by around 6.4% and 2.5%, respectively. In contrast, the Slovak Republic continued to record the strongest growth, at around 4% over the same time. In the United States, MFP growth reached 1.4% over 2023-24, ranking second out of 21 countries with available data. Positive MFP growth suggests improvements in efficiency, allowing more output to be generated from a given combination of labour and capital inputs.
Per cent
Note: Data for most countries refer to the unweighted average of year-on-year MFP growth over the years 2023 and 2024. When 2024 data is missing, it refers to 2023 for Croatia, Estonia, Luxembourg, Lithuania, and Spain and to 2022 for New Zealand and Norway.
Source: OECD Productivity Database.
However, the economy-wide average masks significant variation across industries (Figure 4.13). Comparing MFP developments across selected industries over 2023-24 shows that mining and quarrying, manufacturing, and electricity and water management industries generally acted as a drag on overall MFP growth in most countries. In these industries, MFP growth typically lagged behind the aggregate.
Declines in MFP in these industries appear to reflect a combination of declining gross value added (GVA) and relatively limited adjustments in inputs. For example, in Germany, GVA in manufacturing over 2023-2024 declined by approximately 3% while hours worked in the same industry only declined by 1%. This pattern is consistent with the well-documented procyclical nature of MFP: during downturns, firms often retain labour and maintain capital capacity due to adjustment costs and expectations of future demand, leading to lower measured efficiency as output contracts more rapidly than inputs (OECD, 2025[8]; Comin et al., 2025[34]).
Percentage points
Note: The category “Finance” encompasses finance and insurance services, corresponding to section K of ISIC Rev 4. “Trade, transport, accommodation” covers wholesale and retail trade (Section G), transportation (Section H), as well as accommodation and food services (Section I). Manufacturing corresponds to section C, while information and communication correspond to section J of ISIC Rev 4.
Source: Authors' calculations based on OECD Productivity Database.
By contrast, information and communication services were the main contributors to overall MFP growth in many countries in the same period. Similar, though generally more moderate, contributions were observed in professional and administrative services across a number of countries. In trade and other market services, which are typically characterised by relatively low productivity levels, MFP growth nevertheless outperformed the total economy in more than half of the countries with available data.
The prominence of MFP growth in information and communication services may partly stem from productivity gains related to investments in ICT and other intangible assets, whose growth has outpaced other forms of investment since the mid-2010s (Gal et al., 2026[35]). The accumulation of these assets fosters the diffusion of frontier technologies, such as AI, enhancing operational efficiency and production processes, and gradually contributing to productivity gains.
Multifactor productivity explained the overwhelming part of labour productivity growth
Labour productivity growth can be split into two main sources: capital deepening and MFP growth (Figure 4.11). This decomposition shows whether labour productivity improvements arise from greater capital intensity or from more efficient use of inputs.
Over the 2023-24 period, labour productivity growth in most industries was largely driven by MFP growth (Figure 4.14). As a result, differences in industry-level labour productivity growth across countries largely reflected variation in MFP performance across industries in these countries. The greater contribution of MFP growth relative to capital deepening has been a persistent feature of productivity growth over time (OECD/APO, 2022[30]). This pattern is consistent with subdued investment dynamics observed in recent years, partly linked to heightened economic uncertainty (Dlugosch et al., 2025[36]).
The dominant role of MFP growth in explaining labour productivity growth is particularly pronounced in manufacturing and trade, transport and accommodation services. By contrast, in information and communication services, capital deepening makes a relatively larger contribution to productivity growth. This increase in capital deepening in this industry is consistent with investment in ICT and intangible assets (see Chapter 3).
Percentage points
Note: The category “Finance” encompasses finance and insurance services, corresponding to section K of ISIC Rev 4. “Trade, transport, accommodation” covers wholesale and retail trade (Section G), transportation (Section H), as well as accommodation and food services (Section I). Manufacturing corresponds to section C, while information and communication correspond to section J. Professional services encompass professional (Section M) and administrative service activities (Section N).
Source: OECD Productivity database.
Multifactor productivity growth continued its downward trend over time, with increasing divergence across countries
MFP growth has shown a downward trend since the early 2000s (Figure 4.15). This slowdown is visible not only at the level of the total economy but also across several major industries in the business sector (Nikolov et al., 2026[37]; Acemoglu, Autor and Patterson, 2023[38]; Fernald, Inklaar and Ruzic, 2023[39]). Professional and administrative services represent a notable exception, as MFP growth in this industry has, on average, strengthened over the same period.
Per cent
Note: The category “Finance” encompasses finance and insurance services, corresponding to section K of ISIC Rev 4. “Trade, transport, accommodation” covers wholesale and retail trade (Section G), transportation (Section H), as well as accommodation and food services (Section I). Manufacturing corresponds to section C, while information and communication correspond to section J of ISIC Rev. 4. Professional services include professional, scientific, technical (Section M), and administrative and support activities (Section N). The horizontal line in the boxes represents the median, the upper and lower edges of each boxes reflect the 25th and 75th percentiles of the distribution of MFP growth for a given industry and period.
Source: OECD Productivity Database.
Although MFP growth has generally slowed at the total economy level, the extent and nature of this deceleration vary markedly across industries. At the aggregate level, the shift has been relatively moderate: unweighted average MFP growth across OECD countries in the period of 2023-24 is estimated at around -0.3%, compared with 0.8% in the pre-COVID period (2010-19) and about 1.5% during the period of 2000-07. Much sharper declines are observed in manufacturing and financial activities, where MFP growth in 2023-24 stands approximately 4.5 percentage points below the levels recorded during the 2000-07 period.
While cross-country dispersion in MFP growth is relatively modest at the aggregate level, it is much more pronounced at the industry level. This reflects the nature of MFP itself, which captures industry-specific differences in technology adoption, management practices, innovation, and organisational capabilities.
Differences in MFP growth have become increasingly visible in digital-intensive industries, including finance and insurance, and information and communication services. Compared with 2010-19, these industries displayed greater dispersion in 2023-24 (Figure 4.15).
Cross-country MFP trajectories since the 2000s point to increasing divergence across countries (Figure 4.16). In the information and communication services, this pattern has been driven by a small group of countries – notably the United Kingdom and Australia – that have experienced comparatively strong gains in MFP performance. By 2023-24, their MFP growth rates were around twice those observed in the benchmark period (2000-07). In contrast, many other economies have shown a flat MFP path, suggesting little MFP growth since the early 2000s.
Conversely, in the finance and insurance services, the widening dispersion in MFP performance reflects a group of economies, including Estonia and Luxembourg, that have more frequently recorded negative growth, while others have seen modest gains. In these economies, MFP in 2023-24 remained below 2000-07 levels.
The observed dispersion in MFP growth across countries likely reflects underlying structural differences, including country’s resilience to successive economic shocks and differences in the pace of technology adoption (see Chapter 3), which influence countries’ capacity to sustain growth and catch up (Andrews, Criscuolo and Gal, 2016[40]; Tsvetkova, 2024[41]).
Looking ahead, the widening dispersion in MFP performance across countries could potentially translate into more persistent differences in economic performance across economies if sustained over time. The uneven adoption and diffusion of AI and other advanced digital technologies are likely to exacerbate this pattern. These technologies require significant investments – such as data infrastructure, digital skills, organisational change and other intangible assets – which differ widely across countries due to differences in digital readiness, innovation capacity, regulatory frameworks and policy support for technology adoption (Brynjolfsson, Rock and Syverson, 2021[27]; Corrado et al., 2021[42]; Denmou and Franco, 2021[43]).
Index, 2000-07=100
Note: The category “Finance” encompasses finance and insurance services, corresponding to section K of ISIC Rev 4. “Trade, transport, accommodation” covers wholesale and retail trade (Section G), transportation (Section H), as well as accommodation and food services (Section I). Manufacturing corresponds to section C, while information and communication correspond to section J of ISIC Rev. 4. Professional services include professional, scientific, technical (Section M), and administrative and support activities (Section N). MFP is expressed as an index relative to each country’s 2000–07 average (2000-07=100). A value of 100 indicates that the country has the same MFP level as it did in 2000-07. The country shown in dark blue recorded the highest MFP growth among the 21 countries with available data in 2024 or the latest available year. The country shown in light blue recorded the lowest MFP growth over the same comparison.
Source: OECD Productivity database.
References
[15] Abdirahman, M. et al. (2022), “Telecoms Deflators: A Story of Volume and Revenue Weights”, Economie et Statistique / Economics and Statistics 530-31, pp. 43-59, https://doi.org/10.24187/ecostat.2022.530.2063.
[38] Acemoglu, D., D. Autor and C. Patterson (2023), Bottlenecks: Sectoral Imbalances and the US Productivity Slowdown, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w31427.
[40] Andrews, D., C. Criscuolo and P. Gal (2016), “The Best versus the Rest: The Global Productivity Slowdown, Divergence across Firms and the Role of Public Policy”, OECD Productivity Working Papers, No. 5, OECD Publishing, Paris, https://doi.org/10.1787/63629cc9-en.
[1] Ariu, A. (2016), “Crisis-proof services: Why trade in services did not suffer during the 2008–2009 collapse”, Journal of International Economics, Vol. 98, pp. 138-149, https://doi.org/10.1016/j.jinteco.2015.09.002.
[24] Benz, S., A. Jaax and E. van Lieshout (2024), “Digital trade and labour markets in the United Kingdom”, OECD Trade Policy Papers, No. 284, OECD Publishing, Paris, https://doi.org/10.1787/f3de4400-en.
[2] Bourlès, R. et al. (2013), “Do Product Market Regulations In Upstream Sectors Curb Productivity Growth? Panel Data Evidence For OECD Countries”, The Review of Economics and Statistics, Vol. 95/5.
[27] Brynjolfsson, E., D. Rock and C. Syverson (2021), “The Productivity J-Curve: How Intangibles Complement General Purpose Technologies”, American Economic Journal: Macroeconomics 13, https://doi.org/10.1257/mac.20180386.
[11] Byrne, D. (2022), “The Digital Economy and Productivity”, Finance and Economics Discussion Series 2022-038, Board of Governors of the Federal Reserve System, Washington, DC, https://doi.org/10.17016/FEDS.2022.038.
[34] Comin, D. et al. (2025), “Revisiting Productivity Dynamics in Europe: A New Measure of Utilization-Adjusted TFP Growth”, Journal of the European Economic Association, Vol. 23/4, pp. 1598-1633, https://doi.org/10.1093/jeea/jvaf003.
[42] Corrado, C. et al. (2021), “New evidence on intangibles, diffusion and productivity”, OECD Science, Technology and Industry Working Papers, No. 2021/10, OECD Publishing, Paris, https://doi.org/10.1787/de0378f3-en.
[12] Coyle, D. and D. Nguyen (2023), “Free Digital Products and Aggregate Economic Measurement”, INSEE Economie et Statistique, Vol. 539, https://doi.org/10.24187/ecostat.2023.539.2096.
[43] Denmou, L. and G. Franco (2021), “Mind the financing gap: Enhancing the contribution of intangible assets to productivity”, OECD Economics Department Working Papers, No. 1681, OECD Publishing, Paris, https://www.oecd.org/content/dam/oecd/en/publications/reports/2021/09/mind-the-financing-gap-enhancing-the-contribution-of-intangible-assets-to-productivity_d2e72eb1/7aefd0d9-en.pdf.
[36] Dlugosch, D. et al. (2025), “Understanding the weakness in business investment: A cross-country analysis”, OECD Economics Department Working Papers, No. 1836, OECD Publishing, Paris, https://doi.org/10.1787/89bd437d-en.
[23] El-Rayyes, T. et al. (2024), Defining and Measuring the UK Digital Economy, Department for Science, Innovation and Technology, https://assets.publishing.service.gov.uk/media/66f50b2f30536cb92748274b/defining_and_measuring_the_uk_digital_economy.pdf.
[39] Fernald, J., R. Inklaar and D. Ruzic (2023), “The Productivity Slowdown in Advanced Economies: Common Shocks or Common Trends?”, Federal Reserve Bank of San Francisco, Working Paper Series, Vol. 2023/07, pp. 01-47, https://doi.org/10.24148/wp2023-07.
[35] Gal, P. et al. (2026), “Business investment in the face of the digital transformation: Initial evidence”, OECD Economics Department Working Papers, No. 1859, OECD Publishing, Paris, https://doi.org/10.1787/4f89aa3e-en.
[19] Heussaff, C. (2024), “Decarbonising for competitiveness: Four ways to reduce European energy prices”, Bruegel Policy Brief, Vol. 32, https://hdl.handle.net/10419/322557.
[4] INSEE (2025), “Le compte provisoire de l’agriculture en 2024”, Insee Première, No. 2062, https://www.insee.fr/fr/statistiques/8600618.
[16] Ishi, K. (2024), “Productivity Shocks to the Pharmaceutical Sector and the Danish Economy: Denmark”, Selected Issues Papers, Vol. 2024/043, https://doi.org/10.5089/9798400296383.018.
[5] Kuchler, A. et al. (2025), “Danish productivity and competitiveness in a globalised world”, Analysis, No. 6, https://www.nationalbanken.dk/en/news-and-knowledge/publications-and-speeches/analysis/2025/danish-productivity-and-competitiveness-in-a-globalised-world.
[33] Martin, J. and K. Jones (2023), “An Occupation and Asset-Driven Approach to Capital Utilization Adjustment in Productivity Statistics”, Technology, Productivity, and Economic Growth (University of Chicago Press, 2023), https://www.nber.org/books-and-chapters/technology-productivity-and-economic-growth/occupation-and-asset-driven-approach-capital-utilization-adjustment-productivity-statistics.
[22] MPF (2024), Re-thinking approaches to labour migration? Potentials and gaps in EU member states’ migration, https://www.icmpd.org/file/download/63761/file/20241115_Overall_Report.pdf.
[37] Nikolov, L. et al. (2026), “Mid-Tech Europe? Sectoral Patterns of Total Factor Productivity Growth in the EU vs. the US”, Review of European Economic Policy, Vol. 1, https://doi.org/10.2478/ie-2026-0009.
[25] OECD (2026), OECD Productivity Dashboard, http://www.oecd.org/en/data/dashboards/oecd-dashboard-of-productivityindicators.html.
[8] OECD (2025), OECD Compendium of Productivity Indicators 2025, OECD Publishing.
[20] OECD (2025), OECD Economic Surveys: European Union and Euro Area 2025, OECD Publishing, Paris, https://doi.org/10.1787/5ec8dcc2-en.
[21] OECD (2025), OECD Reviews of Labour Market and Social Policies: Croatia 2025, OECD Reviews of Labour Market and Social Policies, OECD Publishing, Paris, https://doi.org/10.1787/90b78cc3-en.
[29] OECD (2025), The revamp of the OECD Productivity Database, OECD Publishing, Paris, https://doi.org/10.1787/f07c55d4-en.
[7] OECD (2024), OECD Compendium of Productivity Indicators 2024, OECD Publishing, Paris, https://doi.org/10.1787/b96cd88a-en.
[9] OECD (2024), OECD Employment Outlook 2024: The Net-Zero Transition and the Labour Market, OECD Publishing, Paris, https://doi.org/10.1787/ac8b3538-en.
[26] OECD (2023), Advancing Digital Maturity in Croatia’s Higher Education System, Higher Education, OECD Publishing, Paris, https://doi.org/10.1787/c3c8d452-en.
[13] OECD (2019), Measuring the Digital Transformation: A Roadmap for the Future, OECD Publishing, Paris, https://doi.org/10.1787/9789264311992-en.
[10] OECD (2001), Measuring Productivity - OECD Manual: Measurement of Aggregate and Industry-level Productivity Growth, OECD Publishing, Paris, https://doi.org/10.1787/9789264194519-en.
[30] OECD/APO (2022), Identifying the Main Drivers of Productivity Growth: A Literature Review, OECD Publishing, Paris, https://doi.org/10.1787/00435b80-en.
[14] OECD/ILO/European Union (2023), Handbook on Measuring Digital Platform Employment and Work, OECD Publishing, Paris, https://doi.org/10.1787/0ddcac3b-en.
[6] Office for National Statistics (ONS) (2025), Blue Book 2025: Globalisation., https://www.ons.gov.uk/economy/grossdomesticproductgdp/articles/bluebook2025globalisation/2025-08-19.
[32] 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, Standards, Centre for the Study of Living, Vol. 45, pp. 155-177, https://ideas.repec.org/a/sls/ipmsls/v45y20236.html.
[17] Reinstaller, A. and R. Sellner (2024), Industrial energy prices and the competitiveness of the Austrian business sector, Büro des Produktivitätsrates, https://www.produktivitaetsrat.at/publikationen/reports/2024003.html.
[28] 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.
[3] STATEC (2025), La baisse de la productivité du travail au Luxembourg influencée par la rétention de main-d’œuvre ?, STATEC, https://statistiques.public.lu/dam-assets/catalogue-publications/note-conjoncture/2025/ndc-1-25/ndc-1-25-etude-7-2.pdf.
[41] Tsvetkova, A. (2024), “Forty years of productivity and labour market resilience in European regions”, OECD Local Economic and Employment Development (LEED) Papers, No. 2024/08, OECD Publishing, Paris, https://doi.org/10.1787/3ba1e1e6-en.
[31] Ward, A. and B. Zinni (2024), “The composition of labour input: Sensitivity testing and results for productivity analysis”, OECD Statistics Working Papers, No. 2024/06, OECD Publishing, Paris, https://doi.org/10.1787/5d9b866a-en.
[18] Weyerstraß, K. et al. (2025), “Energy prices, Competitiveness, and Austria’s exports”, FIW-Policy Brief, No. 70, https://www.econstor.eu/bitstream/10419/334018/1/1941756212.pdf.
← 1. The methodology document can be accessed through the following link: www.oecd.org/content/dam/oecd/en/publications/support-materials/2023/02/oecd-compendium-of-productivity-indicators-2023_bdbeba7d/oecd-compendium-of-productivity-indicators-2023-methodology.pdf.