## wpiea2020038-print-pdf

## Source details

**Canonical URL:** [wpiea2020038-print-pdf](https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020038-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2020/english/wpiea2020038-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2020/english/wpiea2020038-print-pdf.pdf.json)

---

### Introduction — key findings and interpretation
- Italian real wages have grown faster than labor productivity since joining the euro; post-global financial crisis (GFC) both real wage growth and labor productivity stalled, leaving elevated unit labor costs relative to euro area peers.
- National accounts indicate a growing share of value added going to employee compensation (unit labor cost), mechanically eroding corporate profits and capital returns and coinciding with a decline in corporate investment.
- Main empirical estimate: a 1 percent increase in real wages causes a ⅓ percent fall in fixed capital over one year.
- Distributional response: profits absorb only ½ of the associated cost increase; the remaining share is passed through to capital formation reductions.
- Mechanism: given short-term rigidity of labor quantities in Italy, liquidity constraints are identified as the main channel linking higher wages to lower investment; evidence: investment–wage elasticity is relatively more negative during the crisis years (2009–2012).
- Heterogeneity: firms in the South and medium firms display lower elasticities (likely due to lower effective coverage of national wage contracts). Results are robust to specification changes and to excluding leading firms in a sector.

### Aggregate facts and institutional background
- Decomposition of non-financial corporate (NFC) value added shows rising employee compensation share since Italy joined the euro in 2002; rising depreciation also compressed net returns to capital (interest + dividends + profits).
- Comparison with euro area: capital returns and investment fell across the euro area after the GFC, but the decline was much more pronounced in Italy, coinciding with Italy-specific growth of labor costs as a share of value added.
- Wage-bargaining institutional facts:
  - More than 97 percent of Italian dependent workers are subject to national-level wage contracts (CCNLs).
  - CCNL minima bind for about 90 percent of workers.
  - CCNLs typically last 3 years.
  - 80 percent of workers remain on permanent contracts with relatively high firing costs.
- Implication: downwardly rigid nominal wages combined with low inflation contributed to a real exchange rate overvaluation identified in IMF assessments.

### Theory and empirical model
- Theoretical channels from wages to investment:
  - Negative: factor complementarity (wage hike reduces labor, lowers marginal product of capital); financial constraints/liquidity (wage hike reduces cash flow, limiting investment).
  - Positive: efficiency wages or substitution of labor for capital could imply positive wage–investment relationship (not supported by main findings).
- Empirical identification:
  - Uses negotiated sectoral wage growth (national IStat CCNL-based series) as exogenous to individual firms.
  - Interaction of sectoral wage growth with firm-level lagged labor share provides an exogenous proxy for firm-level labor cost shocks.
- Empirical specification (panel):
  - Dependent variable: ∆k_{i,t} / k_{i,t−1} (real capital growth) or cash flow / lagged capital.
  - Key regressor: α_{i,t−1} * (∆w_{j,t} / w_{j,t−1}), with firm and sector-year fixed effects and controls (lagged dependent variable, labor share, interaction of capital share and interest rate growth, lag capital growth, lag debt/total assets, low ICR dummy, firm size and age categories).
  - Regression estimated using Correia (2017) procedure (Stata “reghdfe”).

### Micro data and sample statistics
- Firm-level data: Orbis panel 2003–2017; cleaned sample for 2005–2017: around 200,000 firms per year; covers 2/3 of NFC value added.
- Sectoral wage growth: IStat negotiated national wage contract growth by NACE4 sectors for 2005–2017; 221 sectors with more than 20 valid observations in at least two consecutive years.
- Contractual real wage growth across sector-years ranged between -1 and 3 percent (2005–2017).
- Selected descriptive statistics (all variables deflated; monetary quantities are 2010 euro thousands):
  - Fixed Assets (EUR thousand): Mean 10,300,000; Std. Deviation 435,000,000
  - Cash flow (EUR thousand): Mean 1,350,024; Std. Deviation 45,700,000
  - Costs of employees (EUR thousand): Mean 2,250,687; Std. Deviation 27,900,000
  - Value Added (EUR thousand): Mean 4,260,339; Std. Deviation 94,000,000
  - Fixed Asset growth: Mean -0.00860; Std. Deviation 0.3174
  - Cash flow / Lag Fixed Assets: Mean 0.36; Std. Deviation 0.39
  - Negotiated wage growth: Mean 0.0061; Std. Deviation 0.0098
  - Lag labor share: Mean 0.67; Std. Deviation 0.18
  - Interest rate growth: Mean -0.0022; Std. Deviation 0.0070
  - Lag Debt / Total Assets: Mean 0.18; Std. Deviation 0.19
  - Lag Low ICR (dummy): Mean 0.47; Std. Deviation 0.50
  - Number of Employees: Mean 43.2; Std. Deviation 533
  - age (years): Mean 20.8; Std. Deviation 14.0
  - Sales share in sector: Mean 0.0015; Std. Deviation 0.0105
  - Lag labor share (alternative): Mean 0.65; Std. Deviation 0.19

### Empirical results — sector and firm levels
- Sector-level correlations (sector-year observations):
  - wage_gr on lab_productivity_gr: beta 0.004 (N 2,745; R2 0.000) — wage growth unrelated to productivity growth.
  - fixed_asset_gr on wage_gr: beta -0.744*** (N 2,745; R2 0.362) — sectors with higher wage growth tend to have lower fixed asset growth. (Standard errors clustered at sector level.)
- Firm-level headline result:
  - Multiplying estimated β̂2 by average labor share ᾱ = 0.67 yields: a one percentage point increase in wage growth causes a 0.33 percent (⅓ percent) reduction in fixed capital over one year.
- Summary headline magnitudes reported:
  - 0.36 percentage point fall in fixed asset growth and a 0.78 percentage point fall in cash flow.
- Table 3 (selected coefficients reported exactly as in source):
  - dependent variable: fixed_assets_gr | cashflow_k
  - L.dep var: -0.129*** | 0.166***
  - L.labor_share: 0.096*** | 0.091***
  - l_lab_sh*wage_gr: -0.541** | -1.157***
  - l_cap_sh*loan_rate_gr: -0.229 | 2.501***
  - L.debt/assets: -0.122*** | -0.121***
  - L.lowICR: -0.053*** | -0.028***
  - L.log employees: -0.019*** | -0.023***
  - age bin=2: -0.016*** | 0.013***
  - age bin=3: -0.001 | 0.018***
  - age bin=4: 0.008 | 0.020***
  - age bin=5: 0.014** | 0.022***
  - age bin=6: 0.009 | 0.021***
  - age bin=7: 0.005 | 0.015**
  - r2_a_within: 2.3% | 3.3%
  - N: 888,653 | 844,484
- Estimation details:
  - Includes firm and sector-year fixed effects.
  - Sample excludes one-employee firms, insolvent firms, SOEs, non-market sectors, and is winsorized.
  - Variables deflated; standard errors clustered at the firm level.
  - Asterisks indicate statistical significance at the 10, 5, and 1 percent level respectively.

### Theoretical channels and interpretation
- Employment rigidity evidence:
  - Labor quantities appear very rigid with respect to wages at the one-year horizon; employment growth regressions yield insignificant coefficients.
  - High firing costs and the Cassa Integrazione Guadagni subsidy likely contribute to employment rigidity.
- Channel attribution:
  - With employment relatively rigid, factor complementarity cannot be the main channel.
  - Large impact of wages on cash flows suggests liquidity constraints may play a role.
  - The coefficient of interest is significantly more negative during the global financial and European confidence crises (2009–2012), reinforcing the liquidity-constraint hypothesis.
  - Partial anticipation of future real wage growth from triennial nominal wage contracts can interact with liquidity constraints to hinder optimal adjustment of capital and labor.

### Heterogeneity, robustness, and elasticity range
- Period, region, and firm-size heterogeneity (selected Table 4 values; omitted groups are normal years, firms in the North, and large firms):
  - omitted group: -0.089 | -0.649*** | -0.990*** (standard errors reported in source)
  - crisis (2009–2012): -1.044*** (0.399)
  - Center: 0.179 (0.160)
  - South: 0.593*** (0.168)
  - small (2–14 empl.): 0.276 (0.251)
  - medium (15–199 empl.): 0.647*** (0.247)
  - r2_a_within: 2.3%
  - N: 888,653
- Robustness checks (summarized):
  - Baseline (Table 3, after multiplying by average labor share).
  - Weighted by average real fixed assets of a firm.
  - Regression without controls (γ⃗ = 0).
  - Simple regression plus interaction between interest rates and capital share.
  - Main regression without lagged dependent variable.
  - Lags taken as average of previous two years (c−1 and c−2); two-year lag (c−2) gives very similar results.
  - Exclude firms with share of sales above 5 percent in a year-sector.
  - Alternative labor share definition using total factor payments and cost of capital r_k = lending rate + equity premium approximated to 7 percent.
- Estimated elasticities across robustness cases:
  - Always negative across robustness cases, mostly ranging from -¼ to -¾.
  - Statistical significance lost in some models due to large number of fixed-effect parameters (more than 200,000) absorbing variation.
- Heterogeneity findings:
  - Estimated investment-wage elasticities significantly weaker for firms in the South and slightly weaker for medium firms.
  - Possible explanations: higher prevalence of temporary contracts, CCNLs covering a smaller share of workers, or balance sheet variables being less precisely measured in those groups.

### Aggregate implications and quantitative illustrations
- External rebalancing and investment:
  - IMF External Balance Assessment (IMF, 2019b) concluded Italy’s unit labor cost real exchange rate is overvalued by 0 to 10 percent.
  - Illustrative extrapolation: a 6 percent wage devaluation would increase investment over fixed assets by 2.2 percentage points, bringing it back to its pre-GFC average level (from the 2017 level).
  - Complementary model-based evidence: Andrle et al. (2018) predicts a 15 percent reduction in wage markups would boost steady-state investment by around 3 percent.
  - Higher investment would translate into higher labor productivity, enabling expansion of employment and higher steady-state wages.
  - Decentralization of wage bargaining and product and service market reforms could reduce need for initial wage adjustment and yield better allocation of labor.
- Minimum wage illustrative calculation:
  - INPS (2019) estimates a minimum wage of €9 per hour would increase labor costs by €9.7 billion, or 2.2 percent of total labor costs.
  - Extrapolation of regression results suggests such minimum wage would reduce the fixed capital stock by 0.8 percent.
  - Caveats:
    - Prediction based on average worker; responses may vary across wage distribution.
    - Low-skilled workers may be more substitutable with capital, potentially mitigating or overturning predicted investment impact.
    - INPS estimate considers only workers currently below the proposed minimum; upward pressure on wages above the minimum could imply a larger investment decline.

### Policy takeaways and recommendations
- Correcting wage overvaluation should boost investment; introducing a minimum wage at an excessive level would harm investment.
- Need for a wage bargaining system that takes into account investment impacts—specifically, realigning real wages with labor productivity, for instance through decentralized wage bargaining.
- Complementary reforms recommended:
  - Product market reforms to spur labor productivity.
  - Financial policies to strengthen banking system support to the real economy (IMF, 2019a).
  - Reduction of the labor tax wedge as part of comprehensive tax reform (Cammeraat and Crivelli, 2020) would help reduce labor cost overvaluation, but would not be sufficient to link wages with productivity at the firm level.

*Source: wpiea2020038-print-pdf (IMF Working Paper content provided).*

### REFERENCES ______________________________________________________________15

### REFERENCES

### Introduction — key findings and interpretation
- Italian real wages have grown faster than labor productivity since joining the euro; post-global financial crisis (GFC) both real wage growth and labor productivity stalled, leaving elevated unit labor costs relative to euro area peers.
- National accounts indicate a growing share of value added going to employee compensation (unit labor cost), mechanically eroding corporate profits and capital returns and coinciding with a decline in corporate investment.
- Main empirical estimate: a 1 percent increase in real wages causes a ⅓ percent fall in fixed capital over one year.
- Distributional response: profits absorb only ½ of the associated cost increase; the remaining share is passed through to capital formation reductions.
- Mechanism: given short-term rigidity of labor quantities in Italy, liquidity constraints are identified as the main channel linking higher wages to lower investment; evidence: investment–wage elasticity is relatively more negative during the crisis years (2009–2012).
- Heterogeneity: firms in the South and medium firms display lower elasticities (likely due to lower effective coverage of national wage contracts). Results are robust to specification changes and to excluding leading firms in a sector.

### Aggregate facts and institutional background
- Decomposition of non-financial corporate (NFC) value added shows rising employee compensation share since Italy joined the euro in 2002; rising depreciation also compressed net returns to capital (interest + dividends + profits).
- Comparison with euro area: capital returns and investment fell across the euro area after the GFC, but the decline was much more pronounced in Italy, coinciding with Italy-specific growth of labor costs as a share of value added.
- Wage-bargaining institutional facts:
  - More than 97 percent of Italian dependent workers are subject to national-level wage contracts (CCNLs).
  - CCNL minima bind for about 90 percent of workers.
  - CCNLs typically last 3 years.
  - 80 percent of workers remain on permanent contracts with relatively high firing costs.
- Implication: downwardly rigid nominal wages combined with low inflation contributed to a real exchange rate overvaluation identified in IMF assessments.

### Theory and empirical model
- Theoretical channels from wages to investment:
  - Negative: factor complementarity (wage hike reduces labor, lowers marginal product of capital); financial constraints/liquidity (wage hike reduces cash flow, limiting investment).
  - Positive: efficiency wages or substitution of labor for capital could imply positive wage–investment relationship (not supported by main findings).
- Empirical identification:
  - Uses negotiated sectoral wage growth (national IStat CCNL-based series) as exogenous to individual firms.
  - Interaction of sectoral wage growth with firm-level lagged labor share provides an exogenous proxy for firm-level labor cost shocks.
- Empirical specification (panel):
  - Dependent variable: ∆k_{i,t} / k_{i,t−1} (real capital growth) or cash flow / lagged capital.
  - Key regressor: α_{i,t−1} * (∆w_{j,t} / w_{j,t−1}), with firm and sector-year fixed effects and controls (lagged dependent variable, labor share, interaction of capital share and interest rate growth, lag capital growth, lag debt/total assets, low ICR dummy, firm size and age categories).

### Micro data and sample statistics
- Firm-level data: Orbis panel 2003–2017; cleaned sample for 2005–2017: around 200,000 firms per year; covers 2/3 of NFC value added.
- Sectoral wage growth: IStat negotiated national wage contract growth by NACE4 sectors for 2005–2017; 221 sectors with more than 20 valid observations in at least two consecutive years.
- Contractual real wage growth across sector-years ranged between -1 and 3 percent (2005–2017).
- Selected descriptive statistics (all variables deflated; monetary quantities are 2010 euro thousands):
  - Fixed Assets (EUR thousand): Mean 10,300,000; Std. Deviation 435,000,000
  - Cash flow (EUR thousand): Mean 1,350,024; Std. Deviation 45,700,000
  - Costs of employees (EUR thousand): Mean 2,250,687; Std. Deviation 27,900,000
  - Value Added (EUR thousand): Mean 4,260,339; Std. Deviation 94,000,000
  - Fixed Asset growth: Mean -0.00860; Std. Deviation 0.3174
  - Cash flow / Lag Fixed Assets: Mean 0.36; Std. Deviation 0.39
  - Negotiated wage growth: Mean 0.0061; Std. Deviation 0.0098
  - Lag labor share: Mean 0.67; Std. Deviation 0.18
  - Interest rate growth: Mean -0.0022; Std. Deviation 0.0070
  - Lag Debt / Total Assets: Mean 0.18; Std. Deviation 0.19
  - Lag Low ICR (dummy): Mean 0.47; Std. Deviation 0.50
  - Number of Employees: Mean 43.2; Std. Deviation 533
  - age (years): Mean 20.8; Std. Deviation 14.0
  - Sales share in sector: Mean 0.0015; Std. Deviation 0.0105
  - Lag labor share (alternative): Mean 0.65; Std. Deviation 0.19

### Empirical results — sector and firm levels
- Sector-level correlations (sector-year observations):
  - wage_gr on lab_productivity_gr: beta 0.004 (N 2,745; R2 0.000) — wage growth unrelated to productivity growth.
  - fixed_asset_gr on wage_gr: beta -0.744*** (N 2,745; R2 0.362) — sectors with higher wage growth tend to have lower fixed asset growth. (Standard errors clustered at sector level.)
- Firm-level main result (multiplying estimated β̂2 by average labor share ᾱ = 0.67):
  - A one percentage point increase in wage growth causes a 0.33 percent (⅓ percent) reduction in fixed capital over one year (reported as main regression finding).
- Crisis-period amplification: elasticity more negative during 2009–2012.
- Robustness: similar results when using two-year lag or average of t-1 and t-2 for lagged labor share; robust to excluding leading firms in a sector.

### Policy implications and quantitative illustrations
- Repairing corporate and bank balance sheets and complementary structural reforms (product and service market reforms) are important to restore investment and productivity growth.
- Illustration from extrapolation of regression estimates:
  - Correcting the overvaluation in the unit-labor-cost real exchange rate identified in IMF External Balance Assessment could bring investment back to its pre-crisis average.
  - A €9 minimum wage (draft legislation in 2019) could reduce the fixed capital stock by 0.8 percent, illustrating that labor cost increases would curtail investment.

*Source: wpiea2020038-print-pdf (IMF Working Paper content provided).*

### 0.36 percentage point fall in fixed asset growth and a 0.78 percentage point fall in cash flow

### 0.36 percentage point fall in fixed asset growth and a 0.78 percentage point fall in cash flow

### Main regression results and key estimates
- Baseline interpretation: firms absorb about half of the increase in their production cost after a wage increase by lowering their profits, and pass through the other half by cutting investment. Both estimates are statistically different from zero at the 5 percent confidence level.
- Table 3 (selected coefficients reported exactly as in source):
  - dependent variable: fixed_assets_gr | cashflow_k
  - L.dep var: -0.129*** | 0.166***
  - L.labor_share: 0.096*** | 0.091***
  - l_lab_sh*wage_gr: -0.541** | -1.157***
  - l_cap_sh*loan_rate_gr: -0.229 | 2.501***
  - L.debt/assets: -0.122*** | -0.121***
  - L.lowICR: -0.053*** | -0.028***
  - L.log employees: -0.019*** | -0.023***
  - age bin=2: -0.016*** | 0.013***
  - age bin=3: -0.001 | 0.018***
  - age bin=4: 0.008 | 0.020***
  - age bin=5: 0.014** | 0.022***
  - age bin=6: 0.009 | 0.021***
  - age bin=7: 0.005 | 0.015**
  - r2_a_within: 2.3% | 3.3%
  - N: 888,653 | 844,484
- Estimation details:
  - Regression estimated using Correia (2017) procedure (Stata “reghdfe”).
  - Includes firm and sector-year fixed effects.
  - Sample excludes one-employee firms, insolvent firms, SOEs, non-market sectors, and is winsorized.
  - Variables deflated; standard errors clustered at the firm level.
  - Asterisks indicate statistical significance at the 10, 5, and 1 percent level respectively.

### Theoretical channels: factor complementarity vs. financial constraints
- Empirical observations:
  - Labor quantities appear very rigid with respect to wages at the one-year horizon; a regression with employment growth as dependent variable yields insignificant coefficients (not shown).
  - High firing costs and the Cassa Integrazione Guadagni subsidy likely contribute to employment rigidity.
  - With employment relatively rigid, factor complementarity (which relies on labor falling) cannot be the main channel.
  - Large impact of wages on cash flows suggests liquidity constraints may play a role.
  - The coefficient of interest is significantly more negative during the global financial and European confidence crises (2009–2012), reinforcing the liquidity-constraint hypothesis.
  - Partial anticipation of future real wage growth (triennial nominal wage contracts) can interact with liquidity constraints to hinder optimal adjustment of capital and labor.
- Heterogeneity:
  - Interactions with macro-regions and firm size show:
    - Estimated investment-wage elasticities significantly weaker for firms in the South and slightly weaker for medium firms.
    - Possible explanations for weaker elasticities in the South or medium firms: higher prevalence of temporary contracts, CCNLs covering a smaller share of workers, or balance sheet variables being less precisely measured.

### Period, region, and firm-size heterogeneity (Table 4 reported values)
- Dependent variable: fixed_asset_gr
  - omitted group: -0.089 | -0.649*** | -0.990***
    - (0.295) (0.249) (0.335)
  - crisis (2009–2012): -1.044*** (0.399)
  - Center: 0.179 (0.160)
  - South: 0.593*** (0.168)
  - small (2–14 empl.): 0.276 (0.251)
  - medium (15–199 empl.): 0.647*** (0.247)
  - r2_a_within: 2.3% (for each column)
  - N: 888,653 (for each column)
- Notes: omitted groups are normal years, firms in the North, and large firms. Standard errors in parentheses.

### Robustness checks and elasticity range
- Robustness cases (Figure 5 summary; descriptions preserved):
  - Baseline (Table 3, after multiplying by average labor share).
  - Weighted by average real fixed assets of a firm.
  - Regression without controls (γ⃗ = 0).
  - Simple regression plus interaction between interest rates and capital share.
  - Main regression without lagged dependent variable.
  - Lags taken as average of previous two years (c−1 and c−2); two-year lag (c−2) gives very similar results.
  - Exclude firms with share of sales above 5 percent in a year-sector.
  - Alternative labor share definition using total factor payments and cost of capital r_k = lending rate + equity premium approximated to 7 percent.
- Estimated elasticities:
  - Always negative across robustness cases, mostly ranging from -¼ to -¾.
  - Statistical significance lost in some models due to large number of fixed-effect parameters (more than 200,000) absorbing variation.

### Aggregate implications and illustrative extrapolations
- External rebalancing:
  - IMF External Balance Assessment (IMF, 2019b) concluded Italy’s unit labor cost real exchange rate is overvalued by 0 to 10 percent.
  - Illustrative extrapolation: a 6 percent wage devaluation would increase investment over fixed assets by 2.2 percentage points, bringing it back to its pre-GFC average level (from the 2017 level).
  - Complementary model-based evidence: Andrle et al. (2018) predicts a 15 percent reduction in wage markups would boost steady-state investment by around 3 percent.
  - Higher investment would translate into higher labor productivity, enabling expansion of employment and higher steady-state wages.
  - Decentralization of wage bargaining and product and service market reforms could reduce need for initial wage adjustment and yield better allocation of labor.

- Minimum wage illustrative calculation:
  - INPS (2019) estimates a minimum wage of €9 per hour would increase labor costs by €9.7 billion, or 2.2 percent of total labor costs.
  - Extrapolation of regression results suggests such minimum wage would reduce the fixed capital stock by 0.8 percent.
  - Caveats:
    - Prediction based on average worker; responses may vary across wage distribution.
    - Low-skilled workers may be more substitutable with capital, potentially mitigating or overturning predicted investment impact.
    - INPS estimate considers only workers currently below the proposed minimum; upward pressure on wages above the minimum could imply a larger investment decline.

### Policy takeaways
- Correcting wage overvaluation should boost investment; introducing a minimum wage at an excessive level would harm investment.
- Need for a wage bargaining system that takes into account investment impacts—specifically, realigning real wages with labor productivity, for instance through decentralized wage bargaining.
- Complementary reforms recommended:
  - Product market reforms to spur labor productivity.
  - Financial policies to strengthen banking system support to the real economy (IMF, 2019a).
  - Reduction of the labor tax wedge as part of comprehensive tax reform (Cammeraat and Crivelli, 2020) would help reduce labor cost overvaluation, but would not be sufficient to link wages with productivity at the firm level.

### Data, sample construction, and variable definitions (selected)
- National accounts (Table A.1) and firm-level variable definitions (Table B.1) used; key Orbis definitions:
  - Labor share = Costs_of_employees / Added_value
  - Capital share = 1 – Labor share
  - Debt = Loans + Long_Term_Debt
  - Firm age = (Closing_date - Date_of_incorporation) / 365.25
  - Alternative labor share = Costs_of_employees / (Costs_of_employees + (rrate_outstanding + 0.07) * Total_Assets)
  - rrate_outstanding is the interest rate on NFC outstanding loans (Bank of Italy).
- Sample exclusions applied to Orbis firm-year observations (selected thresholds preserved):
  - non-market sectors: agriculture, mining, financials, real estate, public administration and extraterritorial bodies
  - insolvent firms: any firm status other than “Active” or "Active (default of payment)"
  - state-owned enterprises with controlling shareholder "Public authority, state, government"
  - duplicate consolidated accounts (C2 category)
  - non-positive fixed assets, cash flow, value added or age
  - one or less employees
  - fixed asset growth above 100 percent or below -50 percent (contemporaneous or lagged)
  - cash flow over fixed assets below 200 percent (contemporaneous or lagged)
  - lagged labor share < 10 percent or > 99 percent
  - debt over total assets > 95 percent
  - year-sector cluster with less than 20 observations
  - sector-level regressions drop sectors with yearly productivity growth outside -20 to 20 percent or capital growth outside -30 to 50 percent
- Merging: remaining Orbis data merged with negotiated wages by sector (Istat “hourly index of wages according to collective labour agreements (Nace rev.2)”), deflators (IStat), and interest rates on NFC outstanding loans (Bank of Italy).

*Sources: Orbis, IStat and Bank of Italy; regression estimated using the procedure in Correia (2017).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020038-print-pdf.pdf_