Abstract
The recent shift in policy focus in the EU to competitiveness and productivity growth is likely to lead to regulatory simplification, the removal of remaining barriers in the Single Market, and a greater emphasis on innovation. This contribution explores the three key ingredients missing from the policy mix that are needed to boost labour productivity growth: improved labour markets; greater fiscal elasticity throughout the business cycle in relation to the labour productivity in the public sector; and the structural accumulation of fiscal space needed to reduce tax burdens.
Introduction
The European Union’s cumulative labour productivity lag behind the US was 13.7% as of 2023, using 1998 as the base year. The annualised labour productivity drag rate was 0.5% over a quarter-century (see Figure 1 and Table 1). This finding, or, more precisely, the frustration it motivated, led to the analysis and publication of policy documents including the Draghi Report on EU Competitiveness (Draghi, 2024) and the Letta Report on the Single Market (Letta, 2024). These two documents provided the background for the new European Commission’s (2024–2029) policy strategy under the umbrella of the EU Competitiveness Compass (European Commission, 2025), aimed at restoring innovation capacity, productivity, and competitiveness in the EU.
The quality of policies depends, among other things, on the accuracy of diagnosis. However, diagnosing productivity gaps is notoriously tricky. The absence of reliable labour productivity statistics in the services sector, especially in the public sector, requires a cautious approach to productivity measurement and policy design. Cross-sector externalities further complicate the assessment process. Caution is also necessary because the EU’s productivity gap compared with the US has widened mainly during recessions. Therefore, there is a risk of misdiagnosis and implementation of policies that lack focus and effectiveness.
A puzzling feature of the European labour productivity gap, besides its cyclical nature, is that it cannot be explained by the difference in labour productivity growth in manufacturing and the information and communications technology (ICT) industries. Labour productivity growth in US manufacturing has been lagging behind that of Europe since 2010. This article shows that the productivity growth differential in ICT is positive for the US, but it explains only a fraction of the total productivity gap since the late 1990s. The sluggish labour productivity growth in other service industries explains the bulk of the economy-wide productivity gap.
Policy discussions have so far focused on two crucial aspects of future productivity growth: (i) the potential for increased productivity related to the completion of the EU Single Market, including simplification of regulation, and (ii) a considerable potential for fostering competition and innovation by promoting cross-industry spillovers related to the innovation and implementation of new technologies. However, there are policy areas that deserve more attention: (i) improvements in counter-cyclical macroeconomic policies in Europe, (ii) measurement of the value of public goods and improvements in the efficiency of their provision, and (iii) dealing with EU fiscal malaise that inhibits the moderation of tax burdens, which can contribute to investment and productivity growth.
The first section presents an analysis of productivity developments. The second section discusses policies.
Productivity growth and the productivity gap across the business cycle
According to OECD data on real GDP per hour worked, cumulative labour productivity growth in the US over the 25 years from 1999 to 2023 was 50.3% (1.64% annualised). In the euro area (EA), it was 23.5% (0.85% annualised), and in the EU, it was 32.2% (1.12% annualised, see Table 1). The difference between the EA and the EU-27 was due to faster productivity growth in Central, Eastern, and Southeastern Europe (CESEE). The average annual rate of increase in the productivity gap between the EA, which was not influenced by the CESEE convergence, and the US was 0.79% over a quarter-century. With the CESEE convergence effect included in the EU-27 data, the productivity gap relative to the US, as mentioned, accumulated at a lower annualised rate of 0.52%.[1]
While the title of this article suggests it is about the EU, the remainder of this section will focus on the EA. Besides the fact that the EA’s 2024 GDP was 85% of the EU’s GDP, the measurement focuses on the EA because productivity comparisons between it and the US are more meaningful, as they compare more developed parts of the EU with even more developed US. Therefore, both the indicators and the conclusions presented below also hold at the EU level.
The labour productivity gap index (US labour productivity divided by EA labour productivity – see Figure 1) shows that a significant portion of the cumulative labour productivity gap of 21.7% accumulated over three relatively short periods totalling six years: (i) 4.4% gap (20% of the total) emerged in 2002–2003, (ii) 5.9% gap (27% of the total) emerged in 2008–2009, and (iii) 3.7% gap (17% of the total) emerged in 2020–2021. Therefore, close to two-thirds (64%) of the 25-year labour productivity gap was accumulated during 24% of the time marked by recessions. Equally important is the fact that 75% of the gap was accumulated before 2010. After a pause from 2011 to 2017, the productivity gap has started to widen again.
The labour productivity gap has also been widening during periods of business expansion, but to a much smaller extent. There is a concentration during short periods when GDP growth was slow: in 2004–2005 (a 2.5% gap or 11% of the total) and in 2023 (a 2.5% gap or 12% of the total). When these three years are combined with the years of recession mentioned earlier, 87% of the labour productivity gap has accumulated in just 36% of the time, specifically in nine critical years out of 25. The productivity gap has a strong cyclical component.
Explanations of productivity differentials usually emphasise fundamentals such as the quality of institutions, knowledge accumulation, innovation, and technical progress; however, these differences tend to accumulate over the long term. The analysis presented here suggests that potential explanations should be extended to include inertia (hysteresis effects) during and after recessions. Flaws in counter-cyclical macroeconomic policies during times of crisis may lead to prolonged adverse effects on productivity. We will revisit this topic in the policy discussion.
It is also essential to look beyond the aggregate figures. The economy-wide productivity gap arises from industry-level developments that need not be correlated: (i) long-term industry productivity gaps over 25 years tend to vary significantly across industries, (ii) the volatility of industry-level productivity gaps throughout the business cycle differs, and (iii) there are externalities, that is, spillovers among industry-level productivity gaps. For example, a negative productivity shock in energy generation harms productivity growth in energy-intensive industries. At the same time, a positive productivity shock in ICT or infrastructure may affect productivity growth in other sectors.
It is difficult to measure all three aspects at the industry level because there are no comparable OECD statistics on hourly labour productivity at the industry level for the US and the EA. Still, an educated guess is possible based on data from the US Bureau of Labor Statistics and Eurostat for selected industries.
Labour productivity in the EA manufacturing sector outperformed that of the US manufacturing sector from 1999 to 2023. Annualised growth rates were 2.21% in the EA and 1.72% in the US (see Table 1 and Figure 2). The productivity gap with the US in the numerator (see Figure 2) accumulated at an annualised rate of −0.48% (EA productivity growing faster), indicating a structural break: US manufacturing accumulated a productivity advantage up to 2009. The manufacturing productivity differential in this period recorded two short-term increases in favour of the US during the recessions of 2002–2003 and 2008–2009. Therefore, cyclical volatility in the first decade of the century in manufacturing partially explains the dynamics of the aggregate productivity gap (see Figure 1). However, after 2010, a long-term decline in US manufacturing productivity began: labour productivity in manufacturing in the US compared with labour productivity in EA manufacturing was 11.3% lower in 2023 than in 1998.
This is puzzling because the growth of US Big Tech and the vibrant start-up scene centred around Silicon Valley were expected to create a constant flow of positive spillovers in US manufacturing through the adoption of new digital technologies and the fostering of research and development (R&D) and industrial innovation. However, no such positive effect is evident in the US manufacturing productivity data over the last 15 years, as labour productivity in manufacturing and ICT in the US have exhibited contrasting trends during this period (see Figure 2). This serves as a sobering warning: dynamism in the ICT industries does not necessarily translate into increased manufacturing productivity, as other factors affecting productivity in manufacturing (e.g., global competition in the case of the US) may prevail.[2]
Figure 2. Index of relative labour productivity in selected industries in US vs euro area (1998 = 100)
Note: PST – professional, scientific, and technical services.
Sources: Eurostat, US Bureau of Labor Statistics, author’s calculations.
The annualised growth rate of labour productivity in the ICT sector from 1999 to 2023 was 2.93% in the US and 2.55% in the EA. The ICT sector exhibits an annualised rate of productivity gap accumulation of 0.36%, resulting in a cumulative differential of 9.48% (2023/1998) in favour of the US. However, ICT’s share in GDP has been single-digit on average during this period. Thus, its direct influence on the accumulation of the total productivity gap was of minor importance.[3] This contradicts assumptions regarding the role of ICT in the accumulation of the productivity growth differential, at least when its direct effects are taken into account. The indirect effects of ICT’s impact on productivity related to ICT applications in other industries may be much more substantial.
We will focus on other sectors soon, but first, notice that the productivity gap in ICT shows no clear cyclical pattern (see Figure 2). Still, the pace of accumulation was very different in 1999–2013 compared with 2014–2023: in the first period, the EA accumulated a positive productivity differential, which was reversed after a structural break in 2013 when the US took the lead.[4] As pointed out in the preface of the Draghi Report, the latter trajectory may be explained by global economies of scale achieved by the US Big Techs after the Great Recession.
In the energy sector and in the professional, scientific, and technical (PST) services, the annual rates of productivity gap accumulation in favour of the US were substantial: 1.49% and 1.78%, respectively. However, the productivity gap in the energy sector does not display a cyclical pattern. The gap arose due to a negative structural break in 2021–2022 related to a sudden and extreme contraction in energy production in the EA. This reflects the broader impact of the energy crisis in Europe, as the EA recorded faster productivity growth in the energy sector before the recent shock.
The productivity gap in the PST services exhibits a more regular cyclical pattern, with 71% of the cumulative productivity differential until 2023 emerging between 2002 and 2005 and again between 2008 and 2009.[5] The vibrant US ICT sector had a positive impact on the US services industries earlier in this century. The rate of adoption of new productivity-enhancing technologies in the PST services in the US was particularly high up to 2012, while EA productivity in the PST services was on a long-term decline: annualised rates of labour productivity growth in 1999–2012 in the PST services were 2.53% and −1.10% in the US and the EA respectively, as European PSTs were late adopters of technological change. Then, trends reversed from 2013, and labour productivity started to grow faster in the EU PSTs (1.14% vs 0.55% in the US). Productivity growth was generally slower after the Great Recession. However, as far as the PST services are concerned, the US still enjoyed a significant accumulated productivity advantage of 55.3% in 2023 compared with 1998, because the annualised rates of labour productivity growth in PST in the 2013–2023 period, when the EA took the lead, were lower than they had been previously (see Table 1).
Table 1. Labour productivity per hour and productivity differential, US vs EA – compounded annual growth rates (CAGR) in respective periods in %
Industry | Description | 2023 vs 1998 | Selected Periods* | |
Calculation of productivity differential growth for selected industries | ||||
Total economy | 2010 vs 1998 | 2023 vs 2010 | ||
EA | 0.85 | 1.03 | 0.73 | |
US | 1.64 | 2.32 | 1.10 | |
Diff. US vs EA | 0.79 | 1.28 | 0.37 | |
Manufacturing | 2009 vs 1998 | 2023 vs 2009 | ||
EA | 2.21 | 1.95 | 2.42 | |
US | 1.72 | 3.73 | 0.17 | |
Diff. US vs EA | −0.48 | 1.74 | −2.19 | |
ICT | 2013 vs 1998 | 2023 vs 2013 | ||
EA | 2.55 | 2.97 | 1.94 | |
US | 2.93 | 2.23 | 3.99 | |
Diff. US vs EA | 0.36 | −0.72 | 2.01 | |
Professional, scientific, and technical services (PST) | 2012 vs 1998 | 2023 vs 2012 | ||
EA | −0.12 | −1.10 | 1.14 | |
US | 1.65 | 2.53 | 0.55 | |
Diff. US vs EA | 1.78 | 3.67 | −0.58 | |
Energy (electricity, gas, steam, air cond.) | 2020 vs 1998 | 2023 vs 2020 | ||
EA | −1.10 | 1.11 | −15.93 | |
US | 0.36 | 0.45 | −0.26 | |
Diff. US vs EA | 1.49 | −0.65 | 18.64 | |
Labour productivity growth in other selected industries in the EA | ||||
Construction | EA | −0.67 | No structural break | |
Finance and insurance | EA | 1.55 | No structural break | |
Trade | EA | 1.95 | No structural break | |
Transportation and storage | EA | 2013 vs 1998 | 2023 vs 2013 | |
0.82 | 1.46 | −0.53 | ||
*Selection of periods is based on visual inspection.
Sources: Eurostat, US Bureau of Labor Statistics, author’s calculations.
Development in the PST sector is indicative of a broader productivity gap accumulated in the services industries in Europe. The key question is: what has happened to the services industries in Europe? If the manufacturing–ICT nexus, which comprises around one-fifth of GDP, does not explain a significant fraction of the economy-wide labour productivity gap, then the gap related to the remaining four-fifths of GDP, consisting of non-ICT services, must be even larger than the economy-wide productivity gap.
The calculation of the productivity gap in other industries is beyond the scope of this analysis; however, the construction industry warrants special attention, as real hourly labour productivity in the EU construction sector declined by 14.9% from 1998 to 2024 (CAGR is −0.67% up to 2023, without a visible structural break – see Table 1). This decline may be related to the slow adoption of new technologies, the prolonged effects of cyclical downturns, insufficient economies of scale and scope linked to costly regulation, supply constraints (e.g., restrictive zoning regulations), demand constraints (limited public infrastructure investment that could support the growth of large construction companies), and regulatory barriers within the imperfect EU Single Market.
Three other service industries in the EA reported modest labour productivity growth over a quarter-century: finance and insurance (annualised rate of labour productivity growth of 1.55%), trade (1.95%), and transportation and storage (0.82%), with the latter exhibiting a structural break in 2013 (productivity began a long-term decline after 2013 – see Table 1). The slow adoption of new technologies and a lack of market dynamism are widely accepted explanations for the relatively slow productivity growth in Europe, which also applies to the services sector. Adilbish et al. (2025) pointed out that (i) large leading firms across Europe are lagging in terms of innovation and productivity, (ii) too few young high-growth firms make it to the top, and (iii) many early high-growth firms settle too early and remain mature low-growth firms. In the light of these findings, practical recommendations for boosting productivity growth across Europe involve the broader adoption of digital technologies and general technical progress, along with deregulation and the elimination of remaining barriers in the EU Single Market, as well as increasing the availability of risk-tolerant capital for company growth. All policy documents reiterate these recommendations. But does it cover the whole problem?
The dynamics of labour productivity in the public sector have been overlooked in discussions on productivity and policy design, despite the public sector representing a substantial portion of GDP in the EA/EU. Policy documents acknowledge the importance of externalities generated by the security, education, science, and health sectors; however, they are largely silent on the public sector’s labour productivity level and growth. This is partly due to a lack of data. Neither Eurostat nor the US Bureau of Labour Statistics measures labour productivity for public administration, defence, education, science, human health, and social care activities. Estimating labour productivity in the production of public goods is notoriously tricky. GDP measurement assumes the impossibility of directly measuring the output of the part of the public sector that produces non-market goods.[6] Therefore, Eurostat’s ESA2010 methodology (like the United Nations’ System of National Accounts) calculates the real value of government output by summing the actual costs of its production, which include the cost of labour, the cost of intermediary inputs, and the consumption of fixed capital required to produce public services. Economists sometimes use proxies – for example, productivity in PST services may serve as a proxy for productivity in public administration – but still, wherever a public service in terms of the provision of non-market goods involves a complex interplay of production cost, quantity, and quality alongside externalities, the absence of market data (prices) makes it nearly impossible to calculate the real value produced per worker hour.
The measurement problem is closely related to the management problem: it is difficult to manage what is hard to measure. Hence, the government sector is a productivity black box, although government final consumption accounted for 21.6% of the EU’s and the EA’s GDP in 2024. There may be an internal productivity problem related to inertia and a lack of incentives for labour productivity growth in the public sector, which affects the entire economy due to spillovers. This problem is related to broader themes of public management quality and the EU’s fiscal malaise. In addition to the productivity measurement and management problem, the overall political atmosphere in nearly all Member States is such that it appears almost impossible to discuss significant reductions in tax burdens, despite the EU having the highest taxes on average in the world. Research has shown that high direct taxes, particularly corporate income taxes, are a significant hindrance to investment and productivity growth (e.g., Hanappi, Milot, & Turban, 2023); therefore, a comprehensive discussion about long-term labour productivity growth must also address tax policy.
[1] The EU-27 vs EA productivity gap is a good proxy for the Central and Eastern Europe (CEE) convergence effect because only small economies joined the euro area after its inception (Slovenia 2007, Slovakia 2009, Estonia 2011, Latvia 2014, Lithuania 2015, Croatia 2023), while more populous CEE member states (Poland, Romania, Hungary, and Czechia) decided not to join.
[2] Long-term strengthening of the US dollar after 2008 may be one element of the explanation.
[3] For example, with a GDP share of 10%, the annualised contribution to the accumulation of the productivity gap would be around 0.04%, that is, less than one-tenth of the total.
[4] This finding aligns with those of other studies. Gervais (2025) quotes a McKinsey 2023 study that showed average annual labour productivity growth rates in ICT of 3.5% in the US and 1.7% in Europe. The corresponding growth rates in this study, when taking 2011 as a starting point, are 3.64% in the US and 1.94% in the EA, which is very close. Here, the ICT labour productivity index is constructed with annual employment shares as weights for four sub-sector activities: (i) wired and wireless telecommunication carriers; (ii) software publishers; (iii) cable and other subscription programming; (iv) radio and television broadcasting; and (v) newspaper, periodicals, books, and directory publishers.
[5] 71% applies to both periods together.
[6] Non-market goods are characterised by zero or insignificant prices that are not significant enough to affect supply or demand.
Policy discussion
The Draghi Report on EU competitiveness, published on 9 September 2024, laid out a path for Europe’s new industrial policy (the New Industrial Deal). The report reflects on the productivity gap vis-à-vis the US as an existential threat to the EU. Written in the aftermath of the 2022 energy price shock, the report acknowledges the productivity issues in energy generation, as Europeans pay around three to four times more per unit of energy compared with the US and China. Moreover, the report assumes that the tech sector is the primary factor in the productivity gap vis-à-vis the US. Mario Draghi’s foreword to the report emphasised that only four out of the world’s 50 largest Big Tech companies are based in the EU. Consequently, the report stresses the role of the tech–innovation–manufacturing nexus.[1] Five horizontal policies are also outlined in the Draghi Report: accelerating innovation, closing the skills gap, sustaining investment, revamping competition, and strengthening governance. Several of these policy recommendations have the potential to boost productivity growth in the critical service industries: closing the skills gap will require a revamp of the education sector, and sustained investment, if successful, will enhance productivity in finance and insurance, with positive spillovers to other growing industries.
The Draghi Report’s findings and recommendations shaped the strategy and policies of the new European Commission (EC). The Commission Vice-President is now responsible for regulatory simplification, aiming to reduce the reporting burden by 25% for all companies and 35% for small and medium-sized enterprises (SMEs). The EU Competitiveness Compass was adopted in January 2025 as a political translation of the Draghi Report into an official document that will guide the EC’s work through the mandate, focusing on three pillars of transformation: closing the innovation gap, decarbonisation, and competitiveness, as well as reducing dependencies and increasing security.[2]
Deregulation and the removal of the remaining barriers to the Single Market’s functioning, along with digitalisation and innovation, are additional cornerstones of European policy. It is expected that these policies will have the most potent effects in the service industries, as the Letta Report indicates that the greatest productivity growth potential lies in completing the Single Market in electronic communications, energy, financial markets, and defence.
Therefore, the discussion on the Single Market in the remainder of this section is brief, as it is already well established and firmly centred as a focus of policy. This section is devoted primarily to three policy areas that have been underrepresented in European policy documents and which may be critical for policy focus and effectiveness: (i) potential reform of counter-cyclical macroeconomic management to avoid the protracted accumulation of labour productivity lag during future recessions, (ii) improvements in the measurement and management of labour productivity in the production of public goods, and (iii) resolving the EU’s fiscal malaise by lowering direct tax burdens to promote investment and productivity growth in Europe.
[1] The New Industrial Deal for Europe is based on ten critical sectors: energy is the first, followed by critical raw materials and the IT/AI block of digitalisation and new advanced technologies, which comprise broadband networks, computing and AI, and semiconductors. This is followed by energy-intensive industries (EII, e.g., metals, chemicals, and non-metallic minerals) and clean technologies, whose primary goal is to facilitate EII’s further decarbonisation. Five other critical industries are the automotive industry, pharmaceuticals, defence, space, and transportation, including logistics.
[2] The first significant initiative of the new EC was the Stop-the-Clock Directive, which alleviated the ESG (environmental, social, and governance) reporting burden for smaller enterprises and simplified it for larger ones. Draghi’s call for raising the EU gross investment-to-GDP ratio to 26–27% (2024: 21%), which was last seen about half a century ago, led the EC to adopt the Strategy on a Savings and Investment Union (SIU), aiming to mobilise private savings to finance investment through perfected financial markets.
Counter-cyclical macroeconomic policy and labour productivity
In this century, US recessions have been milder and shorter than those in the EU/EA. Interventions via fiscal and monetary policies in the US have been faster and stronger. In 2008–2010, the general government fiscal balance as a percentage of GDP in the US recorded figures of −6.6%, −13.2%, and −11.0%, while the corresponding figures for the EA were −2.1%, −6.1%, and −6.0%. This difference was repeated during the COVID-19 pandemic and its aftermath, as the average annual fiscal balance-to-GDP ratio for the US (2020–2024) was −8.7%, compared with −4.4% in the EU.
The fiscally fuelled recovery of aggregate demand after a recession can be beneficial for subsequent productivity growth if it leads to a more rapid increase in investment. A faster investment recovery can promote the reallocation of capital and labour, thereby contributing to long-run productivity growth through Schumpeterian creative destruction. However, evidence of enhanced investment responses in the US following recessions is not straightforward. Both the EA and the US experienced low investment periods following recessions, as measured by the investment-to-GDP ratio. Real GDP in the US in the aftermath of recessions grew faster than in the EA on average, so real investment in the US also increased at a faster pace. However, this does not imply a causal relationship between real investment and real GDP.[1]
It is unclear whether more expansionary macroeconomic policies are beneficial when considering their long-term effects on productivity. Thus, the discussion should not focus on replicating the US approach to macroeconomic stabilisation during cyclical downturns; overly strong fiscal expansions can (i) create high indebtedness and loss of fiscal capacity in the long run (the US is facing this problem, as its public debt ratio reached a burdensome 123% of GDP after two US fiscal explosions in this century),[2] (ii) lead to overly lax economic conditions in which low-productivity zombie firms may survive too long, and (iii) share prices of established listed companies may inflate too much, which can slow down the process of reallocation of capital and labour towards less reputable high-productivity firms and industries which are undervalued due to higher risk during early stages of growth. In general, it is hard to distinguish between output and productivity recoveries after recessions that are sustainable in the long run and those that are transitory, driven by unsustainable excess demand.[3]
Therefore, theories that (i) EU fiscal rules were a straitjacket, and that (ii) the euro system’s monetary policy was overly restrictive because its mandate is focused only on low inflation instead of both low inflation and full employment as in the US, may be considered only after other potential explanations have been rejected. We need to examine other aspects of macroeconomic stabilisation mechanisms that may have adversely impacted the allocation of capital and labour, as well as labour productivity, during recessions and their aftermath in this century. The key may be hidden in the labour market.
The coefficient of variation of monthly unemployment rates in the US during this century was almost twice as high as in the EA. Even when peak variations in 2020 are excluded from the calculation, the US coefficient of variation of monthly unemployment rates was 0.32, compared with 0.18 in the EA.[4] Greater short-term variability of unemployment rates in the US may be associated with greater ease of hiring and firing due to less restrictive labour market regulation. Studies have documented faster job-to-job transitions in the US (IMF, 2025). At the same time, there is well-documented labour hoarding by European firms during economic slowdowns (Arce & Sonderman, 2024).
Considerations of labour market dynamics are often presented as a trade-off. Supposedly, a more Darwinian US labour market may be economically efficient, but it leaves too many people behind. Many workers who have no savings and/or no access to family or social safety nets may be forced to accept the first job offered, which, on average, is below their skills and potential. This creates additional downward pressure on wages, resulting in a suboptimal allocation of labour, increased income inequality, and lower labour force participation rates.
However, the difference in labour force participation between the US and the EA is not that significant. Labour force participation rates are very similar in the age bracket of 55–64 (around 69%). A four-percentage-point difference occurs in the prime age bracket of 25–54 in favour of the EU, with many possible explanations: it may be related to a shift in labour supply, where workers enter employment directly from inactivity rather than transitioning from prior unemployment in the US. This may be explained by restrictive or low unemployment benefits and the high share of unregistered immigrant workers, both of which are more prevalent in the US. Therefore, the relatively low variability of unemployment rates in the EA compared with the US in this century, if associated with a slower process of labour reallocation, likely explains the slower growth of labour productivity. Recessions amplify this effect.
Even if some human capital is preserved due to lower variability of unemployment rates, the EU should better balance the pace of labour reallocation with the preservation of human capital. This can be accomplished by adopting labour market organisation models that maximise transitions (i) from inactivity to both employment and unemployment, (ii) from unemployment to employment, and (iii) between jobs. Models of flexicurity, higher retirement ages, and active labour market policies (ALMP) align with well-established best practices that involve both supply-side and demand-side incentives for education, lifelong learning, and promoting entrepreneurship. Best practices for such policies in EU Member States are observed in those that experienced faster than average long-term productivity growth (e.g., Denmark and Sweden).
While ALMP falls under the category of national policies, there is potential for improved EU coordination by adopting best practices and utilising the expertise and resources of other Member States. This can be achieved by activating common financial instruments, such as SURE, but not for funding job retention schemes, as seen during the COVID-19 crisis, when it was anticipated that existing economic activities would resume. Instead, these instruments should support smoother job-to-job and unemployment-to-job transitions during periods of economic strain. By implementing elastic labour market interventions, which are typically amplified during recessions, rigid labour market regulations can be relaxed, thereby preventing labour hoarding by less productive firms.
However, liberalising labour market regulation is politically challenging. It is a policy within the national domain. The strength of labour unions and the tradition of protecting workers, which lie at the core of the European social market model, limit the room for manoeuvre for reform. However, national labour market reforms should become easier as the population ages, as the gap between the number of older workers exiting the workforce to retire and the number of young workers entering the labour force is growing across Europe. The risk of long-term structural unemployment is reduced owing to demographic trends. With this in mind, and through common EU coordination and expert and financial assistance, national policymakers should feel more confident in initiating labour market reforms.[5]
A cyclically elastic labour market policy would accelerate the reallocation of labour towards higher-productivity firms and industries if counter-cyclical public investment policy also becomes more elastic across the business cycle. Stronger upward responses of public investment during recessions, compared with what we have seen in the first two major recessions of this century, would represent a new approach, as the public investment ratio has remained consistently around the average of 3.3% of GDP in this century, with a slight increase to 3.9% and 3.7% in 2009–2010 due to sharp contractions of economic activity, and no significant changes during the pandemic and its aftermath. The rigidity of the public investment ratio stems from the limited fiscal capacities of Member States. Constraints on a more elastic fiscal stabilisation policy are not related to the (in)flexibility of fiscal policy rules.[6] Budgetary cyclical (in)elasticity runs much deeper. The ability to accumulate fiscal space during economic expansions, which is necessary for effective fiscal expansion during recessions, is a structural feature influenced by complex political and economic considerations. We will revisit these topics in the following two sections, as public sector productivity is closely related with accumulation of fiscal space.
[1] In 2002–2003, the average gross fixed capital formation-to-GDP ratio bottomed out at 22% in the US and the EA, followed by a 23.0% peak in 2006 in the US and 24.2% peak in the EA in 2007. During the Great Recession, investment bottomed out at 18.3% in the US in 2010, reached the next maximum of 21.6% in 2020, and remained slightly below that level thereafter. In the EA, the investment ratio bottomed out at 19.7% in 2013, reached the next maximum of 22.5% in 2019, and after a mild decline to 22.1% in 2020, it reached 23.0% and 24.1% in 2021 and 2022, respectively. See IMF (2025: 6) on the faster growth of real investment in the US.
[2] The ratio is around 88% in the EA, and it remains to be seen how this difference between the US and the EA will play out in the coming years in terms of its impact on fiscal policies and labour productivity.
[3] Andersson, Jarvis, and Soudan (2025) focused on the short period between late 2021 and late 2024, attributing weaker tangible investment in the EU to weaker demand and an energy price shock associated with policy uncertainty stemming from Russia’s invasion of Ukraine. There is a larger gap in investment in intellectual property, which is typically explained by weaker Big Tech companies, a lack of digital innovations, and cumbersome regulations. In addition, overly lax fiscal and monetary expansion may obscure the natural rate of unemployment and increase the probability of monetary policy errors reflected in an overly large variability of inflation and interest rates that may create social costs in terms of lost output.
[4] The US standard deviation is higher than the EA’s even before normalisation by the mean.
[5] The EU’s Flexicurity programme (European Commission, 2007) can be refreshed, as most documents relate to 2007, and substantially expanded with a particular focus on the design of counter-cyclical common measures.
[6] The Stability and Growth Pact was reformed in 2024, with additional relaxation in 2025, allowing for 1.5% of GDP in additional defence spending from 2025 to 2028.
Measurement of the value of public goods and improvements in the efficiency of their provision
Greater cyclical flexibility of fiscal policy in countries where general government expenditures converge towards 50% of GDP requires very efficient public sectors operating at high labour productivity standards. However, we do not know the exact labour input (i.e., the number of employed individuals and the hours worked). As of mid-2025, only eight EU Member States irregularly contribute to Eurostat’s comparable statistics on the number of people employed in the general government sector. Measuring real output in the government sector is even more challenging than counting workers and hours of work: there are no market prices or quantities of public services. At the same time, immeasurable externalities, both positive and negative, are omnipresent. For example, how do we measure benefits such as risk reduction (from regulation and public goods such as safety and security), the rule of law, health, or knowledge? There are indirect measures (e.g., healthy life expectancy, indicators of resolved court cases, results in PISA tests, publications in scientific journals, etc.), but with benefits such as security, ownership protection, health, knowledge, and innovation being produced from multiple sources, partly thanks to individual efforts and additional services acquired in the private markets, how can we measure a portion of such social benefits created by the public sector’s entities that are directly involved in the production of services?
Without advancements in measuring the quantity and quality of public services, a significant public sector management issue arises: the quality of public management often suffers when measurement is lacking. Additionally, labour may be hoarded in the government sector, and pressures for technological change and efficiency in the production of public services may remain limited, especially if public sector trade unions are strong. As externalities from the production of public goods are overwhelming, lagging labour productivity growth in the government sector may have a significant negative impact on the entire economy and society.
Improvements in labour productivity in providing public services could have a substantial effect on society’s overall productivity, fiscal capacity, and elasticity. Such improvements would also have positive external effects, such as (i) greater citizen trust in public institutions due to increased democratic control over the administration and its output, (ii) greater transparency, less corruption, and more productive informed policy choices, and (iii) increased and more stable aggregate demand for productivity-enhancing technologies by the government sector, which might facilitate the growth of innovative private companies and industries that can produce such technologies, reaching economies of scale which would not be possible solely on the basis of contracts in the private markets.[1] Procurement by the government sector focused on productivity growth, which acts as an early adopter of new technologies, may have a stronger and longer-lasting impact on the spread of technological innovations than state aid. While fragments of such policies are presented in EU policy documents, such as the adoption of artificial intelligence (AI) in the provision of public services, this welcomed measure is not framed within a prioritised programme for enhancing labour productivity in the public sector. This is a missing link in European policy documents.
There are methods for measuring inputs, outputs, and labour productivity in the provision of public services (Somani, 2021). Additionally, some EU Member States have developed public management techniques that rank among the best in the world. At the EU level, sectoral statistical standards, such as CEPEJ-STAT for judicial systems, serve as a valuable evidence-based tool for public management (CEPEJ, 2024). However, the absence of measurements similar to CEPEJ across public services means that it is not possible to comment in depth on how they are being produced and delivered. Common administrative productivity standards and public management toolkits are lacking. International institutions active in this area, such as the OECD and the World Bank, have developed practical tools and internationally comparable indicators.[2] The EU should do much more and provide these in more detailed and standardised forms shared among Member States to inspire labour productivity growth in the public sector.
[1] Belenzon and Cioaca (2022) explained the importance of the public procurement effect in the development of the IT sector in the US.
[2] See, for example, the World Bank Group Bureaucracy Lab (World Bank Group, n.d.).
Overcoming the EU’s fiscal malaise and tax burden
Greater focus on public sector productivity growth is necessary to achieve increased budgetary elasticity across the business cycle, thereby expanding the fiscal space for counter-cyclical public investment, implementing tax reforms, and addressing the EU regulatory and fiscal malaise. The recent regulatory simplification agenda is welcome, but the EU needs additional public sector management tools to widen the space for tax reforms, as cuts in personal and corporate income tax rates may be essential to accelerate investment and labour productivity growth. Studies have shown a significant impact of corporate income taxes on investment and productivity (e.g., Hanappi, Milot, & Turban, 2023).
This topic has been largely excluded from European policy debates, with the exception of tax credits related to investment in R&D. The likely reason is that it falls under the political jurisdiction of national authorities. Additionally, the anticipated social costs of ageing, which will impact public pension and health systems, limit the ability to reduce direct taxes and social contributions. Awareness of this fundamental problem may hinder more ambitious political communication regarding tax cuts. However, this is a mistake; enhanced efficiency and productivity in the provision of public services can relieve the pressures of ageing on fiscal deficits, paving the way for the substitution of inefficient public expenditures by productive expenditures related to ageing: higher living standards and healthier lives after 60 or 70 years of age may help to sustain aggregate demand and create motivation to work to a more advanced age. The same effect would be produced by the contribution of higher investment and productivity growth due to a more investment-friendly tax environment.
The ratio of direct taxes and social contributions to GDP varies significantly across the EU, with a lower ratio being associated with lower real income per capita (see Figure 3). Deviations from the expected value represented by the regression line are an essential consideration because the distance may be associated with long-term productivity growth. Focusing on the outliers, particularly on the right-hand side of the chart where high-income economies are located, is illustrative. Countries with relatively low direct tax burdens – such as the Netherlands and Sweden – recorded annualised rates of labour productivity growth in 2024 versus 1998 of 0.8% and 1.2%, respectively, while the opposite outliers – Belgium, France, and Italy – recorded annualised growth rates of 0.7%, 0.7%, and 0.2%, respectively.[1]
[1] Malta is not mentioned because it underwent a rapid real convergence period, which distorts its productivity growth upwards (at 2.0% annualised).
This is little more than a statistical anecdote. Isolating the long-term impact of direct taxation on long-term labour productivity growth requires a level of analysis that is beyond the scope of this article. Nevertheless, decades of economic research and the logic of investment convey the same message as does Figure 3: it is the expected net return versus risk that influences investment decisions. Higher direct taxation reduces the net return–risk ratio; hence, it leads to a lower supply of equity and less investment, ceteris paribus, which is critical for labour productivity growth. Thus, reasonable tax policies, particularly concerning the direct taxation of capital and labour, are crucial for boosting investment and productivity growth.
Elimination of the remaining barriers in the EU Single Market
The IMF’s research on the EU’s potential to catch up with the US in terms of labour productivity has prioritised completing the Single Market, as the remaining intra-EU barriers to trade are reportedly much higher than extra-EU barriers, representing tariff equivalents of 44% for goods and 110% for services (Kammer, 2024). Besides removing the usual obstacles through the harmonisation of regulations and the opening of protected sectors, the fundamental measures are significantly more challenging to achieve as they relate to the mobility of labour and capital. European projects in infrastructure and savings and investment union are critical in this regard. Based on the findings concerning industry-level productivity presented in the first section, achieving mobility-related completion of the Single Market and associated public investment may be vital for stimulating labour productivity growth in non-ICT service sectors such as construction, waste management and recycling, transportation and logistics, trade, finance, and public services. These are precisely the sectors where productivity lags behind the US, explaining the bulk of the total economy productivity lag that was accumulated in this century.
Conclusion
The main finding of this article is that the economy-wide labour productivity differential between the US and EU, which has been accumulating since the late twentieth century, has primarily occurred in non-ICT services. This is not to imply that the manufacturing–ICT nexus is unimportant for explaining the productivity gap, as it may have significant indirect effects on the adoption of new technologies and productivity growth in service industries. Therefore, industrial policy informed by the Draghi Report on EU competitiveness, along with deregulation and regulatory harmonisation driven by the Letta Report on the Single Market, may not be sufficient to revive labour productivity growth in the EU. While simplifying regulations, completing the savings and investment union, reforming education, stimulating innovation, and eliminating the skills gap are critical components of the policy mix for revitalising productivity growth, this article identifies three essential elements that are missing from the policy mix.
Firstly, the flexibilisation of labour market regulations and fiscal policy, particularly concerning public investment, may help cushion productivity gaps that emerge during recessions and their aftermath. The labour market and investment are the keys to eliminating the cyclicality of the productivity gap. Secondly, greater fiscal elasticity throughout the business cycle demands significant improvements in public sector productivity, as increased efficiency and effectiveness in the public sector are necessary to expand fiscal space without further raising tax burdens. This is related to the third missing component, which concerns the structural accumulation of fiscal space required for reducing tax burdens. This specifically pertains to direct taxes, as research has indicated that direct taxation impedes investment and labour productivity growth.
Labour market reforms and structural fiscal adjustments based on fundamental improvements in labour productivity in the public sector are politically challenging. That, along with the fact that both policies fall within the national policy domain, explains why the three policy blocks have been underrepresented in European policy debates and documents to date.
However, the reasons for underrepresentation may run deeper. The spirit of new interventionism haunts Europe and the world. State aid, tax incentives, and credit for selected sectors and companies, as well as hidden subsidies, political favouritism towards special sectors and companies, and favourable financial arrangements, which were once considered to be in opposition to free market and competition principles, are now back. It is politically easier to distribute fiscal resources towards industrial priorities than to promote reforms such as labour market liberalisation and enhancing the efficiency of the public sector.
That is a trap. A sobering warning was issued by Brandao-Marques and Toprak (2024) from the IMF about the impact of state aid in the EU: while firms that receive state aid increase revenue and employment, it has no effect on their investment and labour productivity. Moreover, there are adverse spillover effects on competing firms that negate any positive outcomes for recipient firms. The authors leave the door open to the potential efficiency of EU-level state aid over national-level state aid to companies, but this is a normative statement that rests on the belief that competition in the Single Market would eliminate national market distortions created by national-level intervention. However, distortion is distortion, regardless of the level.
There are other effective channels for indirect support of innovative and productive companies. Public procurement related to investment in new technologies in the public sector aims to increase labour productivity, which in turn helps grow innovative and efficient companies able to deliver productivity-enhancing technologies to the public sector.
Therefore, some of the recently planned and implemented EU policies are safe bets, as they address the weak points revealed by productivity diagnostics: an increase in pan-European public investment projects, simplification of regulations, and the elimination of remaining obstacles in the Single Market. However, three missing blocks that are underrepresented in the policy mix remind us that policy success is far from guaranteed. Let us hope that the safe bets prevail and that the missing blocks will soon be included in the policy mix.References
Adilbish, O.-E. et al. (2025). ‘Europe’s Productivity Weakness: Firm-Level Roots and Remedies’, VoxEU Column, 24 February, https://cepr.org/voxeu/columns/europes-productivity-weakness-firm-level-roots-and-remedies.
Andersson, M., Jarvis, V., & Soudan, M. (2025). ‘Business Investment: Why Is the Euro Area Lagging Behind the United States?’. ECB Economic Bulletin, 2, https://www.ecb.europa.eu/press/economic-bulletin/focus/2025/html/ecb.ebbox202502_01~a3130f4f9f.en.html.
Arce, O., & Sonderman, D. (2024). ‘Low for Long? Reasons for the Recent Decline in Productivity’, The ECB Blog, 6 May, https://www.ecb.europa.eu/press/blog/date/2024/html/ecb.blog20240506~f9c0c49ff7.en.html.
Belenzon, S., & Cioaca, L.C. (2022). ‘Government Procurement: A Policy Lever to Revitalize Corporate Scientific Research’. Innovation Frontier Project, 17 February, https://innovationfrontier.org/government-procurement-a-policy-lever-to-revitalize-corporate-scientific-research/.
Brandao-Marques, L., & Toprak, H.H. (2024). ‘A Bitter Aftertaste: How State Aid Affects Recipient Firms and Their Competitors in Europe’. IMF Working Paper No. 2024/250, https://www.imf.org/en/Publications/WP/Issues/2024/12/13/A-Bitter-Aftertaste-How-State-Aid-Affects-Recipient-Firms-and-Their-Competitors-in-Europe-559460
CEPEJ (2024). ‘Special File – Report European Judicial Systems – CEPEJ Evaluation Report – 2024 Evaluation Cycle (2022 Data)’. European Commission for the Efficiency of Justice, https://www.coe.int/en/web/cepej/special-file.
Draghi, M. (2024). ‘The Future of European Competitiveness: Part A – A Competitiveness Strategy for Europe’. European Commission, https://commission.europa.eu/topics/eu-competitiveness/draghi-report_en.
European Commission (2007). ‘Flexicurity’. https://employment-social-affairs.ec.europa.eu/policies-and-activities/eu-employment-policies/flexicurity_en.
European Commission (2025). ‘EU Competitiveness Compass’. https://commission.europa.eu/topics/eu-competitiveness/competitiveness-compass_en.
Gervais, H. (2025). ‘The Shocking Truth Behind US–EU Productivity Gap’. Urbest.io blog, 17 February, https://urbest.io/blog/the-shocking-truth-behind-us-eu-productivity-gap/.
Hanappi, T., Milot, V., & Turban, S. (2023). ‘How Does Corporate Taxation Affect Business Investment? Evidence from Aggregate and Firm-Level Data’. OECD Economics Department Working Paper Series No. 1765, https://www.oecd.org/en/publications/how-does-corporate-taxation-affect-business-investment_04e682d7-en.html.
IMF (2025). ‘World Economic Outlook, April 2025: A Critical Juncture amid Policy Shifts’. April, https://www.imf.org/en/publications/weo.
Kammer, A. (2024). ‘Europe’s Choice: Policies for Growth and Resilience’. IMF, European Department, https://www.imf.org/en/News/Articles/2024/12/15/sp121624-europes-choice-policies-for-growth-and-resilience.
Letta, E. (2024). ‘Much More than a Market: Speed, Security, Solidarity. Empowering the Single Market to Deliver a Sustainable Future and Prosperity for All EU Citizens’. https://www.consilium.europa.eu/media/ny3j24sm/much-more-than-a-market-report-by-enrico-letta.pdf.
Somani, R. (2021). ‘Public-Sector Productivity (1): Why Is It Important and How Can We Measure It?’. Governance: Equitable Growth, Finance and Institutions Insight, The World Bank Group, February, https://documents.worldbank.org/en/publication/documents-reports/documentdetail/099062623201136483.
World Bank Group (n.d.). ‘Bureaucracy Lab’. https://www.worldbank.org/en/about/unit/unit-dec/impactevaluation/partnerships/bureaucracy-lab.
