## Section VII concludes.

## Source details

**Canonical URL:** [Section VII concludes.](https://www.imf.org/-/media/files/publications/wp/2018/wp1873.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp1873.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp1873.pdf.json)

---

### Literature review — context and contribution
- Recent decline in productivity has raised interest in the effects of structural reforms.  
- Prior evidence (sector and firm level) shows structural reforms increase productivity with effects that take time to materialize; much of this literature focuses on advanced economies.  
- Key findings from the literature summarized in the paper:
  - Reductions in labor tax wedges and increases in public spending on active labor market policies have larger effects during contractionary periods; lowering employment protection and benefits are procyclical (IMF (2016); Adascalite and Morano (2016); Duval and Furceri (2017)).  
  - Labor productivity tends to be weaker in industries with more stringent employment protection; industries with more binding layoff restrictions suffer lower productivity growth (Bassanini and Duval (2009); Bassanini et al. (2000)).  
  - Product market reforms have positive effects on output and employment growth independently of the business cycle (IMF (2016)); deregulation in network sectors raises value added and productivity, with larger gains where public administration is more efficient (Lanau and Topolova (2016)).  
  - Trade liberalization evidence shows reallocation toward more productive firms and products; firm initial productivity, size, import intensity, and financial health matter for gains from trade liberalization (Melitz (2003); Pavcnik (2002); Amiti and Konings (2007); Manova (2012); Chaney (2016)).  
  - Quality of institutions is a critical driver of private investment and entrepreneurship; improvements in legal systems increase firm size and reduce idiosyncratic risk (Acemoglu, Johnson and Robinson (2005); Laeven and Woodruff (2007); Beck, Demirgüç-Kunt, and Maksimovic (2005)).  
- Paper’s main contributions:
  - Extends analysis beyond advanced economies by including emerging market firms; finds macroeconomic effects of structural reforms in emerging markets broadly similar to advanced economies.  
  - Analyzes multiple types of structural reforms comprehensively to tease out differential effects on output and employment at the firm level and document heterogeneous effects across firm characteristics.

### Methodology — empirical approach and identification
- Main estimation approach: local projection method (Jordà, 2005; Teulings and Zubanov, 2014) to derive impulse response functions.  
- Baseline regression specification (symbols preserved as in source):
  - y_{isc,t+h} − y_{isc,t−1} = β_{0,h} + β_{1,h} R_{ct} + β_{2,h} X_{isc,t−1} + β_{3,h} (gdp_{c,t+h} − gdp_{c,t−1}) + μ_s + φ_c + ω_t + ε_{isc,t+h}, ∀ h = 0,1,2,3,4,5.
  - Subscripts: isc,t denote firm i, sector s, country c, year t. y is log of output (log of employment). R is reform shock indicator (1 for major structural reform, 0 otherwise). X are firm-specific controls (size proxied by log of total assets). Country cumulative GDP growth (gdp_{c,t+h} − gdp_{c,t−1}) included. Sector, country, year fixed effects μ_s, φ_c, ω_t included.  
- Standard errors: robust and clustered at the frequency of the reform shock — clustered by country and year for labor, trade, and institutional reforms; by country and sector for product market reforms (results unchanged if clustered by country only).  
- Horizons: impulse responses computed for h ∈ {0,..,5}; 90 percent confidence bands reported. Shock assumed at period t=0.  
- Interaction specification to capture heterogeneity (firm characteristic dummy d_i equals 1 if firm is above its country median, using lagged characteristics to avoid endogeneity):
  - y_{isc,t+h} − y_{isc,t−1} = β_{0,h} + β_{1,h} R_{ct} + β_{2,h} R_{ct} × d_i + β_{3,h} d_i + β_{4,h} X_{isc,t−1} + β_{5,h} (gdp_{c,t+h} − gdp_{c,t−1}) + μ_s + φ_c + ω_t + ε_{isc,t+h}, ∀ h = 0,1,2,3,4,5.
  - Focus on sign and significance of β_{2,h} to assess whether firm characteristics influence reform effects.

### Data — reform indices and firm-level sample
- Structural reform measures:
  - Labor market: mandated cost of worker dismissal from World Bank Ease of Doing Business.  
  - Product market: OECD Regulatory Impact index measuring cost of anti-competitive regulation in upstream network sectors (energy, transport, communication) on 37 downstream sectors.  
  - Trade: non-tariff regulatory trade barriers index from the Fraser Institute.  
  - Institutions: legal rights index from the Fraser Institute.  
  - All indices annual. Product market indices available 1975–2013; others 2000–2014. For the analysis the paper considers reforms between 2004 and 2009 (firm-level data available 2003–2014 and analysis focuses on 5-year effects).  
- Reform shock identification:
  - A reform shock is an increase in the structural reform index exceeding two standard deviations of changes over all observations; reform variable is a dummy equal to 1 when a reform shock is observed. For OECD product market reforms, 1 standard deviation cutoff used for sample size.  
  - Robustness checks use: (i) largest 5th percentile changes (Gal and Hijzen (2016)); (ii) absolute changes in the reform indices (to exploit reform intensity).  
  - Incidence: average 4 percent of all potential country-year (or country-sector-year for PMR) observations; mandated cost of dismissal reforms most observed with 6 percent incidence; other reform shocks around 3 percent. All labor market reforms in sample happened in 2009.
- Firm-level data (Orbis):
  - Sample years: 2003–14. Around 4 million firm-year observations (4,222,303). 30 countries included; 11 emerging markets account for around 40 percent of firm-year observations.  
  - Cleaning/sample criteria include: drop micro firms (<4 employees); non-agricultural market sectors (ISIC 5–82); firms present at least 5 consecutive years (panel locally balanced); financial accounts refer to whole fiscal year; non-missing strictly positive employment, assets, gross output in all years; country included only if ≥ 1,000 firm observations each year.  
  - Nominal variables deflated using WEO GDP deflators; robustness check using sector-level WIOD SEA deflators (sample reduced by 25 percent) yields similar results.  
- Outcome variables:
  - Output proxied by operating revenue; employment by number of employees. Labor productivity growth (revenue per employee) used as robustness check.

### Results — baseline and heterogeneous effects
- Baseline impulse response findings:
  - Structural reforms lead to higher output at the firm level and effects take time to materialize (lags).  
  - Effects on employment are generally insignificant, suggesting reforms operate by increasing labor productivity.  
  - Summary (Short term (ST = one year after reform) and Medium term (MT = years 2–5)):
    - Labor reforms: ST (no marked short-term positive effect), MT (+ at t=5 for output in baseline discussions) — employment insignificant in ST and MT.  
    - Product market reforms: positive in ST and MT for output (strengthens over time); employment not significant (baseline) — implies product market reforms raise labor productivity.  
    - Trade reforms: positive but statistically insignificant effects on output and employment in ST and MT.  
    - Institutional reforms: positive effect on output growth; no significant effect on employment.
- Heterogeneous effects by firm characteristics:
  - Labor market reforms (lower mandated dismissal costs):
    - Stronger returns for: Large firms; High-skilled labor firms; Low labor-intensive firms.
    - Other firm characteristics appear not to be significant channels for labor reform effects.
  - Product market reforms (lower input costs from deregulated network sectors):
    - More beneficial for: More leveraged firms; Service-oriented firms; Low-skilled labor firms.
    - When excluding firms from deregulated sectors (ISIC codes 40–1 and 60–4), manufacturing firms profit more from deregulation.
  - Trade reforms (non-tariff barriers reduction):
    - Greater benefits to: More profitable firms; Leveraged firms; Service firms.
    - Interpretation: high-profit firms better positioned to compete in international markets and leverage global value chains; lower fixed cost of trading benefits leveraged firms.
  - Institutional reforms:
    - Stronger gains for service firms (consistent with institutional improvements reducing uncertainty and incentivizing investment, especially where intangible assets and intellectual property protection matter).

### Caveats and limitations
- ORBIS not necessarily representative for every country; sample restricted to countries with ≥ 1,000 firm observations per year.  
- Entry and exit channel for reforms cannot be fully explored because ORBIS lacks clear entry/exit dates; sample restricted to firms present at least 5 consecutive years.  
- Reforms often implemented in packages — analysis considers reforms one at a time.  
- Short sample period for firm-level data (2004–2009 reform window) and the analysis estimates time to effect post-implementation but not time to implement reforms.

### Robustness checks — alternative identification and specifications
- Alternative reform shock identification:
  - Largest 5th percentile changes and absolute changes in reform indices produce results robust to baseline findings.  
- Sectoral deflators:
  - Using WIOD SEA sector-level deflators (sample reduced by 25 percent) yields very similar results.  
- Richer fixed effects:
  - Including firm fixed effects and sector-specific time trends — results robust; main difference: product market reforms show a negative impact on employment under richer fixed effects (consistent with Bassanini (2015)).  
- Labor productivity:
  - Directly measuring labor productivity growth (revenue per employee) confirms structural reforms increase labor productivity.  
- Deleting predominant countries:
  - Excluding Japan, Russia, Spain, and Ukraine (each > 10 percent of firm-year observations) yields very similar baseline results.

### Select numerical results (robustness check using absolute change in the reform index — Table 3 figures preserved exactly)
- Labor market reforms
  - Output growth (t=0 to t=5): -0.0167, 0.0079, 0.0040, -0.0053, 0.0201, 0.0104.  
  - Employment growth (t=0 to t=5): 0.0028, 0.0024, 0.0024, 0.0050, 0.0014, -0.0114.  
  - Output growth standard errors (t=0 to t=5): 0.0182, 0.0207, 0.0287, 0.0257, 0.0488, 0.0376.  
  - Employment growth standard errors (t=0 to t=5): 0.0056, 0.0093, 0.0093, 0.0099, 0.0114, 0.0105.
- Product market reforms
  - Output growth (t=0 to t=5): -0.6111***, 0.2743, 0.8351, 1.1452**, 1.5440**, 1.4608**.  
  - Employment growth (t=0 to t=5): 0.2748***, -0.3274**, -0.4505**, -0.1174, 0.1203, 0.2131.  
  - Output growth standard errors (t=0 to t=5): 0.2283, 0.3608, 0.5804, 0.5917, 0.6601, 0.7632.  
  - Employment growth standard errors (t=0 to t=5): 0.1004, 0.1426, 0.2145, 0.2789, 0.3277, 0.3836.
- Trade reforms
  - Output growth (t=0 to t=5): 0.0105, -0.1267, -0.1246, -0.0531, -0.0890, -0.0055.  
  - Employment growth (t=0 to t=5): 0.0030, -0.0115, -0.0014, 0.0018, -0.0064, -0.0014.  
  - Output growth standard errors (t=0 to t=5): 0.0418, 0.0817, 0.0898, 0.0760, 0.0909, 0.0659.  
  - Employment growth standard errors (t=0 to t=5): 0.0168, 0.0150, 0.0240, 0.0211, 0.0176, 0.0193.
- Institutional reforms
  - Output growth (t=0 to t=5): -0.0413, 0.0179, 0.0684**, 0.1145***, 0.1024***, 0.1135***.  
  - Employment growth (t=0 to t=5): -0.0039, -0.0116, -0.0147, 0.0038, 0.0177**, 0.0046.  
  - Output growth standard errors (t=0 to t=5): 0.0325, 0.0302, 0.0355, 0.0367, 0.0348, 0.0321.  
  - Employment growth standard errors (t=0 to t=5): 0.0102, 0.0110, 0.0141, 0.0165, 0.0083, 0.0075.

### Concluding remarks — policy-relevant takeaways
- Structural reforms raise firm-level output primarily by increasing labor productivity; effects take time to materialize.  
- Over the analyzed period (2004–2009), product market and institutional reforms exhibit higher payoffs than labor market and trade reforms in the sample; caveat that labor market reforms in the sample were identified during downturns and trade reforms were largely implemented before 2004.  
- Benefits of reforms vary with firm characteristics (size, profitability, leverage, sector), highlighting:
  - The importance of targeting structural reforms to country-specific firm and sectoral structures to maximize returns.  
  - The usefulness of complementary policies to mitigate redistributional costs and support inclusive gains from reforms.

### Appendix A. Definition of Structural Reforms
- Labor Market
  - Measure used: mandated cost of worker dismissal based on the World Bank’s Doing Business report.  
  - Captures: requirements for advanced notice, severance payments, and penalties due when dismissing a redundant worker.  
  - Interpretation: the Doing Business report gives a lower rating to countries with costly dismissal requirements.
- Product market Reforms
  - Measures used: OECD indicators on regulatory impact (RI).  
  - Scope: quantifies the potential costs of anti-competitive regulation in upstream sectors (electricity, transportation, and communication) on 37 downstream sectors that use the output of these sectors as intermediate inputs.  
  - Data coverage: available in 32 OECD and 2 non-OECD countries on an annual basis starting in 1975.  
  - Additional index: Energy, Transport, and Communication Regulation (ETCR) index. Captures: the degree of liberalization in the telecommunication and electricity markets, including the extent of competition in the provision of these services, the presence of an independent regulatory authority, and privatization.
- Trade Regulations
  - Measure used: regulatory trade barriers index from the Fraser Institute.  
  - Composition: an average of the non-tariff trade barriers index from the World Economic Forum’s Global Competitiveness Report, the compliance cost of importing and exporting index from World Bank’s Doing Business Report, and the black-market exchange rates index from the MRI Bankers’ Guide to Foreign Currency.  
  - Interpretation: the black-market exchange rates index gives high score to countries with a domestic currency that is fully convertible without restrictions.
- Institutional reforms
  - Measure used: integrity of the legal system index from the Fraser Institute Economic Freedom Index.  
  - Composition: the index is composed using the PRS Group’s International Country Risk Guide measure of risk component.  
  - Components: the risk component is an average of law and order, with the “law” sub-component measuring the strength and impartiality of the legal system and the “order” sub-component measuring the popular observance of the law.

*Source: wp1873 - Section VII concludes.*

### Section VII concludes.

### Section VII concludes.

### Literature review — context and contribution
- Recent decline in productivity has raised interest in the effects of structural reforms.  
- Prior evidence (sector and firm level) shows structural reforms increase productivity with effects that take time to materialize; much of this literature focuses on advanced economies.  
- Key findings from the literature summarized in the paper:
  - Reductions in labor tax wedges and increases in public spending on active labor market policies have larger effects during contractionary periods; lowering employment protection and benefits are procyclical (IMF (2016); Adascalite and Morano (2016); Duval and Furceri (2017)).  
  - Labor productivity tends to be weaker in industries with more stringent employment protection; industries with more binding layoff restrictions suffer lower productivity growth (Bassanini and Duval (2009); Bassanini et al. (2000)).  
  - Product market reforms have positive effects on output and employment growth independently of the business cycle (IMF (2016)); deregulation in network sectors raises value added and productivity, with larger gains where public administration is more efficient (Lanau and Topolova (2016)).  
  - Trade liberalization evidence shows reallocation toward more productive firms and products; firm initial productivity, size, import intensity, and financial health matter for gains from trade liberalization (Melitz (2003); Pavcnik (2002); Amiti and Konings (2007); Manova (2012); Chaney (2016)).  
  - Quality of institutions is a critical driver of private investment and entrepreneurship; improvements in legal systems increase firm size and reduce idiosyncratic risk (Acemoglu, Johnson and Robinson (2005); Laeven and Woodruff (2007); Beck, Demirgüç-Kunt, and Maksimovic (2005)).  
- Paper’s main contributions:
  - Extends analysis beyond advanced economies by including emerging market firms; finds macroeconomic effects of structural reforms in emerging markets broadly similar to advanced economies.
  - Analyzes multiple types of structural reforms comprehensively to tease out differential effects on output and employment at the firm level and document heterogeneous effects across firm characteristics.

### Methodology — empirical approach and identification
- Main estimation approach: local projection method (Jordà, 2005; Teulings and Zubanov, 2014) to derive impulse response functions.  
- Baseline regression specification (symbols preserved as in source):
  - y_{isc,t+h} − y_{isc,t−1} = β_{0,h} + β_{1,h} R_{ct} + β_{2,h} X_{isc,t−1} + β_{3,h} (gdp_{c,t+h} − gdp_{c,t−1}) + μ_s + φ_c + ω_t + ε_{isc,t+h}, ∀ h = 0,1,2,3,4,5.
  - Subscripts: isc,t denote firm i, sector s, country c, year t. y is log of output (log of employment). R is reform shock indicator (1 for major structural reform, 0 otherwise). X are firm-specific controls (size proxied by log of total assets). Country cumulative GDP growth (gdp_{c,t+h} − gdp_{c,t−1}) included. Sector, country, year fixed effects μ_s, φ_c, ω_t included.  
- Standard errors: robust and clustered at the frequency of the reform shock — clustered by country and year for labor, trade, and institutional reforms; by country and sector for product market reforms (results unchanged if clustered by country only).  
- Horizons: impulse responses computed for h ∈ {0,..,5}; 90 percent confidence bands reported. Shock assumed at period t=0.
- Interaction specification to capture heterogeneity (firm characteristic dummy d_i equals 1 if firm is above its country median, using lagged characteristics to avoid endogeneity):
  - y_{isc,t+h} − y_{isc,t−1} = β_{0,h} + β_{1,h} R_{ct} + β_{2,h} R_{ct} × d_i + β_{3,h} d_i + β_{4,h} X_{isc,t−1} + β_{5,h} (gdp_{c,t+h} − gdp_{c,t−1}) + μ_s + φ_c + ω_t + ε_{isc,t+h}, ∀ h = 0,1,2,3,4,5.
  - Focus on sign and significance of β_{2,h} to assess whether firm characteristics influence reform effects.

### Data — reform indices and firm-level sample
- Structural reform measures:
  - Labor market: mandated cost of worker dismissal from World Bank Ease of Doing Business.  
  - Product market: OECD Regulatory Impact index measuring cost of anti-competitive regulation in upstream network sectors (energy, transport, communication) on 37 downstream sectors.  
  - Trade: non-tariff regulatory trade barriers index from the Fraser Institute.  
  - Institutions: legal rights index from the Fraser Institute.  
  - All indices annual. Product market indices available 1975–2013; others 2000–2014. For the analysis the paper considers reforms between 2004 and 2009 (firm-level data available 2003–2014 and analysis focuses on 5-year effects).  
- Reform shock identification:
  - A reform shock is an increase in the structural reform index exceeding two standard deviations of changes over all observations; reform variable is a dummy equal to 1 when a reform shock is observed. For OECD product market reforms, 1 standard deviation cutoff used for sample size.  
  - Robustness checks use: (i) largest 5th percentile changes (Gal and Hijzen (2016)); (ii) absolute changes in the reform indices (to exploit reform intensity).  
  - Incidence: average 4 percent of all potential country-year (or country-sector-year for PMR) observations; mandated cost of dismissal reforms most observed with 6 percent incidence; other reform shocks around 3 percent. All labor market reforms in sample happened in 2009.
- Firm-level data (Orbis):
  - Sample years: 2003–14. Around 4 million firm-year observations (4,222,303). 30 countries included; 11 emerging markets account for around 40 percent of firm-year observations.  
  - Cleaning/sample criteria include: drop micro firms (<4 employees); non-agricultural market sectors (ISIC 5–82); firms present at least 5 consecutive years (panel locally balanced); financial accounts refer to whole fiscal year; non-missing strictly positive employment, assets, gross output in all years; country included only if ≥ 1,000 firm observations each year.  
  - Nominal variables deflated using WEO GDP deflators; robustness check using sector-level WIOD SEA deflators (sample reduced by 25 percent) yields similar results.  
- Outcome variables:
  - Output proxied by operating revenue; employment by number of employees. Labor productivity growth (revenue per employee) used as robustness check.

### Results — baseline and heterogeneous effects
- Baseline impulse response findings:
  - Structural reforms lead to higher output at the firm level and effects take time to materialize (lags).  
  - Effects on employment are generally insignificant, suggesting reforms operate by increasing labor productivity.  
  - Summary (Text Table 1-style): Short term (ST = one year after reform) and Medium term (MT = years 2–5) indications:
    - Labor reforms: ST (no marked short-term positive effect), MT (+ at t=5 for output in baseline discussions) — employment insignificant in ST and MT.
    - Product market reforms: positive in ST and MT for output (strengthens over time); employment not significant (baseline) — implies product market reforms raise labor productivity.
    - Trade reforms: positive but statistically insignificant effects on output and employment in ST and MT.
    - Institutional reforms: positive effect on output growth; no significant effect on employment.
- Heterogeneous effects by firm characteristics (Text Table 2 and Figure 5 summary):
  - Labor market reforms (lower mandated dismissal costs):
    - Stronger returns for: Large firms; High-skilled labor firms; Low labor-intensive firms.
    - Other firm characteristics appear not to be significant channels for labor reform effects.
  - Product market reforms (lower input costs from deregulated network sectors):
    - More beneficial for: More leveraged firms; Service-oriented firms; Low-skilled labor firms.
    - When excluding firms from deregulated sectors (ISIC codes 40–1 and 60–4), manufacturing firms profit more from deregulation.
  - Trade reforms (non-tariff barriers reduction):
    - Greater benefits to: More profitable firms; Leveraged firms; Service firms.
    - Interpretation: high-profit firms better positioned to compete in international markets and leverage global value chains; lower fixed cost of trading benefits leveraged firms.
  - Institutional reforms:
    - Stronger gains for service firms (consistent with institutional improvements reducing uncertainty and incentivizing investment, especially where intangible assets and intellectual property protection matter).
- Caveats noted by authors:
  - ORBIS not necessarily representative for every country; sample restricted to countries with ≥ 1,000 firm observations per year.
  - Entry and exit channel for reforms cannot be fully explored because ORBIS lacks clear entry/exit dates; sample restricted to firms present at least 5 consecutive years.
  - Reforms often implemented in packages — analysis considers reforms one at a time.
  - Short sample period for firm-level data (2004–2009 reform window) and the analysis estimates time to effect post-implementation but not time to implement reforms.

### Robustness checks — alternative identification and specifications
- Alternative reform shock identification:
  - Largest 5th percentile changes and absolute changes in reform indices produce results robust to baseline findings (Figures 6–7; Table 3 presents absolute-change robustness coefficients).  
- Sectoral deflators:
  - Using WIOD SEA sector-level deflators (sample reduced by 25 percent) yields very similar results (Figure 8).  
- Richer fixed effects:
  - Including firm fixed effects and sector-specific time trends — results robust; main difference: product market reforms show a negative impact on employment under richer fixed effects (consistent with Bassanini (2015)). (Figures 9–10).  
- Labor productivity:
  - Directly measuring labor productivity growth (revenue per employee) confirms structural reforms increase labor productivity (Figure 11).  
- Deleting predominant countries:
  - Excluding Japan, Russia, Spain, and Ukraine (each > 10 percent of firm-year observations) yields very similar baseline results (Figures 12–13).

### Select numerical results (preserved exactly as in source Table 3 — robustness check using absolute change in the reform index)
- Output growth coefficients (t=0 to t=5) and associated employment growth columns as presented in Table 3 (rows by reform type, numbers preserved):
  - Labor market reforms (Output growth t=0 to t=5): -0.0167, 0.0079, 0.0040, -0.0053, 0.0201, 0.0104.  
    - Employment growth (t=0 to t=5): 0.0028, 0.0024, 0.0024, 0.0050, 0.0014, -0.0114.  
    - Lower panel standard errors (for output growth row): 0.0182, 0.0207, 0.0287, 0.0257, 0.0488, 0.0376. For employment growth row: 0.0056, 0.0093, 0.0093, 0.0099, 0.0114, 0.0105.
  - Product market reforms (Output growth t=0 to t=5): -0.6111***, 0.2743, 0.8351, 1.1452**, 1.5440**, 1.4608**.  
    - Employment growth (t=0 to t=5): 0.2748***, -0.3274**, -0.4505**, -0.1174, 0.1203, 0.2131.  
    - Lower panel standard errors (for output growth row): 0.2283, 0.3608, 0.5804, 0.5917, 0.6601, 0.7632. For employment growth row: 0.1004, 0.1426, 0.2145, 0.2789, 0.3277, 0.3836.
  - Trade reforms (Output growth t=0 to t=5): 0.0105, -0.1267, -0.1246, -0.0531, -0.0890, -0.0055.  
    - Employment growth (t=0 to t=5): 0.0030, -0.0115, -0.0014, 0.0018, -0.0064, -0.0014.  
    - Lower panel standard errors (for output growth row): 0.0418, 0.0817, 0.0898, 0.0760, 0.0909, 0.0659. For employment growth row: 0.0168, 0.0150, 0.0240, 0.0211, 0.0176, 0.0193.
  - Institutional reforms (Output growth t=0 to t=5): -0.0413, 0.0179, 0.0684**, 0.1145***, 0.1024***, 0.1135***.  
    - Employment growth (t=0 to t=5): -0.0039, -0.0116, -0.0147, 0.0038, 0.0177**, 0.0046.  
    - Lower panel standard errors (for output growth row): 0.0325, 0.0302, 0.0355, 0.0367, 0.0348, 0.0321. For employment growth row: 0.0102, 0.0110, 0.0141, 0.0165, 0.0083, 0.0075.

### Concluding remarks — policy-relevant takeaways
- Structural reforms raise firm-level output primarily by increasing labor productivity; effects take time to materialize.  
- Over the analyzed period (2004–2009), product market and institutional reforms exhibit higher payoffs than labor market and trade reforms in the sample; caveat that labor market reforms in the sample were identified during downturns and trade reforms were largely implemented before 2004.  
- Benefits of reforms vary with firm characteristics (size, profitability, leverage, sector), highlighting:
  - The importance of targeting structural reforms to country-specific firm and sectoral structures to maximize returns.
  - The usefulness of complementary policies to mitigate redistributional costs and support inclusive gains from reforms.

*Source: wp1873 - Section VII concludes.*

### Appendix A. Definition of Structural Reforms

### Appendix A. Definition of Structural Reforms

### Labor Market
- Measure used: mandated cost of worker dismissal based on the World Bank’s Doing Business report.
- Captures: requirements for advanced notice, severance payments, and penalties due when dismissing a redundant worker.
- Interpretation: the Doing Business report gives a lower rating to countries with costly dismissal requirements.

### Product market Reforms
- Measures used: OECD indicators on regulatory impact (RI).
- Scope: quantifies the potential costs of anti-competitive regulation in upstream sectors (electricity, transportation, and communication) on 37 downstream sectors that use the output of these sectors as intermediate inputs.
- Data coverage: available in 32 OECD and 2 non-OECD countries on an annual basis starting in 1975.
- Additional index: Energy, Transport, and Communication Regulation (ETCR) index.
  - Captures: the degree of liberalization in the telecommunication and electricity markets, including the extent of competition in the provision of these services, the presence of an independent regulatory authority, and privatization.

### Trade Regulations
- Measure used: regulatory trade barriers index from the Fraser Institute.
- Composition: an average of
  - the non-tariff trade barriers index from the World Economic Forum’s Global Competitiveness Report,
  - the compliance cost of importing and exporting index from World Bank’s Doing Business Report, and
  - the black-market exchange rates index from the MRI Bankers’ Guide to Foreign Currency.
- Interpretation: the black-market exchange rates index gives high score to countries with a domestic currency that is fully convertible without restrictions.

### Institutional reforms
- Measure used: integrity of the legal system index from the Fraser Institute Economic Freedom Index.
- Composition: the index is composed using the PRS Group’s International Country Risk Guide measure of risk component.
- Components: the risk component is an average of law and order, with the “law” sub-component measuring the strength and impartiality of the legal system and the “order” sub-component measuring the popular observance of the law.

*Source: Appendix A. Definition of Structural Reforms (wp1873).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp1873.pdf_