## 1espea2023002

## Source details

**Canonical URL:** [1espea2023002](https://www.imf.org/-/media/files/publications/cr/2023/english/1espea2023002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2023/english/1espea2023002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2023/english/1espea2023002.pdf.json)

---

### Introduction — key observations
- Labor productivity performance in Spain:
  - Significantly lower than in some peer economies; poor convergence.
  - Large differences in output per hour worked across Spanish regions (IMF, 2018).
  - Weak performance across sectors, suggesting cross-cutting drivers.
- Post-Global Financial Crisis dynamics:
  - Counter-cyclical pattern: productivity growth 2009–2013 driven by large reduction in employment and rapid decline in working hours, increasing capital-labor ratio.
  - Capital deepening settled at a lower contribution (negative in 2014–19).
  - Pandemic measures (strengthened furlough scheme, ERTE) led to better labor market dynamics than in past crises.
- Total factor productivity (TFP):
  - Consistently low and lagging peers for decades.
  - Contributing factors identified: high share of temporary workers; wide use of sector-level collective bargaining agreements; capital misallocation linked to financial frictions; product market regulation; size-related regulations and policies; weak business innovation.
  - Reforms noted: December 2021 changes in regulation of temporary contracts (early results show a decline in the share of temporary workers); incentives for worker training; revamped vocational and training system; Spain’s Recovery, Transformation and Resilience Plan (RTRP) contemplates improvements to active labor market policies.
- Policy importance:
  - Sustained policy focus on raising productivity needed to increase living standards, rebuild fiscal buffers, and make growth more inclusive.

### Data and empirical strategy — scope and model features
- Dataset and coverage:
  - Firm-level data covering over 1.2 million Spanish firms during the period of 2003–2019.
  - Unbalanced panel with about 8 million total observations.
  - Selected 10 economic sectors covering about 80 percent of total value added and 70 percent of total employment.
- Firm investment empirical model:
  - Dependent variable: firm net investment, IINV_v_i_j = change in fixed assets normalized by one-year lagged fixed assets (interpretable as percent change of firm capital stock).
  - Key regressors: lagged debt-to-asset ratio, debt service ratio, log(FIRM_ASSET_VALUE), profit and sales growth indicators, firm size, firm age, local product market concentration, and sector-region-year fixed effects (kappa_s_r_i).
- Misallocation and TFP measures:
  - Misallocation measured by dispersion of firm-level marginal revenue product of capital (MRPK) and labor (MRPL), derived from profit-maximization framework (Hsieh and Klenow, 2009; Hsieh and Song, 2015; Gopinath et al., 2017).
  - Friction parameters τ_{i s i}^L and τ_{i s i}^K capture reduced-form measures of labor and capital market frictions; higher τ implies higher factor costs and more frictions.
  - Firm-level TFP estimated using De Loecker and Warzynski (2012) and Ackerberg, Caves, and Frazer (2015) methods, using value-added and Orbis material costs.
- Additional constructions:
  - Local market: NACE 2-digit sector × autonomous community pair; total of [209] local product markets.
  - Market concentration: Herfindahl–Hirschman index (HHI).
  - Firm size: by employees (0–9, 10–49, 50–249, 250+) and by book value of fixed assets (quartiles labeled micro, small, medium, large).
  - Firm age: years since establishment (firm considered one year old at establishment).

### Results — firm investment and misallocation highlights
- Firm-level investment:
  - Pre-GFC: firm net investment rate averaged about 20 to 30 percent (interpretable as annual percent change of firm capital stock).
  - Post-GFC: investment dropped to negative territory and only recently started to increase again.
  - Size-dependence:
    - Pre-GFC, a 10-percent increase in firm asset value could raise investment rate by about 0.6 percentage points (chart interpretation example).
    - By 2015, almost no difference between investment rates of large and small firms.
  - Macro-financial drivers:
    - Real interest rate, total deposits growth, and economic policy uncertainty were main drivers of size-differential changes.
    - Large firms disproportionately reduced investment when real interest rate increases, total deposit growth declines, and economic policy uncertainty rises.
  - Firm-level determinants:
    - Lower leverage ratios and lower debt service ratios associated with higher investment regardless of cycle.
    - Balance sheet stress relevance increases when financial conditions are tighter.
    - Tangible capital investment responds more strongly to financial condition changes.
    - Higher firm-level investment positively correlated with stronger future sales growth and return on assets (ROA).
- Resource misallocation:
  - Dispersion in MRPK and MRPL used to detect distortions; deviations from equalized marginal products indicate misallocation.
  - Significant positive correlations between model-based MRPK and observed indicators of financial constraints.
  - Firms with higher MRPK pay higher effective interest rates and on average hold a larger share of intangible assets.
  - Interpretation: capital and labor market frictions (tax and labor regulations, discretionary governance or credit provision, imperfect competition) are potential drivers of misallocation.

### Allocative efficiency, firm size, and firm age
- MRPK dispersion:
  - MRPK dispersion normalized to 2005 = 1 for cross-country comparison.
  - MRPK dispersion in Spain has stabilized in recent years, consistent with rising TFP; Spain remains more inefficient relative to the early 2000s compared to France and Germany.
  - Potential drivers of stabilization: “cleansing effect” of credit contractions (Osotimehin and Pappadà, 2017); accommodative monetary policy (Albrizio et al., 2021).
- Cost of capital and capital stock:
  - Micro, small, and medium-sized firms (by fixed asset value) pay a higher cost of capital compared to large firms, both before the euro area crisis and in recent years.
  - Finding consistent with collateral constraints: maximum external financing determined by value of firms’ existing capital stock.
  - Firm size by employment does not show robust correlation with cost of external financing.
  - Robustness: firms with less capital stock on average pay higher effective interest rates (proxy for average borrowing cost).
- Firm age and startup frictions:
  - Start-ups age < 2 years face significantly higher frictions in the capital market.
  - Spain’s business sector shows lower entry rates and lower post-entry survival rates compared to EU peers.
  - Gaps in entry rates and survival within first two years declined before 2015 but widened again in recent years.
  - Conditional on surviving after 2 years, probability of surviving 3 to 5 years is larger in Spain than in a median EU country.
  - Only firms 1 or 2 years old face significantly higher cost of capital relative to established firms with 10 or more years of experience.
  - Policy implication: focus support on startups to help them survive through the first two years.
- Size-dependent regulations and “small business trap”:
  - Spain has more than 100 size related regulations spanning accounting, financial, insurance, labor, and tax areas; these can create advantages for small firms but reduce incentives to grow.
  - Examples of size thresholds:
    - Not eligible for simplified accounting if company has either 1) assets of more than EUR 1 mn or 2) sales of more than EUR 2 mn.
    - Monthly VAT and income tax retention settlement if company has sales over EUR 6mn.
    - Not eligible for simplified balance sheets if company has either 1) assets of more than EUR 4mn or 2) sales of more than EUR 8mn.
  - Growth accounting: firms with less than 10 employees and with 10–49 employees pay less for labor, suggesting size-related labor market misallocations.
  - Prevalence of small and micro firms is high; SMEs on average have significantly lower TFP compared to large firms.
- Labor market frictions affecting young firms:
  - Young firms use larger share of temporary contracts (Dolado et al., 2011), implying higher employment instability and precautionary use of flexible contracts.
  - Young firms face higher labor search costs (Minkler, 1992) and rigidities may prevent scaling.
  - Growth accounting: startups (age 0–2 years) and young firms (age 3–5 years) exhibit a higher labor wedge, indicating greater labor market distortions.
  - Recent labor regulation modifications aim to promote permanent contracts over temporary ones and have strengthened short-time work schemes; implementation has increased the share of permanent contracts but it is too early for full assessment.
  - Suggested non-monetary supports: promoting employee ownership and facilitating startup spin-offs from science and universities.

### Innovation, human capital, and regional fragmentation
- Innovation and R&D:
  - Spain ranks relatively low in business innovation; business R&D spending as a share of GDP lags European and OECD peers and has not increased materially since 2008.
  - Law on Science, Technology and Innovation target: R&D public funding of 1.25 percent of GDP by 2030 (3 percent including private investment).
  - Recovery plan increases budgetary support for R&D and innovation; further review of incentives framework recommended.
  - Number of firms conducting R&D and number receiving R&D tax reliefs have both been declining in recent years.
- Education and skills:
  - Early leavers from education and training (age 18–24) declined by 13-percentage point over the last decade but remain among the highest in Europe.
  - Enrollment in vocational education and training (VET) remains low compared to the region.
  - Structural transformation will raise demand for communication & creation, and information skills.
  - Recent reforms to VET, incentives for training, and RTRP reforms to active labor market policies aim to promote learning, re-skilling, and personalized itineraries for low employability groups.
- Product market competition and regional fragmentation:
  - Market Unity Law (2013) intended to align central, regional, and local regulatory requirements; application slowed after 2017 Constitutional Court decisions.
  - Fragmentation of Spanish market has been increasing since the GFC; sub-national regulations significant contributors (Llano-Verduras et al., 2021).
  - Empirical finding: more competitive local product markets (lower local HHI) associated with both higher investment and higher firm-level TFP.
  - Literature: increases in market power of large firms can weaken investment, deter innovation, and reduce labor income shares (Gutiérrez and Philippon, 2017; IMF, 2019).

### Green investments and carbon pricing — labor market impacts and scenarios
- Policy context and targets:
  - Law on Climate Change and the Energy Transition (2021) establishes climate neutrality by 2050 at the latest; Spanish electricity system must be 100 percent renewable by 2050.
  - Intermediate target: reduce emissions by 23 percent relative to 1990 levels by 2030.
  - 2030 targets: renewables to account for 42 percent of the country’s energy mix, to generate 74 percent of its electricity (on a primary energy basis), and to reduce primary energy consumption by at least 39.5 percent by 2030.
  - Additional provisions: no later than 2040 new passenger cars and light commercial vehicles will be vehicles with emissions of 0gCO2/km; establishment of low emission zones in urban centers with population at least 50 thousand by 2023.
- RTRP green investment magnitude:
  - RTRP allocates about 40 percent of announced investments (€77.2 billion, 5.9 percent of 2022 GDP) for 2021–23 to green projects, prioritizing clean energy, sustainable mobility and building efficiency renovations.
- Carbon pricing instruments:
  - Most emissions subject to some carbon pricing (EU ETS in power and industry; fuel duties in road transport).
  - Explicit domestic carbon tax on fluorinated greenhouse gases in place since 2014, covering 3 percent of Spain’s total emissions.
  - Effective carbon rates are low relative to estimates of carbon emission damage and typically lower than in peer economies in sectors not covered by a common EU framework.
- Modeling approach:
  - Green investments modeled as a labor productivity shock because they increase physical capital and improve sectoral labor productivity.
  - Initial sector-specific labor productivity impacts computed using RTRP announced investments and stock of fixed assets by industry from Eurostat.
  - Analysis uses NACE Rev.2 sectoral disaggregation at the 1-digit level, covering 18 sectors.
- Carbon-pricing scenarios (Aguilar, González and Hurtado, 2022):
  - (i) an increase in the price of CO2 emissions from €25 to €100 per ton;
  - (ii) an expansion of the coverage of the ETS system, to fully cover all emissions from all firms, of all sectors;
  - (iii) a combination of (i) and (ii);
  - (iv) a combination of (i) and (ii) with the revenue from those measures used to finance a reduction in labor taxes.
  - Baseline analysis maps changes in sectoral GVA to changes in employment using results from scenario (iv). Alternative scenario uses scenario (iii).
- Aggregate employment impacts:
  - Baseline (scenario iv): combined green investments and carbon pricing would increase overall employment by about 4 percent.
  - Alternative (scenario iii without redistribution of proceeds via lower labor taxes): employment effect falls to 0.6 percent.
  - Baseline assumes CO2 price increases from €25 to €100 per ton and ETS expansion with revenue used to finance reduction in labor taxes.
- Sectoral heterogeneity:
  - Coke and Refined Petroleum would experience large declines in employment; other sectors see increases that offset declines in aggregate.
  - Carbon pricing would negatively impact employment in land transport and agriculture, but green investments in these sectors can more than offset negative effects.
- Occupational and skills impacts:
  - Mapping from sectoral employment to occupations uses 2019 Labor Force Survey (ISCO 1-digit and 2-digit).
  - More than half of overall employment increase concentrated in services and sales workers, professionals, and elementary occupations—together contributing more than 2 percentage points to the 4-percent increase.
  - Mapping to skills uses ESCO skills-occupation matrix:
    - More than 60 percent of growth in employment attributable to three sets of skills: Communication, collaboration and creativity; Information skills; and Assisting and caring—together contributing 2.4 percentage points to the 4-percent increase.
  - Important sub-skills: Monitoring, inspecting and testing (Information skills); Protecting and enforcing (Assisting and caring).
- Policy complements and design considerations:
  - Employment outcomes depend critically on design and use of carbon pricing proceeds; redistribution to lower labor taxes produces positive employment effects in baseline.
  - Strengthen active labor market policies (ALMPs) with a focus on training and reskilling.
  - Complement with education policies such as expansion of vocational training.
  - Sectoral climate policies, regulations and standards essential to address sector-specific obstacles to emissions reductions.
  - For shrinking sectors, strengthen ALMPs and requalification plans (Sectoral-RED Mechanism requires companies in sectors experiencing structural changes to develop requalification plan for affected workers).

### Input-output methodology and multipliers (Annex I summary)
- Input-output structure:
  - Economy modeled with I competitive sectors; Cobb-Douglas production function used.
  - Production: y_i = z_i S_i^{α_i} ∏_j F_{ij}^{L_{ij}} (equation (1) as presented).
  - Shares: α_i = (S S S S_i) / (p_i y_i); α_{ij} = (p_j F_{ij}) / (p_i y_i); α_i + ∑_j L_{ij} = 1.
- Propagation of shocks:
  - Log-linearized: log y_i = log z_i + α_i log S_i + ∑_j L_{ij} log F_{ij} (equation (3)).
  - Change in sectoral employment: Δlog S_i = Δlog y_i / α_i = (L_{ij} / α_i) Δlog z_j (equation (4)).
  - Input-output matrix A = [L_{ij}]; H ≡ (I − A)^{-1}; ln y = H ln z + c̄ (equation (5)).
  - Framework captures direct and indirect spillovers through input-output linkages.
- Multipliers:
  - Direct and indirect multipliers computed using 2018 input-output table for Spain; mapping from 1-digit to 2-digit assumes same percent change of sectoral capital stock applies to each subsector within a 1-digit sector.
  - Charts (12 two-digit sectors highlighted) report:
    - Green Investments: Output Multiplier by Sector — increase in gross output per Euro of additional output in a given industry (top sectors include 35, 34, 2, 26, 24, 31, 25, 1, 27, 32, 33, 3).
    - Green Investments: Employment Multiplier by Sector — increase in employment per 1 million additional output (top sectors include 35, 2, 31, 34, 25, 1, 26, 3, 24, 27, 32, 33).
  - Sources: IMF staff calculations.

### Policy implications and recommendations
- Labor market and firm policies:
  - Continue reforms to address structural labor market weaknesses (temporary contracts, training, active labor market policies).
  - Monitor labor reform (2022) effects: initial data suggests proportion of permanent hires increased significantly; evaluate Recovery plan education reforms after implementation.
  - Support startups to survive first two years: target financing costs and provide non-monetary supports (employee ownership, science/university spin-offs).
  - Address size-dependent regulatory barriers that may create a small business trap; review size-related regulatory thresholds across labor and tax spheres.
  - Strengthen ALMPs with focus on training and personalized itineraries for low-employability groups.
- Financial and product market reforms:
  - Address financial frictions and firm financing constraints to support investment, particularly in large/systemic firms during downturns.
  - Improve allocative efficiency by targeting sources of misallocation in factor and product markets (regulatory reform, competition policy, credit allocation).
  - Reduce regional regulatory fragmentation and improve application of Market Unity Law objectives to enhance product market competition.
- Innovation, R&D, and public investment:
  - Strengthen incentives and support for business R&D and broader innovation activities to close gaps with European and OECD peers.
  - Use public investment in large strategic projects (PERTEs) to crowd in private investment in tangible and intangible capital.
  - Review existing R&D incentives framework to increase take-up.
- Green transition design:
  - Ensure effective implementation and targeting of RTRP green investments.
  - Design carbon pricing carefully; use of revenues (e.g., to reduce labor taxes) materially affects employment outcomes.
  - Combine carbon pricing with ALMPs, sectoral climate policies, and standards to facilitate worker transitions from shrinking to expanding sectors.

### Selected numeric summary statistics (from Annex II)
- Net Investment Rate: Mean 0.32
- Leverage Ratio: Mean 0.30
- Log(Fixed Assets): Mean 11.5
- Debt Service Ratio: Mean 0.3
- ROA: Mean 0.0
- Age: Mean 13.5
- EMPL: Mean 9.6
- Market Share (percent): Mean 0.2
- Effective Interest Rate (percent): Mean 8.6
- TFP Growth (percent): Mean -0.31
- Sales Growth (percent): Mean 24.4
- HHI (0-1): Mean 0.05

*Source: Extract from "LABOR PRODUCTIVITY DYNAMICS IN SPAIN: A FIRM-LEVEL PERSPECTIVE" (December 16, 2022) — IMF staff analysis as contained in the supplied PDF content.*

### References ________________________________________________________________________________18

### 1espea2023002 - References

### Contents and Structure
- Main report sections included in the supplied extract:
  - LABOR PRODUCTIVITY DYNAMICS IN SPAIN: A FIRM-LEVEL PERSPECTIVE
  - LABOR MARKET IMPLICATIONS OF GREEN INVESTMENTS AND CARBON PRICING IN SPAIN (section headers listed)
  - ANNEXES I and II referenced
  - References section markers

### A. Introduction — key observations
- Labor productivity performance in Spain:
  - Significantly lower than in some peer economies; poor convergence.
  - Large differences in output per hour worked across Spanish regions (IMF, 2018).
  - Weak performance across sectors, suggesting cross-cutting drivers.
- Post-Global Financial Crisis (GFC) dynamics:
  - Counter-cyclical pattern: productivity growth 2009–2013 driven by large reduction in employment and rapid decline in working hours, increasing capital-labor ratio.
  - Capital deepening settled at a lower contribution (negative in 2014–19).
  - Pandemic measures (strengthened furlough scheme, ERTE) led to better labor market dynamics than in past crises.
- Total factor productivity (TFP):
  - Consistently low and lagging peers for decades.
  - Contributing factors identified: high share of temporary workers, wide use of sector-level collective bargaining agreements, capital misallocation linked to financial frictions, product market regulation, size-related regulations and policies, weak business innovation.
  - Reforms noted: December 2021 changes in regulation of temporary contracts (early results show a decline in the share of temporary workers); incentives for worker training; revamped vocational and training system; Spain’s Recovery, Transformation and Resilience Plan (RTRP) contemplates improvements to active labor market policies.
- Policy importance:
  - Sustained policy focus on raising productivity needed to increase living standards, rebuild fiscal buffers, and make growth more inclusive.

### B. Data and Empirical Strategy — data scope and key model features
- Dataset and coverage:
  - Firm-level data covering over 1.2 million Spanish firms during the period of 2003–2019.
  - Microdata source: Orbis Bureau Van Dijk (BvD) database, compiled by IMF’s Research Department (Díez et al., 2018).
  - Unbalanced panel with about 8 million total observations.
  - Selected 10 economic sectors that match census sectoral employment and value-added growth; these sectors cover about 80 percent of the total value added and 70 percent of total employment in the Spanish economy.
- Firm investment empirical model (specification summary):
  - Dependent variable: firm net investment, IINV_v_i_j = change in fixed assets normalized by one-year lagged fixed assets (interpretable as percent change of firm capital stock).
  - Key regressors include: lagged debt-to-asset ratio, debt service ratio, log(FIRM_ASSET_VALUE), profit and sales growth indicators, firm size, firm age, local product market concentration, and sector-region-year fixed effects.
  - Notation highlights: X_i_j includes firm size, age, local product market concentration; kappa_s_r_i denotes sector-region-year fixed effects.
- Misallocation and TFP measures:
  - Misallocation measured by dispersion of firm-level marginal revenue product of capital (MRPK) and labor (MRPL), derived from firm profit-maximization framework (Hsieh and Klenow, 2009; Hsieh and Song, 2015; Gopinath et al., 2017).
  - Friction parameters τ_{i s i}^L and τ_{i s i}^K capture reduced-form measures of labor and capital market frictions; higher τ implies higher factor costs and more frictions.
  - Firm-level TFP estimated using methods by De Loecker and Warzynski (2012) and Ackerberg, Caves, and Frazer (2015), using value-added on the left-hand-side and exploiting Orbis data on material costs.
- Additional variable constructions:
  - Local market defined for each NACE 2-digit sector × autonomous community pair; total of [209] local product markets.
  - Market concentration measured by Herfindahl–Hirschman index (HHI).
  - Firm size: two approaches
    - By employees: 0–9 (micro), 10–49 (small), 50–249 (medium-sized), 250+ (large).
    - By book value of fixed assets: quartiles labeled micro, small, medium, large.
  - Firm age: years since establishment (firm considered one year old at establishment).

### C. Results — firm investment and misallocation highlights
- Firm-Level Investment
  - Pre-GFC: firm net investment rate averaged about 20 to 30 percent (interpretable as annual percent change of firm capital stock).
  - Post-GFC: investment dropped to negative territory and only recently started to increase again.
  - Size-dependence:
    - Before the GFC, a 10-percent increase in firm asset value could raise investment rate by about 0.6 percentage points (chart note: coefficient interpretation example; elsewhere a 0.05 coefficient implies 0.5 percentage points for a 10-percent asset increase).
    - By 2015, almost no difference between investment rates of large and small firms.
  - Macro-financial drivers of size-differential changes:
    - Real interest rate, total deposits growth, and economic policy uncertainty were main drivers.
    - Large firms disproportionately reduced investment when real interest rate increases, total deposit growth declines, and economic policy uncertainty rises.
    - This is consistent with findings that large companies are more sensitive to economic cycles in advanced economies.
  - Firm-level determinants:
    - Healthier balance sheets (lower leverage ratios and debt service ratios) associated with higher investment regardless of cycle.
    - Balance sheet stress relevance increases when financial conditions are tighter.
    - Tangible capital investment responds more strongly to financial condition changes.
    - Higher firm-level investment positively correlated with stronger future sales growth and return on assets.
- Resource Misallocation
  - Focus: allocation of given capital and labor across heterogeneous producers; measure dispersion of MRPK and MRPL to detect distortions.
  - In frictionless markets, marginal products equalize across producers with same costs; deviations indicate misallocation.
  - Empirical findings:
    - Significant and positive correlations between model-based MRPK and observed indicators of financial constraints.
    - Firms with higher MRPK pay higher effective interest rates and on average hold a larger share of intangible assets.
  - Interpretation:
    - Dispersion in MRPK/MRPL and their correlations with firm characteristics point to capital and labor market frictions (tax and labor regulations, discretionary governance or credit provision, imperfect competition in capital, labor, and product markets) as potential drivers of misallocation.

### D. Policy implications and recommendations (as presented)
- Continue reforms to address structural labor market weaknesses (temporary contracts, training, active labor market policies).
- Address financial frictions and firm financing constraints to support investment, particularly in large/systemic firms during downturns.
- Improve allocative efficiency by targeting sources of misallocation in factor and product markets (regulatory reform, credit allocation, competition policy).

*Source: Extract from "LABOR PRODUCTIVITY DYNAMICS IN SPAIN: A FIRM-LEVEL PERSPECTIVE" (December 16, 2022) — IMF staff summary and analysis as contained in the supplied PDF content.*

### 12.      The within-sector a llocative efficiency in Spain’s capital market has stabilized in recent

### 12. The within-sector allocative efficiency in Spain’s capital market has stabilized in recent years, consistent with rising TFP

### Allocative efficiency and MRPK dispersion
- The dispersion of Marginal Revenue Product of Capital (MRPK) is normalized to 2005 = 1 for all countries in the comparison exercise.
- MRPK dispersion in Spain has stabilized in recent years, consistent with rising TFP; nevertheless, Spain remains more inefficient relative to the early 2000s compared to France and Germany.
- Potential drivers of recent stabilization:
  - “Cleansing effect” of credit contractions as bank lending declines (Osotimehin and Pappadà, 2017).
  - Positive impact of accommodative monetary policy on improving allocative efficiency (Albrizio et al., 2021).
- Elevated misallocation implies large variation in firms’ marginal productivity of capital; firms with higher MRPK often face larger financing costs and are more financially constrained.
- Empirical focus in the section: allocative efficiency across firm size and firm age to detect capital market distortions.
- Note: The exercise fixes sectoral share of gross value added at its initial level; Orbis BvD database is relatively representative for Spain but could be less representative for other countries.

### Cost of capital, firm size, and capital stock
- Relevant criterion for capital market distortions: level of capital stock measured as total value of fixed assets.
- Micro, small, and medium-sized firms defined by fixed asset value pay a higher cost of capital compared to large firms, both before the euro area crisis and in recent years.
- Finding consistent with collateral constraints: maximum external financing determined by value of firms’ existing capital stock.
- Firm size measured by employment does not show a robust correlation with cost of external financing.
- Robustness check: firms with less capital stock on average pay higher effective interest rates (proxy for average borrowing cost).

### Firm age and startup frictions
- Start-ups with an age of less than two years face significantly higher frictions in the capital market.
- Spain’s business sector shows lower entry rates and lower post-entry survival rates compared to EU peers.
- Gaps in entry rates and survival within the first two years declined before 2015 but widened again in recent years.
- Conditional on surviving after 2 years, probability of surviving 3 to 5 years is larger in Spain than in a median EU country.
- Only firms 1 or 2 years old face significantly higher cost of capital relative to established firms with 10 or more years of experience.
- Policy implication: focus support on startups to help them survive through the first two years.

### Size-dependent regulations and the “small business trap”
- Spain has more than 100 size related regulations spanning accounting, financial, insurance, labor, and tax areas; these can create advantages for small firms but reduce incentives to grow, creating inefficiencies.
- Examples of size thresholds and regulatory effects:
  - Firms with less than 10 employees and with 10–49 employees have less responsibility in forming workplace representation, implying lower labor costs from firms’ perspectives.
  - Accounting/tax thresholds linked to assets and sales:
    - Not eligible for simplified accounting if company has either 1) assets of more than EUR 1 mn or 2) sales of more than EUR 2 mn.
    - Monthly VAT and income tax retention settlement if company has sales over EUR 6mn.
    - Not eligible for simplified balance sheets if company has either 1) assets of more than EUR 4mn or 2) sales of more than EUR 8mn.
- Growth accounting findings: firms with less than 10 employees and with 10–49 employees pay less for labor, suggesting size-related labor market misallocations.
- Prevalence of small and micro firms in Spain is high compared to other European countries; SMEs on average have significantly lower TFP compared to large firms.

### Labor market frictions affecting young firms
- Young firms use a larger share of temporary contracts (Dolado et al., 2011), implying higher employment instability and precautionary use of flexible contracts.
- Young firms face higher labor search costs (Minkler, 1992) and rigidities may prevent scaling to desirable production.
- Growth accounting shows startups (age 0–2 years) and young firms (age 3–5 years) exhibit a higher labor wedge, indicating greater labor market distortions.
- Recent labor regulation modifications aim to promote permanent contracts over temporary ones and have strengthened short-time work schemes; implementation has increased the share of permanent contracts but it is too early for full assessment.
- Suggested non-monetary supports for startups: promoting employee ownership and facilitating startup spin-offs from science and universities (Comprehensive startup strategy, Germany).

### Product market competition and regional regulatory fragmentation
- Market Unity Law approved in 2013 intended to align central, regional, and local regulatory requirements.
- Application slowed after null and void declarations of key principles by the Spanish Constitutional Court in 2017; regulatory differences across autonomous communities and municipalities have not improved materially.
- Fragmentation of Spanish market has been increasing since the GFC, with sub-national regulations as significant contributors (Llano-Verduras et al., 2021).
- Empirical finding: more competitive local product markets (lower local HHI) are associated with both higher investment and higher firm-level TFP.
- Literature link: increases in market power of large firms can weaken investment, deter innovation, and reduce labor income shares (Gutiérrez and Philippon, 2017; IMF, 2019).

### Innovation, R&D, and human capital
- Spain ranks relatively low in business innovation; business R&D spending as a share of GDP lags European and OECD peers and has not increased materially since 2008.
- Despite relatively high implied tax subsidy rate for R&D activities, take-up of R&D tax incentives has been limited historically.
- Overall government support of business R&D (direct and indirect) falls short compared to the region’s frontier (OECD, 2021).
- Number of firms conducting R&D activities and number receiving R&D tax reliefs have both been declining in recent years (Xifré, 2018; OECD, 2021).
- Spanish businesses also lag in other innovative activities: business and public-private collaborations, process innovations, and non-R&D innovative expenditures.
- Education and skills:
  - Early leavers from education and training (age 18–24) declined by 13-percentage point over the last decade but remain among the highest in Europe.
  - Enrollment in vocational education and training (VET) remains low compared to the region.
  - Structural transformation (digitalization and greening) will raise demand for communication & creation, and information skills.
  - Recent reforms to VET and incentives for training, and reforms to active labor market policies in the RTRP, aim to promote learning and re-skilling and personalized itineraries for low employability groups.

### Policy implications and recommended focus areas
- Support startups to survive the first two years (targeted measures addressing financing costs and non-monetary supports such as employee ownership and science/university spin-offs).
- Address size-dependent regulatory barriers that may create a small business trap and discourage scaling.
- Enhance product market competition by reducing regional regulatory fragmentation and improving application of Market Unity Law objectives.
- Strengthen incentives and support for business R&D and broader innovation activities to close gaps with European and OECD peers.
- Improve education and training outcomes, increase VET enrollment, and scale re-skilling policies to match changing skill demands from digitalization and greening.
- Monitor labor market reforms (permanent contract promotion, short-time work schemes) to assess impact on workforce allocation, especially for young firms.

*Source: IMF staff analysis (chapter content).*

### 20.      Spain has committed to

### 20.      Spain has committed to 

### Human Capital
- Labor reform enacted in 2022 aimed to combat the high prevalence of temporary employment to improve accumulation of workers’ human capital; initial data suggests that the proportion of permanent hires has increased significantly.
- Recovery plan education reforms include:
  - investments in digital skills,
  - expansion of vocational training,
  - modernization of the education system,
  - revamp active labor market policies (important to limit the loss of human capital during unemployment).
- The plan’s programs need evaluation after being in place for some time.

### Innovation and R&D
- Law on Science, Technology and Innovation sets a goal for R&D public funding of 1.25 percent of GDP by 2030 (3 percent including private investment).
- Recovery plan increases budgetary support for R&D and innovation but should be complemented with a review of the existing incentives framework.
- Further efforts needed to:
  - strengthen collaborations between the public and private sectors,
  - identify impediments to business innovation.
- Adoption of large strategic projects (PERTEs) is an opportunity to foster synergies between public and private investment and promote innovation.
- Recovery plan target noted elsewhere: target 2 percent of GDP in R&D investment (listed among area key initiatives).

### Business demographics and dynamics
- Recently approved Law on Business Growth includes measures to eliminate financial and administrative barriers to firm growth and to reduce delinquency rates.
- A new Start-up Law is being legislated to provide incentives and facilitate the creation of innovative companies.
- Sectoral conference for Regulatory Improvement and the Business Climate (established in 2021) aims to enhance cooperation across government levels to foster business activity.
- Recommended further efforts:
  - review size-dependent regulatory thresholds in the labor and tax spheres,
  - better align regulatory framework across regional and local authorities to promote market integration and business competition.

### Key empirical findings on productivity and firm dynamics
- Low labor productivity growth in recent years was mostly driven by a lack of capital deepening because of weak firm investment amid a continued deleveraging process since the GFC.
- Large firms lowered investment disproportionately more in response to the unfavorable macroeconomic shock.
- Firm balance sheet health and growth potential are significant factors affecting investment decisions.
- Allocative inefficiency remains relatively high in Spain compared to neighboring countries.
- Misallocation across firms’ size and age groups is identified in both the capital market and the labor market, suggesting size-dependent financial constraints remain a significant determinant of TFP.
- Startups with less than 2 years of experience are the most vulnerable age group, with a much lower survival rate compared to other European countries and significant disadvantages in capital and labor markets versus older firms.
- High market concentration is associated with both weak investment and weak TFP growth.

### Policy conclusions and recommendations
- Safeguarding firm balance sheets and supporting large and systemic companies during downturns can help limit the impact of negative macroeconomic shocks on investment.
  - Pandemic firm support measures (state-backed guarantee program and solvency support to strategic companies) successfully preserved the environment for resuming corporate investment.
- Reducing economic policy uncertainty and reforms in product market regulation would help boost investment.
- Public investment in large strategic projects (PERTEs), and effective implementation of structural reforms, could crowd in private investment in tangible as well as intangible capital.
- To improve allocative efficiency and enhance firm TFP growth, continued policy effort should:
  - alleviate distortions faced by startups and SMEs, including improving access to finance for disadvantaged firms (e.g., startups with less than two years of experience or firms with limited physical capital to use as collaterals),
  - review the large number of size-related rules and regulations,
  - expand non-monetary support to startups (including facilitating the search for talent, providing easier access to data, and reducing regulatory costs),
  - continue aligning the regional regulatory framework and improving product market competition.

### Green policies and labor market impacts
- Spain’s Law on Climate Change and the Energy Transition (2021) establishes:
  - climate neutrality by 2050 at the latest,
  - Spanish electricity system must be 100 percent renewable by 2050,
  - intermediate target to reduce emissions by 23 percent relative to 1990 levels by 2030,
  - targets for 2030: renewables to account for 42 percent of the country’s energy mix, to generate 74 percent of its electricity (on a primary energy basis), and to reduce primary energy consumption by at least 39.5 percent by 2030.
- Additional law provisions:
  - no later than 2040 new passenger cars and light commercial vehicles will be vehicles with emissions of 0gCO2/km,
  - establishment of low emission zones in urban centers with a population of at least 50 thousand by 2023,
  - creation of an Expert Committee to submit an annual report to Congress with policy evaluation and recommendations.
- National Integrated Energy and Climate Plan 2021–2030 (NECP) envisages sectoral actions including: increasing renewable power installations, boosting renewable gases in power sector, modal shifts and electrification in transport, refurbishments and renewable heating in residential and commercial sectors, energy efficiency and fuel switching in industry, and energy efficiency in agriculture.
- Green policy instruments analyzed:
  - Carbon pricing: most emissions are subject to some form of carbon pricing (EU ETS in power and industry; fuel duties in road transport). Explicit domestic carbon tax on fluorinated greenhouse gases has been in place since 2014, covering 3 percent of Spain’s total emissions. Effective carbon rates are low relative to estimates of carbon emission damage and typically lower than in peer economies in sectors not covered by a common EU framework.
  - Green investments: Spain’s Recovery, Transformation and Resilience Plan (RTRP) allocates about 40 percent of announced investments (€77.2 billion, 5.9 percent of 2022 GDP) for the period 2021–23 to green projects, prioritizing clean energy, sustainable mobility and building efficiency renovations.
- Modeling approach: Green investments are modeled as a labor productivity shock because they increase the level of physical capital and improve sectoral labor productivity. Initial sector-specific labor productivity impacts are computed using information on the RTRP announced investments and on the stock of fixed assets by industry from Eurostat. Analysis uses NACE Rev.2 sectoral disaggregation at the 1-digit level, covering 18 sectors.
- Main labor market result: The labor market impact of the two policies studied (the announced green investments in the recovery plan and an increase in carbon pricing with an expansion of the EU ETS) is net positive, although results depend on policy design, particularly on the use of proceeds from increased carbon pricing.
- Policy complements to facilitate the transition:
  - Strengthen active labor market policies with a focus on training,
  - Complement with education policies such as expansion of vocational training to facilitate worker transitions from shrinking to expanding sectors.

### Selected numeric summary statistics (from Annex II)
- Net Investment Rate: Mean 0.32
- Leverage Ratio: Mean 0.30
- Log(Fixed Assets): Mean 11.5
- Debt Service Ratio: Mean 0.3
- ROA: Mean 0.0
- Age: Mean 13.5
- EMPL: Mean 9.6
- Market Share (percent): Mean 0.2
- Effective Interest Rate (percent): Mean 8.6
- TFP Growth (percent): Mean -0.31
- Sales Growth (percent): Mean 24.4
- HHI (0-1): Mean 0.05

*Source: https://www.imf.org/-/media/files/publications/cr/2023/english/1espea2023002.pdf*

### 5.      We use input-output linkages across sectors to estimate the impact of green

### 1espea2023002 - 5.      We use input-output linkages across sectors to estimate the impact of green

### Methodology: input-output linkages and multipliers
- Direct and indirect multipliers for each sector computed using the 2018 (latest available) input-output table for Spain.
- NACE 2-digit sectors used to analyze propagation of green investments along the input-output network.
- Rationale: higher labor productivity in a sector implies lower relative prices and lower production costs for downstream sectors; well-connected upstream sectors (mining, transportation, support services) expected to see large indirect impacts on aggregate output.
- Mapping assumption: "In order to map the investment shock at the 1-digit sector to the 2-digit level input-output table, we assume the same investment shock (in percent change of sectoral capital stock) to each subsector with in the same 1-digit sector."
- See Annex I for calculation details of output and employment multipliers.

### Green investment projects and magnitudes (as presented)
- ProjectSector EUR million 2021–23 % of 2018 Capital Stock Sustainable, safe and connect mobilityTransport132036.39
- Building renovation and urban renewalConstruction68201.93
- Deployment and integration of renewable energyEletricity, gas, power3165
- Roadmap for renewable hydrogenEletricity, gas, power1555
- Energy infrastructures, smart networks, storageEletricity, gas, power1365
- Conservation and restoration of ecosystems and biodiversityAgriculture1642
- Preservation of the coastline and water resourcesWater and waste management20913.64
- Sources: OECD and IMF staff calculations.
- Main Green Investments Under the Recovery Plan: 6.62

### Carbon pricing: scenarios and mapping to employment
- Sectoral GVA changes taken from Aguilar, González and Hurtado (2022), which incorporate propagation through input-output linkages.
- Four carbon-pricing scenarios computed by Aguilar, González and Hurtado (2022):
  - (i) an increase in the price of CO2 emissions from €25 to €100 per ton;
  - (ii) an expansion of the coverage of the ETS system, to fully cover all emissions from all firms, of all sectors;
  - (iii) a combination of (i) and (ii);
  - (iv) a combination of (i) and (ii) with the revenue from those measures used to finance a reduction in labor taxes.
- Baseline analysis maps changes in sectoral GVA to changes in employment using results from scenario (iv).
- Alternative scenario produced based on scenario (iii); detailed results reported in Annex II.
- The difference between scenarios highlights the importance of using carbon revenues to reduce distortionary taxes and support growth.

### Aggregate employment impacts
- Baseline (scenario iv): combined green investments and carbon pricing would increase overall employment by about 4 percent.
- Alternative (scenario iii without redistribution of proceeds via lower labor taxes): employment effect falls to 0.6 percent.
- Note: baseline scenario assumes increase in CO2 price from €25 to €100 per ton and expansion of ETS, with revenue used to finance a reduction in labor taxes; alternative is same but without redistribution via lower labor taxes.
- Graphical decomposition in text separates contributions from: Green investments and Carbon pricing (baseline vs alternative).

### Sectoral heterogeneity of employment effects
- Impact of green investments and carbon pricing is heterogeneous across sectors.
- Coke and Refined Petroleum would experience large declines in employment; other sectors see significant increases that offset these declines in aggregate.
- In the baseline scenario, positive effects on sectoral employment are mainly due to the carbon pricing shock—specifically, positive effects of labor tax reductions enabled by additional fiscal revenues.
- Examples:
  - Carbon pricing would negatively impact employment in land transport and agriculture, but green investments in these sectors can more than offset those negative employment effects.
- The chapter lists sectoral employment change magnitudes in thousands of persons (graphical): wide variation across sectors, with some sectors showing large positive and negative net changes.

### From sectoral employment changes to occupations
- Mapping approach: use distribution of occupations by sectors from the 2019 Labor Force Survey for Spain; occupations disaggregated by ISCO 1-digit and 2-digit.
- More than half of the overall increase in employment concentrated in three occupations:
  - Services and sales workers, professionals, and elementary occupations together contribute more than 2 percentage points to the 4-percent increase in employment due to the greening shocks.
- Drivers of occupational changes:
  - Growth in services and sales workers and professionals largely driven by the carbon pricing shock (when revenues reduce labor taxes).
  - For elementary occupations, both green investments and carbon pricing contribute roughly equally.
  - Investment shock more prominent among skilled workers in primary occupations and among plant and machine operators and assemblers, but these groups have small shares in total employment, limiting their contribution to aggregate employment growth.
- Within-occupation heterogeneity:
  - Services and sales workers: main sub-categories contributing are personal service and sales workers; personal care and protective services contribute less.
  - Professionals: contributions more evenly distributed, largest from teaching professionals.
  - Elementary occupations: larger contributions by cleaners and helpers and by laborers in mining, construction, manufacturing and transport.
  - Notable sub-categories with large contributions: building and related trades workers (craft and related trades); drivers and mobile plant operators (plant and machine operators and assemblers); skilled agricultural workers (skilled workers in primary activities).

### From occupations to skills
- Mapping approach: use distribution of skills for each 1-digit ISCO occupation from ESCO skills-occupation matrix tables; skills disaggregated by ESCO 1-digit and 2-digit.
- Skill-set concentration:
  - More than 60 percent of growth in employment can be attributed to three sets of skills (with heterogeneity across sub-categories).
  - Communication, collaboration and creativity; Information skills; and Assisting and caring together contribute 2.4 percentage points to the 4-percent increase in employment.
- Important sub-skill contributors:
  - Monitoring, inspecting and testing (within Information skills).
  - Protecting and enforcing (within Assisting and caring).
  - These sub-skills include activities associated with the green economy, e.g., testing vehicles, monitoring environmental conditions, complying with environmental protection laws and standards.
- Distribution of skills in 2019 (percent of total employment) reported in figures and used for mappings.

### Conclusions and policy recommendations
- Framework contribution:
  - Provides a tractable framework accounting for input-output linkages to assess labor market impact of green investments and policies.
  - Applied to: (i) announced green investments in the recovery plan; and (ii) a possible increase in carbon pricing and expansion of ETS, based on scenarios from Aguilar, González and Hurtado (2022).
- Green investments:
  - Productive green investments can support employment growth and lead to labor reallocation during the green transition.
  - Investments in the recovery plan correctly target sectors needing upscaling, but results may change with more granular information on investments or occupation distributions across sectors.
  - Ensuring effective implementation and continued proper identification of future investment projects is critical for positive effects to materialize.
- Shrinking polluting sectors and active labor market policies (ALMPs):
  - Even in optimistic scenarios, most polluting sectors will suffer employment losses and their workers will likely need to transition to expanding sectors.
  - Coke and Refined Petroleum is the main sector identified as potentially suffering employment losses from the combined greening shocks because the negative impact from carbon pricing is not offset by investment projects.
  - For shrinking sectors, strengthening ALMPs with focus on training to facilitate reallocation is crucial.
  - Spain has introduced reforms to promote reskilling and requalification, including training incentives to short-time work schemes (ERTEs and the new RED mechanism).
  - Sectoral-RED Mechanism requires companies in sectors experiencing structural changes to develop a requalification plan for affected workers.
  - Spain’s RTRP contemplates reforms to strengthen ALMPs to improve labor matching and address skill mismatches; proposed legal framework foresees creation of personalized itineraries to support training and re-skilling of low-employability groups, focusing on skills demanded by companies.
- Design of carbon pricing:
  - Employment outcomes depend critically on the design and use of carbon pricing proceeds.
  - Baseline results rely on proceeds being redistributed as a uniform reduction of labor taxes, which leads to positive employment effects across-the-board without creating incentives for labor reallocation.
  - Alternative uses of proceeds would lead to different employment outcomes.
- Complementary greening policies:
  - Sectoral climate policies, regulations and standards will be essential to address sector-specific obstacles to reducing emissions, particularly in transport, building and power sectors.
  - For more detailed discussion on complementary policies, see Spain: Selected Issues (2022).

*Source: IMF staff calculations.*

### Annex I. Estimation of Output and Employment Multipliers Using

### Annex I. Estimation of Output and Employment Multipliers Using the Input-Output Approach

### The Input-Output Structure of the Economy
- Economy modeled with 퐼퐼 competitive sectors {1,2 , ...,퐼퐼}; each sector produces a distinct product using intermediate goods and labor.
- Production function: Cobb-Douglas. Output of industry 푖:
  - 푦푦푖 = 푧푧푖 푆푖 훼푖 ∏푗 퐹푖푗 푎푖푗  (equation (1) as presented)
- Definitions:
  - 푆푖: amount of labor hired by firms in sector 푖.
  - 퐹푖푗: quantity of good 푗 used for production of good 푖 as intermediate inputs.
  - 푧푧푖: labor productivity in sector 푖.
  - 훼푖 and 퐿푖푗 are labor and intermediate input shares.
- Shares characterized (assuming 푝푖 is price of good 푖 and 푆 is equilibrium wage):
  - 훼푖 = (푆푆푆푆푖) / (푝푖 푦푖)
  - 훼푖푗 = (푝푗 퐹푖푗) / (푝푖 푦푖)  (equations (2))
  - 훼푖 + ∑푗 퐿푖푗 = 1

### The Propagation of Supply/Productivity Shocks
- Log-linearized production function:
  - log푦푖 = log푧푖 + 훼푖 log푆푖 + ∑푗 퐿푖푗 log퐹푖푗  (equation (3))
- Direct impact of a supply shock to labor productivity 푧푖 described by (3).
- Example: if industry 푗 is hit by a negative shock to 푧푗 reducing production and increasing price of good 푗, industries using good 푗 as intermediate input experience direct impact via 퐿푖푗 (input-output coefficient).
- Change in sectoral employment:
  - Δlog푆푖 = Δlog푦푖 / 훼푖 = (퐿푖푗 / 훼푖) Δlog푧푗  (equation (4))
- Input-output matrix:
  - 퐴 = [퐿푖푗]
- Log-price response (partial equilibrium, wages in other sectors do not respond):
  - 푝̂ = −(I − A)^{-1} log푧푧
  - where relative price vector 푝̂ = (log(푝1/푤), log(푝2/푤), ..., log(푝푛/푤))′
- Combining households’ Cobb-Douglas utility and defining 퐻 ≡ (퐼 − 퐴)^{-1}, log output:
  - ln푦 = 퐻 ln푧 + c̄  (equation (5) in text; expressed as ln푦푖 = ∑ℎ푖푗 ln푧푗 + c̄)
- Interpretation of equation (5):
  - Captures direct and indirect spillovers: initial impact (equation (4)) propagates downstream through input-output linkages, with price increases in affected industries creating further indirect negative effects on their customers.
- Modeling assumptions that prevent upstream propagation:
  - (a) Cobb-Douglas preferences and technologies,
  - (b) a single factor of production (labor),
  - (c) constant returns to scale.
  - (b) and (c) together ensure productivity shocks only affect upstream sectors through wage changes; prices relative to the wage are not affected.

### Supply Shock Multipliers
- The report summarizes direct and indirect multipliers of sectoral supply shocks (green investment shock) with charts showing:
  - Green Investments: Output Multiplier by Sector — "Increase in gross output--measured in Euros--per Euro of additional output in a given industry".
    - The chart highlights the 12 two-digit sectors with the largest multipliers, including sectors numbered: 35, 34, 2, 26, 24, 31, 25, 1, 27, 32, 33, 3 (sector names shown in figure captions).
  - Green Investments: Employment Multiplier by Sector — "Increase in employment--measured in persons--per 1 million additional output in each industry".
    - The chart highlights the 12 two-digit sectors with the largest multipliers, including sectors numbered: 35, 2, 31, 34, 25, 1, 26, 3, 24, 27, 32, 33 (sector names shown in figure captions).
- Sources and notes:
  - Sources: IMF staff calculations.
  - Note: Charts show the 12 two-digit sectors with the largest multipliers.

*Annex I. Estimation of Output and Employment Multipliers Using the Input-Output Approach — source: IMF staff calculations as presented in the provided content.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2023/english/1espea2023002.pdf_