## United Kingdom–United States Aggregate Labor Productivity Decoupling

## Source details

**Canonical URL:** [United Kingdom–United States Aggregate Labor Productivity Decoupling](https://www.imf.org/-/media/files/publications/cr/2025/english/1gbrea2025002-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2025/english/1gbrea2025002-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2025/english/1gbrea2025002-source-pdf.pdf.json)

---

### Introduction and framing
- The UK’s productivity gap with the US has widened over the past two decades, with productivity growth rates decoupling after the Global Financial Crisis (GFC).
- In the early 2000s the UK experienced strong aggregate labor productivity growth, nearly matching the US's average annual rate of 2 percent.
- Since 2010 the UK has achieved only half the productivity gains seen in the US.
- Today, the UK’s aggregate labor productivity level is approximately 20 percent lower than in the US, while it was just about 10 percent lower in the early 2000s.
- After the GFC the US maintained a growth rate of around 1 percent, while the UK's rate fell to 0.5 percent.

### Data, measurement, and scope
- Primary firm-level data source: Compustat (publicly-listed firms, US and UK), covering 2000 to 2023.
- Sectoral analysis data source: EU-KLEMS (harmonized sectoral data constructed from national accounts), covering 2000–21.
- Labor productivity definitions:
  - In Compustat: real revenues per number of employees (revenues converted to a common currency using PPPs and deflated using one-digit sectoral deflators).
  - In EU-KLEMS: real sectoral value added per total employment.
- Limitations of Compustat:
  - Covers only publicly-listed firms (on average larger than the economy’s average firm).
  - Does not allow entry and exit of firms to be observed explicitly.
  - Financial sector firms are excluded.

### Macro-to-micro linkage and decomposition findings
- Macro decompositions attribute most of the UK’s slowdown in aggregate labor productivity growth to a decline in TFP growth; labor supply and capital intensity played a minor role.
- Microeconomic decomposition focus: between-firm reallocation versus within-firm efficiency.
- Key finding: evidence favors the within-firm (internal efficiency) explanation—UK publicly listed firms, especially frontier firms, lag US counterparts due to a significant decline in post-GFC TFP growth.

### Sectoral patterns and finance
- Sectoral focus (pre-GFC 2000–2008; post-GFC 2010–2021): manufacturing, construction, retail, finance, food services and ICT (these sectors make up more than 60 percent of total employment).
- Financial sector:
  - Pre-GFC, the UK’s financial sector had a high and likely unsustainable productivity contribution.
  - Post-GFC, the financial sector’s contribution turned from largely positive to negative and stands out relative to the US.
  - Main explanation: a leverage-driven boom pre-2008 that reversed with falls in asset prices, wealth, and higher uncertainty.

### Microeconomic drivers: TFP, frontier firms, intangible investment, and R&D
- Olley-Pakes decomposition (Compustat):
  - US average within-productivity growth: 2.5 percent per year.
  - UK average within-productivity growth: 0.9 percent per year.
  - Strong within-firm growth in the US is the main factor driving the widening productivity gap.
- Frontier firms (top decile):
  - UK frontier firms performed relatively well compared to other UK firms but not as well as US frontier firms.
  - UK frontier firms experienced a much stronger decline in TFP growth post-GFC than laggards; US frontier firms’ TFP growth distribution remained broadly unchanged.
  - UK frontier firms’ cumulative labor productivity growth has lagged US frontier firms since the GFC and fell below US laggard firms in cumulative terms.
- Intangible investment and R&D:
  - Pre-GFC UK firms’ investment rate into intangible capital: approximately 4 percent.
  - Post-GFC average intangible investment rate: US publicly-listed firms: 3 percent; UK: 1 percent.
  - By 2019 the UK's intangible capital stock level was lower than at the onset of the financial crisis.
  - Aggregate R&D spending as a share of GDP in the UK has lagged the US and other advanced countries.
  - Publicly listed UK firms spend less on R&D as a share of sales compared to US firms; gap pronounced in tech sectors.
  - Returns to knowledge capital are more dispersed in the UK than in the US.

### Drivers of lower firm-level R&D and scaling constraints
- Financing patterns:
  - UK listed firms rely more on debt financing relative to US counterparts; debt is less suited for R&D and intangible investment.
  - Net equity issuance: averaged approximately 10 percent in both countries in the four years preceding the GFC; post-GFC the difference widened with UK listed firms relying less on equity issuance.
- Market size and scaling:
  - Scaling up high-growth ICT firms is more difficult in the UK; UK venture capital in the ICT sector (2023, percent of GDP, decomposed by stage) lags the US and some other advanced economies.
  - The UK provides adequate early-stage funding but faces shortages in later-stage funding for scale-ups.

### Policy implications and recommendations
- Two-pronged strategy to revive UK productivity:
  1. Build on the UK’s strengths and revitalize traditional growth engines, especially the financial and ICT sectors.
  2. Foster a more conducive environment to business innovation through greater access to scale-up finance and continued efforts to retain high skilled individuals.
- Additional policy levers:
  - Address low business investment as a share of GDP.
  - Ease SME access to finance.
  - Improve translation of scientific advances into productivity gains.
  - Tackle skills mismatches and strengthen management practices.
- Specific firm-level recommendations:
  - Support scaling of high-growth firms (improving access to later-stage equity financing).
  - Encourage equity issuance and risk-taking by large firms.
  - Promote investment in intangible capital and R&D.
  - Facilitate reallocation of knowledge capital toward high-return firms.
- Reallocation policies can raise aggregate productivity, but primary UK–US divergence reflects within-firm inefficiencies.

### Brexit and trade openness effects on innovation and talent
- Lower trade openness and higher trade costs following Brexit may have reduced incentives to invest in R&D and adopt technology.
- Evidence cited:
  - Ampudia and Pardy (2023): firms’ rate of technological adoption decreased after Brexit.
  - Brexit complicated firms' ability to attract and retain global talent, impacting R&D human capital.

### Compustat and national accounts key statistics (Annex I)
- Compustat (average labor productivity growth and aggregate labor productivity growth):
  - UK pre- GFC (2000-2008): 3.1% (Average Labor Productivity Growth), 4.3% (Aggregate Labor Productivity Growth)
  - UK post- GFC (2010-2021): -0.3% (Average Labor Productivity Growth), -0.2% (Aggregate Labor Productivity Growth)
  - US pre- GFC (2000-2008): 3.4% (Average Labor Productivity Growth), 2.2% (Aggregate Labor Productivity Growth)
  - US post- GFC (2010-2021): 1.1% (Average Labor Productivity Growth), 0.9% (Aggregate Labor Productivity Growth)
- National accounts (ONS measures):
  - Growth Rate Output per Worker:
    - UK pre- GFC (2000-2008): 1.3%
    - UK post- GFC (2010-2021): 0.5%
    - US pre- GFC (2000-2008): 1.5%
    - US post- GFC (2010-2021): 1.4%
  - Growth Rate Output per Hour Worked:
    - UK pre- GFC (2000-2008): 1.7%
    - UK post- GFC (2010-2021): 0.5%
    - US pre- GFC (2000-2008): 2.1%
    - US post- GFC (2010-2021): 1%

*Source: IMF staff analysis (Compustat, EU-KLEMS, OECD datasets) as presented in the supplied content unit.*

### 1. United Kingdom-U nited States Aggregate Labor Productivity Decoupling ___________ 4

### United Kingdom-United States Aggregate Labor Productivity Decoupling

### Introduction and framing
- The UK’s productivity gap with the US has widened over the past two decades, with productivity growth rates decoupling after the Global Financial Crisis (GFC).
- In the early 2000s the UK experienced strong aggregate labor productivity growth, nearly matching the US's average annual rate of 2 percent.
- Since 2010 the UK has achieved only half the productivity gains seen in the US.
- Today, the UK’s aggregate labor productivity level is approximately 20 percent lower than in the US, while it was just about 10 percent lower in the early 2000s.
- After the GFC the US maintained a growth rate of around 1 percent, while the UK's rate fell to 0.5 percent.

### Data, measurement, and scope
- Primary firm-level data source: Compustat (publicly-listed firms, US and UK), covering 2000 to 2023.
- Sectoral analysis data source: EU-KLEMS (harmonized sectoral data constructed from national accounts), covering 2000–21.
- Labor productivity definitions:
  - In Compustat: real revenues per number of employees (revenues converted to a common currency using PPPs and deflated using one-digit sectoral deflators).
  - In EU-KLEMS: real sectoral value added per total employment.
- Limitations of Compustat highlighted:
  - Covers only publicly-listed firms (on average larger than the economy’s average firm).
  - Does not allow entry and exit of firms to be observed explicitly.
  - Financial sector firms are excluded given challenges with estimating TFP for these firms.

### Aggregate and macro-to-micro linkage
- Macro decompositions show most of the UK’s slowdown in aggregate labor productivity growth can be attributed to a decline in TFP growth; labor supply and capital intensity played a minor role.
- The paper takes a microeconomic and comparative approach to identify whether divergence stems from:
  - Between-firm reallocation (allocation of resources across firms), or
  - Within-firm efficiency (internal firm TFP improvements).
- Finding emphasized: evidence favors the within-firm (internal efficiency) explanation—the UK’s publicly listed firms, especially frontier firms, lag the US due to a significant decline in post-GFC TFP growth.

### Sectoral patterns and the role of finance
- Sectoral decomposition uses pre-GFC (2000–2008) and post-GFC (2010-2021) periods and focuses on manufacturing, construction, retail, finance, food services and ICT (these sectors make up more than 60 percent of total employment).
- Financial sector dynamics:
  - Pre-GFC, the UK’s financial sector had a high and likely unsustainable productivity contribution.
  - Post-GFC, the financial sector’s contribution turned from largely positive to negative and stands out relative to the US.
  - Main explanation: a leverage-driven boom that artificially boosted profits and income pre-2008; when the GFC hit, falls in asset prices, wealth, and higher uncertainty reversed those gains and exposed structural weaknesses.

### Microeconomic drivers: TFP, intangible investment, and innovation
- Firm-level TFP and productivity patterns:
  - Aggregate evidence and Compustat firm-level data show the productivity gap between UK and US-listed firms widened over time, with divergence after the GFC.
  - UK frontier firms performed relatively well compared to other UK firms but not as well as US frontier firms.
- Potential contributors to subdued UK TFP growth post-GFC:
  - Reduced investment in intangible capital following the GFC.
  - Lower R&D spending compared to the US.
  - Business dynamism declines and resource misallocation.
  - Skills gaps and weaker management practices that hinder technology adoption and organizational change.

### Policy implications and recommendations
- A two-pronged strategy is suggested to revive UK productivity:
  1) Build on the UK’s strengths and revitalize traditional growth engines, especially the financial and ICT sectors.
  2) Foster a more conducive environment to business innovation through greater access to scale-up finance and continued efforts to retain high skilled individuals.
- Additional policy levers discussed in the surrounding analysis include addressing low business investment as a share of GDP, easing SME access to finance, improving the translation of scientific advances into productivity gains, and tackling skills mismatches.

### Analytical contributions and novelty
- This paper adds to the literature by:
  - Conducting a cross-country (UK–US) comparison rather than a solely domestic analysis.
  - Focusing on publicly-listed firms to shed light on large-firm contributions to aggregate trends.
  - Decomposing aggregate changes into sectoral and firm-level between/within components and linking slower UK TFP growth to lower intangible investment and R&D.

*Prepared by Leonardo Indraccolo; International Monetary Fund.*

### 12.      Beyond the financial sector, wholesale and retail trade has also contributed to the

### 12.      Beyond the financial sector, wholesale and retail trade has also contributed to the decoupling.

### Sectoral contributions to the productivity decoupling
- Wholesale and retail trade contributed to widening productivity gaps between the UK and the US. The sector’s lower contribution to the UK’s productivity growth after the GFC resulted from lower sectoral productivity growth, rather than a declining sectoral share in the economy.
- The decline of the UK manufacturing sector explains the productivity slowdown over time, but not the decoupling from the US: both the UK and the US saw manufacturing’s contribution to productivity growth decline after the GFC, indicating common structural forces (e.g., competition from lower-cost global manufacturers).  
- The ICT sector showed sustained productivity growth in the UK over two decades but could not offset slowdowns elsewhere. The ICT sector’s recent signs of slowdown are documented (OECD 2019; Ilzetzki 2020; Pittaway 2025). A potential constraint is lagging access to digital infrastructure (e.g., high-speed internet), and limited diffusion of skills needed to leverage new technologies (Tuckett and others 2017; Deb and Li 2024; Criscuolo and others 2021).

### Resource misallocation versus within-firm performance
- Data note: This section is based primarily on the Compustat dataset, which does not cover financial sector firms and firms that are not publicly listed.
- Aggregate productivity decomposes into within-firm (average firm efficiency) and between-firm (allocation of resources across firms) components following Olley and Pakes (1996).
- Olley-Pakes decomposition results:
  - Between-firm factors contribute negatively to the level of aggregate productivity in the US, while explaining a much smaller fraction in the UK.
  - Despite a larger negative between-firm contribution in the US, aggregate productivity is higher in the US because of stronger within-firm productivity growth.
  - Quantified divergence: US average within-productivity growth was 2.5 percent per year, versus 0.9 percent for the UK over the sample period. Strong within-firm growth in the US was the main factor driving the widening productivity gap between the two countries.
- Additional stylized facts:
  - The share of workers employed in the lowest productivity decile is higher in the UK than in the US (post GFC).
  - Over time, the US experienced stronger reallocation towards less productive firms after the GFC, with a sharper decline in employment shares in the most productive firms (top two deciles).

### Drivers of within-firm productivity divergence: TFP and capital intensity
- Approach: Within-firm labor productivity growth is decomposed into capital intensity and TFP components by estimating firm-level production functions.
- Comparative patterns:
  - Growth in capital per worker has been broadly flat in both countries post-GFC.
  - TFP growth diverged: US firms’ annual TFP growth declined from approximately 2.5 percent to 1 percent after the GFC; the decline was more pronounced for UK businesses and explains most of the UK–US divergence.
- Frontier firms:
  - Frontier firms (top decile of labor productivity) experienced a much stronger decline in TFP growth in the UK than among laggards; US frontier firms’ TFP growth distribution remained broadly unchanged.
  - UK frontier firms’ cumulative labor productivity growth has lagged that of US frontier firms since the GFC and fell below US laggard firms in cumulative terms.

### Intangible investment and R&D as proximate causes
- Intangible capital and R&D matter for innovation-led TFP growth, especially in ICT where innovations are often intangible (software, code, algorithms).
- Investment patterns:
  - Pre-GFC, UK firms’ investment rate into intangible capital was approximately 4 percent, similar to the US.
  - Post-GFC, US publicly-listed firms maintained an average investment rate of 3 percent in intangible assets, whereas in the UK the rate fell to 1 percent.
  - By 2019, the UK's intangible capital stock level was lower than at the onset of the financial crisis.
- R&D spending:
  - Aggregate R&D spending as a share of GDP in the UK has lagged the US and other advanced countries, despite generous R&D tax policy measures.
  - Publicly listed UK firms spend less on R&D as a share of sales compared to US firms; this gap is particularly pronounced in tech sectors that have helped sustain US productivity growth.
  - Returns to knowledge capital are more dispersed in the UK than in the US (implying potential gains from reallocating knowledge capital from low- to high-return firms).
- Definition/measurement notes:
  - Frontier firms are those in the top decile of the labor productivity distribution; middle firms are between the 90th and 50th percentiles; laggard firms are the bottom 50 percent.
  - The firm’s investment rate in this discussion is defined as the growth rate of intangible capital.

### Possible determinants of lower firm-level R&D investment
- Access to finance:
  - UK listed firms rely more on debt financing relative to US counterparts. Debt financing is less suited for financing R&D and intangible investment because these assets are hard to collateralize and returns are riskier/longer‑dated.
  - Net equity issuance: After averaging approximately 10 percent in both countries in the four years preceding the GFC, the difference in equity issuance has widened over time (UK listed firms rely less on equity issuance).
  - Less reliance on equity financing implies UK’s largest firms have less capacity for undertaking riskier projects, which may have contributed to subdued R&D spending.
- Market size and scaling:
  - Difficulties in scaling up high-growth firms, especially in ICT, may limit incentives to invest in R&D because firms that cannot grow cannot recoup R&D costs in larger markets.
  - UK venture capital in the ICT sector (2023, percent of GDP, decomposed by stage) has recently lagged not only the US but also other advanced economies like Canada and Denmark. The UK provides adequate early-stage funding but firms that need later-stage funding to scale often prefer the US where more funds are available.

### Implications and policy-relevant points
- The decoupling from the US is largely explained by weaker within-firm TFP growth in the UK, concentrated among frontier firms, and linked to lower investment in intangible capital and R&D.
- Policies that:
  - Support scaling of high-growth firms (improving access to later-stage equity financing),
  - Encourage equity issuance and risk-taking by large firms,
  - Promote investment in intangible capital and R&D,
  - Facilitate reallocation of knowledge capital toward high-return firms,
could help boost within-firm productivity and narrow the UK–US productivity gap.
- Reallocation policies that move workers and financial capital toward the most productive businesses could also raise aggregate productivity, though the primary UK–US divergence reflects within-firm inefficiencies rather than pure resource misallocation.

*Source: IMF staff analysis (Compustat, EU-KLEMS, OECD datasets) as presented in the supplied content unit.*

### 25.      Lower trade openness and higher trade costs following Brexit may have also

### 25.      Lower trade openness and higher trade costs following Brexit may have also 

### Impact of Brexit on R&D, innovation, and talent
- Trade barriers and reduced openness can limit firms’ market size, lowering incentives to invest in R&D that yields higher returns in larger markets (as noted in IMF (2024b)).
- Higher trade costs following Brexit have negatively affected firms' productivity by lowering incentives to spend on R&D and develop better products to compete on the global market (Bloom and others (2019)).
- Ampudia and Pardy (2023) find that firms’ rate of technological adoption decreased after Brexit due to future bureaucratic costs and decreased demand.
- Hiring highly skilled individuals is a key component of R&D investment; Brexit has complicated firms' ability to attract and retain global talent by making immigration to the UK more difficult (Van Reenen 2016).

### G. Conclusions and Policy Recommendations — Revitalizing the UK’s Traditional Growth Engines
- Strategic pillar: build on sectoral strengths such as finance and ICT.
- Financial sector:
  - It is unlikely the financial sector will again make the same contribution to productivity growth as pre-GFC without unsustainable leverage increases.
  - Authorities have initiated measures to enhance the financial sector's contribution to growth.
  - Reforms aimed at streamlining data collection and revising firms’ listing requirements have potential to boost sector efficiency.
  - Simplification of existing regulatory rules should be conducted cautiously to preserve financial stability.
- ICT sector and AI:
  - The IMF's AI readiness index shows the UK is well-positioned to capitalize on AI technologies, with the index score surpassing the average of other advanced economies because of the large share of workers employed in cognitive-intensive jobs (Cazzaniga and others 2024).
  - Authorities’ reforms in construction planning are expected to speed up delivery of critical infrastructure needed for AI development.
  - Improving skills can allow wider adoption of digital technologies so productivity gains from AI are not concentrated in too few firms.
  - Expanding tax credits or tax allowances for SMEs that invest in employees’ training can facilitate upskilling and narrow skill gaps.

### G. Conclusions and Policy Recommendations — The Road Ahead: Boosting Firm Innovation
- Strategic pillar: foster firms’ innovation by improving access to scale-up finance and retaining talent.
- Access to scale-up finance:
  - Start-ups and young businesses benefit from a vibrant venture capital market, but scaling up funding can be more challenging in the UK compared to the US.
  - Policies that incentivize participation of institutional investors, like pension funds, in domestic venture capital markets can help high-growth firms expand domestically.
  - Authorities’ plans to consolidate pension funds are welcome as they have potential to expand access to diverse asset classes.
- Innovation hubs and university-business collaboration:
  - Ongoing efforts to create innovation hubs and intensify university-business collaboration support development of commercializable ideas.
  - Hiring workers with high human capital and advanced skills is crucial; talent retention has become harder.
  - Measures that help retain talent and encourage labor mobility of high-skilled workers should be prioritized to provide incentives for researchers and highly educated individuals to come and stay in the UK.

### Annex I. Productivity Trends in Compustat Data and National Accounts — Key statistics
- Compustat (average labor productivity growth and aggregate labor productivity growth):
  - UK pre- GFC (2000-2008): 3.1% (Average Labor Productivity Growth), 4.3% (Aggregate Labor Productivity Growth)
  - UK post- GFC (2010-2021): -0.3% (Average Labor Productivity Growth), -0.2% (Aggregate Labor Productivity Growth)
  - US pre- GFC (2000-2008): 3.4% (Average Labor Productivity Growth), 2.2% (Aggregate Labor Productivity Growth)
  - US post- GFC (2010-2021): 1.1% (Average Labor Productivity Growth), 0.9% (Aggregate Labor Productivity Growth)
- National accounts (ONS measures):
  - Growth Rate Output per Worker:
    - UK pre- GFC (2000-2008): 1.3%
    - UK post- GFC (2010-2021): 0.5%
    - US pre- GFC (2000-2008): 1.5%
    - US post- GFC (2010-2021): 1.4%
  - Growth Rate Output per Hour Worked:
    - UK pre- GFC (2000-2008): 1.7%
    - UK post- GFC (2010-2021): 0.5%
    - US pre- GFC (2000-2008): 2.1%
    - US post- GFC (2010-2021): 1%

### Appendix I. Mathematical Appendix — Analytical frameworks and decompositions
- Decomposition of aggregate labor productivity growth by sector:
  - Aggregate labor productivity growth in year t approximated as: G_prod_d_t ≈ Σ_{i=1}^N y_{i,t-1} * g_prod_d_{i,t}, where y_{i,t-1} is nominal value-added share of sector i in year t−1 and g_prod_d_{i,t} is sector i labor productivity growth rate.
  - Nominal value added shares computed as shares of total value-added of the non-agricultural market economy; sectors at NACE 1-digit level.
- Olley-Pakes decomposition:
  - Aggregate productivity P_t decomposed into within-firm (average firm productivity) and between-firm (covariance between firm productivity and employment share) components.
  - Higher covariance between firm-level productivity and employment share implies more efficient resource allocation.
- Firm-level production function and TFP estimation:
  - Cobb-Douglas in logs: y_it = α + β_l l_it + β_k k_it + ω_it + e_it, with ω_it unobserved TFP.
  - Estimation challenges due to correlation between ω_it and inputs; employ non-parametric approach used by IMF (2024b) and developed by Gandhi and others (2020).
- Decomposition of labor productivity growth into capital intensity and TFP:
  - Δprod_it = (β_l − 1) Δl_it + β_k Δk_it + Δu_it, and averaging across firms under constant returns to scale (β_k = 1 − β_l) yields:
    - Δprod_t = (1 − β_l)(Δk_t − Δl_t) + Δω_t
  - First term: growth in capital intensity; second term: growth in TFP.

*UNITED KINGDOM, INTERNATIONAL MONETARY FUND.*

### 1.      There has been a recent resurgence of IP in advanced economies, while it continues to

### There has been a recent resurgence of IP in advanced economies, while it continues to

### Recent international trends and policy context
- Since 2021, the United States has implemented a large program of fiscal support for manufacturing of electric vehicles, clean energy projects, batteries and semi-conductors through the Inflation Reduction Act (2022) and CHIPS Act (2022).
- European countries have increased use of IP over the past decade via the state aid framework and EU-level initiatives such as the Green Deal Industrial Plan, Batteries Strategy and European Chips Act.
- China has continued industrial policy via the Made in China 2025 program, using state-owned enterprises and government subsidies to targeted sectors.
- Advanced Economies (AEs) have increased use of IPs sharply since 2019.
- Restrictive trade practices have also increased globally over the past five years.

### Motivations for the recent resurgence of IP
- Primary objectives motivating renewed IP:
  - Boost growth.
  - Reinforce economic security (e.g., onshoring supply chains for semi-conductors; CHIPS Act).
  - Facilitate the green transition (e.g., US subsidies for electric vehicles; around half of non-crisis state aid in the EU directed to green initiatives on average during 2014–2022).
- Additional context:
  - In Europe, IP is being used against weak growth, low productivity and an ageing population (IMF2024a).
  - The recent escalation of trade tensions could further increase onshoring and protective IP efforts.

### Instruments and targeting of modern industrial policy (IP)
- Modern IP is defined as ‘vertical’ policies targeted at one or more sectors to achieve government objectives, differing from ‘horizontal’ economy-wide policies.
- Policy instruments cited include:
  - Trade policy (tariffs and export subsidies).
  - Production subsidies or grants to firms.
  - Support for Research and Development (R&D) via grants and tax incentives.
  - Special Economic Zones (SEZs).
  - Financing on non-commercial terms (concessional loans, equity injections, guarantees).
  - Regulatory policies such as local content requirements.
  - Direct government procurement limited to domestic suppliers.
- Subsidies account for the majority of new IP in advanced economies, particularly grants and loans to firms.

### The UK experience: recent history and the new industrial strategy
- Prior to the new industrial strategy, the UK emphasized horizontal policies (e.g., generous R&D tax incentives and full expensing of business investment in plant and machinery).
- Vertical interventions since the pandemic were limited and included targeted grants for advanced manufacturing, creative industries and a Green Industries Growth Accelerator, with a fiscal cost of only around 0.1 percent of GDP in FY2023/24.
- Comparative fiscal context:
  - Average EU country spends 1.5 percent of GDP per year on state aid.
  - Conservative estimates indicate the total cost of Chinese IP at 1.7 percent of GDP.
- UK place-based and zone policies:
  - Special Economic Zones called ‘free ports’, 8 Investment Zones and 48 Enterprise Zones with tax concessions or other support.
- Timeline and scope of the new UK strategy:
  - Public consultation completed in November 2024.
  - New industrial strategy published in June 2025.
  - Strategy focused on eight priority sectors: advanced manufacturing; clean energy; creative industries; defense; digital and technologies; financial services; life sciences; professional and business services.
  - Ten-year strategy to be implemented with an Industrial Strategy Advisory Council for monitoring and evaluation.
  - The strategy includes both vertical and horizontal policies, and some initiatives will entail fiscal costs within the spending review’s envelope.
- Specific fiscal commitments described:
  - £4.3 bn in public support for advanced manufacturing (0.15 percent of FY2025/26 GDP), including £2.8 bn in funding for R&D.

### Role and design of public financial institutions in the UK strategy
- Objective: public financial institutions to catalyze private investment while minimizing fiscal cost; supported portfolios required to generate a return at least equal to the cost of government borrowing per the Financial Transactions Control Framework.
- Institutions and endowments:
  - National Wealth Fund (NWF):
    - Successor to the UK Infrastructure Bank.
    - Endowment of around £27.8 billion (approximately 1 percent of 2025 GDP).
    - Will use loans, guarantees and equity investments to crowd in private investment with a target 3:1 mobilization ratio.
    - Key objective is to generate “additionality” by unlocking projects that would not otherwise go ahead.
  - British Business Bank:
    - State-owned bank supporting SMEs.
    - Allocated an additional £4 bn of capital, bringing its total endowment to £25.6 bn.
    - Targets a 3:1 mobilization ratio for private finance in the eight priority sectors.
  - UK Export Finance Agency:
    - Provided with an additional £3 bn of lending capacity, bringing available funds to £13 bn to incentivize foreign demand for UK exports.

### Conceptual framework: benefits of IP and market failures addressed
- Ultimate goal: boost productivity and growth in sectors that will not reach potential without public intervention due to market failures.
- Market failures IP can alleviate:
  - Externalities: positive spillovers (e.g., agglomeration effects, knowledge spillovers in green industries). Example: production subsidies in the Inflation Reduction Act and CHIPS Act for electric vehicles and semiconductors.
  - Imperfect Competition: natural monopolies with high fixed costs (e.g., utilities), where IP can finance fixed costs, incentivize entry, and promote competition.
  - Information Asymmetries: nascent firms face difficulty accessing finance when technologies are uncertain; public concessional finance can bridge this gap (example: KfW support for solar panel industry in early 2000s).
  - Delayed Returns: long R&D horizons reduce private investors’ willingness to finance; public guarantees, equity, grants and loans can raise risk-adjusted returns.
  - Coordination Failures: complex supply chains require simultaneous development; government coordination can enable industry emergence (example: establishment of the Airbus Consortium).

### Risks and potential harms from poorly designed IP
- Key economic and fiscal risks:
  - Picking losers: public support to firms close to technological frontier may fail if R&D is unsuccessful; support to established firms may shield them from competition and disincentivize innovation.
  - Severe market distortions: IP that favors particular firms can draw resources away from competitors, crowd out unsubsidized projects, or shrink market size; IMF staff analysis of EU state aid (including UK pre-2020) found evidence consistent with these risks.
  - Governance failures: political pressure can drive rent-seeking and misallocation of public support away from market-failure-driven needs.
  - Fiscal costs and inflation: production subsidies not subject to strict budgetary limits can entail substantial fiscal costs (example context: the CHIPS Act, Inflation Reduction Act and Infrastructure Investment and Jobs Act).
- Policy implication emphasized by the paper: careful design, monitoring and evaluation are essential to manage these risks and ensure IP generates productivity gains and additional private investment.

*Prepared by Andrew Hodge (EUR) and Samuel Pienknagura (RES); contributions by Kristina Kostial, Daria Zakharova, Luc Eyraud, Romain Duval, Florence Jaumotte, Aleksandra Zdzienicka, Sergio Rodriguez, Oliver Exton, Maximilien Queyranne, Ed Hearne, Graham Prentice and participants in an IMF internal seminar.*

### 0.3 percent of GDP per year on average in direct government spending (Seydl, Matthews and

### Successful IP: Lessons from International Experience

### Trade distortions and macro costs
- IP can be inflationary when it takes the form of trade measures such as tariffs.
- Trade distortions and retaliation:
  - IP can distort the sectoral make-up of the economy away from comparative advantage (Hodge and others, 2024; Ossa, 2011).
  - Distortions can create supply gluts in supported sectors, generating adverse spillovers to trading partners and prompting retaliation or competing IP, starting a counter-productive “subsidy race.”
- Fiscal and direct cost examples:
  - Direct government spending estimated at 0.3 percent of GDP per year on average (Seydl, Matthews and Schaeffer, 2023).
  - Cost of Chinese IP estimated to be around 1.5 percent of GDP in 2019 (DiPippo and others, 2019).

### Lesson #1 — Targeting and concentration risks
- Core finding:
  - IP is most likely to support productivity when implemented using evidence-based targeting, rather than blanket sector-wide support.
- Policy guidance:
  - Public support should be targeted and temporary; blanket support risks subsidizing activity that would occur without government intervention and can crowd out private investment.
  - Firms selected for IP support should be chosen based on a rigorous evaluation process, including analysis of where market failures exist, so that IP catalyzes new economic activity.
- Best-practice example:
  - Advanced Research Projects Agency (ARPA) model: programs run by a director selected from academia or industry, who selects programs based on extensive consultation with firms and researchers (Juhasz, Lane and Rodrik, 2023b).
- Concentration risk:
  - Avoid “picking winners” (concentrating support on one or two firms) because of high mistake/failure risk, especially for new technologies.
  - Negative examples: Synthetic Fuels Corporation (US, 1980's) and Solyndra Corporation guarantee (US, 2009).
  - Positive example: Operation Warp Speed (2020) diversified support across multiple vaccine researchers/manufacturers, managing concentration risk.

### Lesson #2 — Governance, selection and non-discrimination
- Governance and selection:
  - Strong governance arrangements and technical expertise in selection improve targeting and effectiveness and reduce discriminatory, politically driven support.
  - Enlist academic and industry expertise or delegate decisions to external experts to keep targeting evidence-based (ARPA and Peru sectoral roundtables examples).
- Non-discrimination principles:
  - Consider all firms irrespective of age, size or ownership to avoid excluding firms with greatest potential.
  - Rationale:
    - New entrants may face market failures and be more innovative than established firms.
    - Supporting only existing firms can increase crowding out and reduce incentives for innovation.
    - Small and big firms both merit consideration; new entrants are often small and innovative.
    - Both domestic and foreign-owned firms should be eligible to avoid shielding domestic firms from competition or disincentivizing foreign investment (IMF 2024b).

### Lesson #3 — Containing fiscal costs and time-bounding support
- Fiscal design principles:
  - Limit support by making it time bound and subject to financial limits and ongoing monitoring.
  - Use grants with set limits, equity investments, loans or guarantees (recognizing contingent liability risks).
  - Avoid open-ended subsidy programs that make fiscal costs difficult to project.
- Examples and evidence:
  - EV subsidy programs in China and the US proved more popular and costly than envisaged, reflecting open-ended program design (IMF 2024a).
  - China: consumers received subsidies for purchase of EVs based on vehicle range during 2013–22.
  - US: tax credits allowed for purchase of EVs.
- Monitoring and exit:
  - Require recipients to meet performance benchmarks to unlock continued support; design support to expire once objectives are achieved.
  - Under ARPA, supported projects have pre-defined milestones for continued public support (Juhasz, Lane and Rodrik, 2023b).
  - Time limits and periodic reviews preferred to indefinite support that can shield firms from competition.

### Lesson #4 — Need for strong horizontal policies and sequencing
- Preconditions and sequencing:
  - High-quality horizontal policies are a prerequisite for successful IP; vertical interventions are not a substitute for strong horizontal policies.
  - Vertical IP should be considered only if strong horizontal policies are in place and prove insufficient to address market failures.
- Horizontal complements:
  - Skills, high-quality public infrastructure, active labor market policies, and immigration policies supply critical inputs to firms benefiting from IP (IMF 2024b).
  - Public infrastructure examples: access to the electricity grid and EV charging infrastructure complement EV purchase subsidies in the US and China.
- Empirical note:
  - Vertical interventions amplify their benefits when the general business environment provides policy certainty, stable tax and regulatory regimes, including planning laws.

### Targeting UK IP: Evidence-based approaches
- Challenge:
  - Identifying market failures and demonstrating ‘additionality’ is theoretically sound but practically challenging.
- Quantitative approaches:
  - Two broad approaches:
    1. Direct tests of additionality: estimate IP impact versus a counterfactual (e.g., econometric diff-in-diff or comparing similar firms with/without support); data intensive and hard to apply ex ante.
    2. Proxy metrics for market failures: ex ante indicators such as measures of competition or price mark-ups; easier to apply before approval.
  - These quantitative methods should be supplemented by descriptive diagnostics of specific market failures.
- Example metric and UK findings:
  - Liu (2019) “Distortion Centrality (DC)” metric gauges market failures based on industry-level price mark-ups and economic linkages; industries with higher DC index (close to 1 and above) propagate distortions.
  - Data for the UK suggest manufacturing of electrical equipment, metal products, paper products, and warehousing (critical for supply chains) have among the largest distortions and are likely to exhibit market failures (Figure 4).
  - Services such as accommodation, food services and real estate display the least evidence of distortions and likely exhibit fewer market failures; IP in these sectors risks crowding out private investment.
- Policy implication:
  - Use DC index and complementary diagnostics to pinpoint sectors (e.g., UK advanced manufacturing) where IP is most warranted and to design appropriate interventions.

### Modeling the impact of UK IP on productivity
- Rationale:
  - Structural general equilibrium modeling accounts for spillovers and spillbacks across industries and international borders when assessing IP’s effects.
- Context:
  - UK productivity slowdown: trend growth of per capita income sharply slowed since the Global Financial Crisis, leaving UK GDP about a quarter below the pre-GFC trend.
  - Labor productivity growth declined from around 2 percent pre-GFC to approximately ½ percent since then.
  - Contributors include chronic under-investment and lack of firm-level innovation.
- Modeling insights:
  - Krugman-style model of Lashkaripour and Lugovskyy (2023) with industry-specific scale externalities of firm entry (agglomeration/knowledge spillovers) implies:
    - Industrial production subsidies targeted at industries with significant scale externalities can overcome market failures and increase productivity by incentivizing entry.
    - Impact differs across sectors depending on the extent of market failures.
  - Modeling distinguishes increases in labor productivity from total factor productivity for convenience (labor is the only production input in the model); in practice, labor productivity gains can derive from higher total factor productivity and higher capital intensity.
  - IMF staff empirical analysis shows IP can have a greater effect on capital accumulation than on total factor productivity of firms (Baquie and others, 2025).

*Source: Excerpt from IMF chapter on the United Kingdom (text as provided).*

### 28.      The model reveals that subsidies targeted at advanced manufacturing industries in the

### 1gbrea2025002-source-pdf - 28.      The model reveals that subsidies targeted at advanced manufacturing industries in the

### Model findings: sectoral and economy-wide productivity effects
- Subsidies targeted at advanced manufacturing industries in the UK that exhibit large scale externalities will support sectoral and economy-wide productivity; subsidies directed at other sectors will be less effective.
- Production subsidies correcting market failures in an advanced manufacturing industry with large externalities (example given: the UK electrical and optical equipment sector):
  - Incentivize firm entry into the sector, increasing its share of production and total employment.
  - Given the large scale externalities, the productivity of the expanded electrical and optical equipment sector increases by 1½ percent in the long-run, raising economy-wide productivity (Table 2).
- Production subsidies that correct market failures in sectors with smaller externalities (examples: transport and plastics, agriculture):
  - Expand less-productive industries while failing to increase their productivity significantly.
  - These effects can be viewed as adverse cross-industry spillovers by drawing resources towards less productive sectors.

### International spillovers and spillbacks through trade channels
- IP can lift productivity of sectors with large market failures but can cause unintended international spillovers and spillbacks via trade channels.
- Expansion of production in targeted sectors (notably niche industries or advanced manufacturing of differentiated products near the technological frontier) can create supply gluts in export markets, depressing export prices and worsening the terms of trade.
- Hypothetical simulation: if the UK implements industry-specific production subsidies that completely eliminate scale externalities in every sector while other countries maintain zero subsidies:
  - Productivity benefits of correcting market failures can be outweighed by trade-channel losses, generating an overall income loss for households.
  - Household income could decline in many countries, particularly those more open to trade, such as the UK, because of spillbacks through trade channels (Figure 5).
- Retaliation by trading partners, including with trade policy, could further reduce welfare (not shown in this paper but referenced).

### Table 2 — Model-implied productivity impacts of externality-correcting IP subsidies in selected UK industries
- IP Subsidy (in percent of output price)* and Improvement in industry productivity (ppts, weighted by industry share)**:
  - Electrical & Optical Equipment: 55 ; 1.4
  - Chemicals and Chemical Products: 23 ; 0.8
  - Basic Metals & Fabricated Metal: 21 ; 0.3
  - Rubber and Plastics: 14 ; 0.1
  - Transport Equipment: 13 ; 0.2
- Notes:
  - * Larger subsidies are required when scale externalities are higher.
  - ** Denotes productivity improvement in the model's steady state. Dynamic effects are beyond the scope of the model.

### Policy implications and recommended design features for UK IP
- Core conclusion: careful targeting of UK IP, under strict financial controls and with international cooperation on IP design, can help manage spillovers and spillbacks and maximize benefits.
- Targeting:
  - Greatest chance of success if targeted at firms and projects exhibiting market failures, most likely in sectors like advanced manufacturing and industries critical to supply chains.
  - Quantitative metrics provide a useful guide to demonstrate additionality of IP support; proxy metrics suggest a role for UK IP in industries close to the technological frontier (digital and technology industries, green industries, clean energy).
- Institutional arrangements:
  - Harness participation of academic and industry experts in the UK’s Industrial Advisory Council to design IP programs, select firms/projects, and oversee performance monitoring.
  - Consider aspects of the ARPA model: expert-managed programs with performance milestones that unlock additional financial support.
- Fiscal controls:
  - Contain fiscal costs through targeted grant programs that are time bound with a fixed budget allocation.
  - Rigorous monitoring with clear performance benchmarks; prefer lump sum grant programs over open-ended tax incentives or subsidies.
  - This approach helps accommodate costs within the medium-term expenditure envelope announced at the spending review.
- Horizontal policies:
  - Structural reforms in planning, skills, and infrastructure are pre-requisites and complements to IP.
  - Ongoing UK investment in skills and training, partnership with businesses for eight priority sectors, continued skilled migration, infrastructure (ports), and planning law reforms are necessary complements to targeted IP.

### Realistic expectations, fiscal constraints, and the NWF challenge
- Realistic expectations:
  - IP is supportive but other policies (horizontal reforms, stable policy environment) will likely be substantially more important for boosting productivity.
  - UK has limited fiscal space given high debt and interest costs and ageing-related spending pressures.
- National Wealth Fund (NWF) challenges and guidance:
  - The NWF aims to catalyze new investment but faces the key challenge of achieving ‘additionality’ while covering financing costs and managing fiscal risks.
  - The bar for achieving the 3:1 catalyzation ratio is high because projects that both demonstrate additionality and generate returns to cover the NWF’s financing costs will likely be close to commercially viable without public support.
  - Recommended enhancements:
    - Rigorous ex ante appraisal processes: cost-benefit analysis, financial viability studies, fiscal risk analyses.
    - Evidence-based approaches to demonstrate additionality using quantitative proxies (e.g., low competition metrics) alongside narrative approaches.
    - Robust risk management framework with quantitative thresholds and at least annual engagement between NWF and HMT.
    - An economic capital limit of £7 billion to ensure losses on unsuccessful projects can be absorbed without requiring additional HMT financial support.
    - Detailed and transparent risk reporting integrated into HMT reporting and considered in OBR fiscal risks reports.

### Annex highlights: identifying market failures and model structure
- Distortion-centrality (DC) index (Liu (2019)):
  - DC combines sector-specific distortions with sector interconnections using input-output tables to proxy for market failures or distortions.
- Alternative measures and results:
  - IMF staff constructed the index for 141 countries, measuring distortions using markups (Ackerberg and others, 2015); robust to using external financial dependence (Rajan and Zingales, 1998).
  - Sectors like electricity production and biotech research rank high on both dimensions.
  - Staff finds that the number of protectionist IPs targeting a sector correlate positively with its DC, with a stronger relationship in recent years and stronger for AEs.
- Structural model (Lashkaripour and Lugovskyy (2023)):
  - Multi-country, multi-industry Krugman-style model where IP is modeled as subsidies removing distortions from external economies of scale at the sectoral level.
  - Industries differ by degree of scale economies or elasticity of trade volumes to prices.
  - Labor is the sole factor of production; workers are perfectly mobile across industries within countries but immobile internationally.
  - Consumers choose industry-level product bundles with nested CES aggregators; parameters σ_k (elasticity of substitution between country-specific varieties) and γ_k (firm-level market power, with γ_k >1) govern demand elasticities and love-of-variety preferences.

*Source: Excerpt from the provided IMF content unit.*

### 3. Firms and Production. Each country 푖푖∈퐶퐶

### 3. Firms and Production. Each country 푖푖∈퐶퐶

### Firms, production technology, and market structure
- Each country 푖푖 is populated with a mass 푀푀푖푖,푘푘 = Ω푖푖,푘푘 of monopolistically competitive firms producing a single product in industry 푘푘∈퐾퐾, using labor as the only factor of production.
- Under free entry, a pool of ex ante identical firms pays an entry cost 푤푤푖푓푘푒 to operate in industry 푘푘 from country 푖푖.
- Each firm 휔휔∈Ω푖푖,푘푘 draws a random productivity 푧푧(휔휔) ≥ 1 from distribution 퐺퐺푖푖,푘푘(푧푧).
- Firms face a marginal cost τ푖푖푗푗,푘푘 · w푖푖 / z(휔휔) for producing and delivering goods to destination 푗푗∈퐶퐶̅, where τ푖푖푗푗,푘푘 is a flat, iceberg transport cost.
- Average unit labor cost in origin 푖푖 declines with the number of firms and varieties.

### Scale elasticity, markups, and producer prices
- The elasticity of the Producer Price Index (PPI) of the composite good in industry 푘푘 produced in country 푖푖 to the number of firms is negative: −휇휇푘푘 = 휕휕푙푙 휕휕푃푃푖푖푖푖,푘푘 / 휕휕푀푀푖푖푘푘 < 0.
- The absolute value 휇휇푘푘 is the industry-level scale elasticity: the elasticity of the PPI to the number of firms.
- 휇휇푘푘 is equivalent to the elasticity by which variety-adjusted total factor productivity increases with industry-level employment, proportional to firm mass 퐿퐿푖푖,푘푘 ∝ 푀푀푖푖,푘푘.
- The scale elasticity equals a constant firm-level mark-up in the industry: (훾훾푘푘 − 1) / 1, which also determines the extent of love of variety.

### Policy instruments in the model
- The simulations use a version of the Lashkaripour and Lugovskyy (2023) model where the only policy instrument available to a government in each country 푖푖 is an industry-specific production subsidy applied to industry 푘푘’s output produced in country 푖푖, irrespective of where the output is sold.
- The subsidy is financed via lump sum taxes to consumers.
- Note: The model can be extended to introduce intermediate goods.

### Welfare implications of industrial policy (IP)
- Welfare implications of unilateral industrial subsidies are quantitative and depend on calibration parameters, notably the scale elasticity 휇휇푘푘 and the price-elasticity of foreign demand 휎휎푘푘.
- Lashkaripour and Lugovskyy (2023) find that 휎휎푘푘 and 휇휇푘푘 are negatively correlated empirically.
- Implications of production subsidies:
  - (i) Correct misallocation by expanding output in high returns-to-scale (high 휇휇푘푘) industries.
  - (ii) Worsen the terms of trade by expanding exports in these same industries which have lower price-elasticity of foreign demand (low 휎휎푘푘).
- Under these parameter values, unilateral implementation of Pigouvian industrial subsidies worsens the terms of trade, which can offset some or all of the welfare gains from establishing allocative efficiency. This effect is referred to as an ‘immiserizing’ welfare effect.

### Data and calibration
- Calibration uses observable data from 43 countries (including all EU member states and the UK) on production and expenditure across 56 industries from the 2014 World Input-Output Data (WIOD).
- The 56 WIOD industries are aggregated into 15 traded industries (for which Lashkaripour and Lugovskyy (2023) estimated trade 휎휎푘푘 and scale elasticities 휇휇푘푘), plus a services sector (assumed 휇휇푘푘 = 0 and 휎휎푘푘 = 11).
- Bilateral applied import tariffs (status quo) are constructed following Kucheryavyy, Lyn and Rodriguez-Clare (2023), based on UNCTAD TRAINS data (mainly from 2022).
- Status quo for export subsidies and industrial Pigouvian subsidies is assumed to be zero.

### Table III.1 — United Kingdom: Industry-level trade elasticities and external scale parameters (econometric estimation results used for model calibration; standard errors in parentheses)
- Agriculture and Mining: Trade Elasticity: 6.2 (2.3); External Scale Parameter: 0.1 (0.1)
- Food: Trade Elasticity: 2.3 (0.8); External Scale Parameter: 0.4 (0.1)
- Textiles, Leather & Footwear: Trade Elasticity: 3.4 (0.4); External Scale Parameter: 0.2 (0.02)
- Wood: Trade Elasticity: 3.9 (1.9); External Scale Parameter: 0.2 (0.1)
- Paper: Trade Elasticity: 2.6 (1.1); External Scale Parameter: 0.3 (0.1)
- Petroleum: Trade Elasticity: 0.6 (0.5); External Scale Parameter: 1.2 (0.9)
- Chemicals: Trade Elasticity: 4 (0.4); External Scale Parameter: 0.2 (0.02)
- Rubber & Plastic: Trade Elasticity: 5.2 (1.2); External Scale Parameter: 0.1 (0.03)
- Minerals: Trade Elasticity: 5.3 (1.7); External Scale Parameter: 0.2 (0.06)
- Metals: Trade Elasticity: 3 (0.5); External Scale Parameter: 0.2 (0.03)
- Machinery: Trade Elasticity: 7.8 (1.3); External Scale Parameter: 0.1 (0.02)
- Elec. & Optical Equip.: Trade Elasticity: 1.2 (0.3); External Scale Parameter: 0.6 (0.1)
- Transport Equip.: Trade Elasticity: 2.8 (0.9); External Scale Parameter: 0.1 (0.04)
- Recycling and others: Trade Elasticity: 6.2 (1); External Scale Parameter: 0.2 (0.02)
- Services sector (model assumption): Trade Elasticity: 11; External Scale Parameter: 0

*Source: Excerpt from "Industrial Policies: Handle with Care. Staff Discussion Note SDN/2025/002" (chapter section on Firms and Production, model calibration and implications).*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2025/english/1gbrea2025002-source-pdf.pdf_