## Section V — Labor Market Institutions, Informality, and Reform Evidence (wpiea2019055)

## Source details

**Canonical URL:** [Section V — Labor Market Institutions, Informality, and Reform Evidence (wpiea2019055)](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019055.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wpiea2019055.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019055.pdf.json)

---

### Methods and data
- Empirical approaches:
  - Simple cross-country regressions to explore effects of labor market institutions (example: employment protection legislation) on informality.
  - Synthetic control method (Abadie and Gardeazabal, 2003) used for case studies of a small number of Latin American labor market reforms since 2000.
  - Heterogeneous panel with common correlated effects (CCE) estimator (Pesaran, 2006) applied in Okun’s law analysis for 127 countries (annual data 1990–2017; countries with <20 years excluded).
- Data and measurement notes:
  - Minimum wage measures sourced from several providers; denominator choice critical (average wage, median wage, value-added).
  - Employment Protection Legislation (EPL) data: Laws & Regulations (L&Rs) from World Bank, OECD, ILO; Perceptions-Based Indicators (PBIs) from WEF executive survey.
  - Appendix A lists data coverage, sources, and comparative analysis.

### Stylized facts: labor markets in Latin America
- Employment and unemployment:
  - Total employment expanded at over 2 percent per year in the period prior to the financial crisis.
  - Argentina and Brazil recorded negative employment growth for at least one year after 2012.
  - Mexico, Peru: very low and relatively stable unemployment rates.
- Labor productivity:
  - Median output per worker in South America in 2017 was about 30 percent of the level for the median advanced economy (20 percent in Central America).
  - Productivity growth volatile: all countries except Peru and Chile achieved positive productivity growth in only two out of the four periods considered.
- Informality:
  - Informality accounts for over 50 percent of total employment in Latin America.
  - Informality in South and Central America broadly comparable to other EMDEs; AEs have significantly lower informality.
  - Large heterogeneity: informality ranges from around 30 to 70 percent among the largest economies.
  - Peru and Mexico (and Argentina to some degree) have higher informality than expected for their level of development.
  - Countries with lowest and most stable unemployment rates (Peru and Mexico) have the highest informality relative to development level.
- Labor market institutions and perceptions:
  - WEF perceptions: wage flexibility similar across Latin America and other groups; hiring and firing practices in South America perceived as substantially more rigid.
  - Aggregate OECD and ILO indices do not show South America as having stronger de jure employment protection.
  - Main drivers of perceived rigidity: cumbersome hiring and especially firing procedures rather than severance pay per se.
  - Specific institutional features: redundancy costs higher than in AEs or other EMDEs; permanent contracts mandatory for permanent tasks in many countries; dismissal of even 1 worker often requires third-party approval.
  - Ratio of minimum wage to value added: Central America stands out as having a very high ratio; no clear evidence that minimum wage is more binding in South America than elsewhere.

### Decomposing unemployment dynamics (sections II–III)
- Decomposition framework:
  - Unemployment changes decomposed into labor demand (output + labor productivity) and labor supply (participation rate + working-age population).
  - Alternative decomposition highlights formal employment and labor formality (formal/total employment ratio) as margins.
- Key period patterns (1997–2000; 2000–07; 2007–11; 2011–17):
  - Late 1990s: unemployment rose as supply outstripped demand.
  - 2000–11 commodity boom: unemployment fell as labor demand outpaced supply.
  - GFC: limited impact on unemployment in sample countries.
  - Post-2012: unemployment broadly stable except Brazil and Argentina where unemployment rose due to sharp demand drops.
- Margins of adjustment and informality role:
  - Working-age population growth largely stable.
  - Participation rate acted as an active adjustment margin (notably Chile, Colombia, Peru during boom years).
  - Labor productivity growth reduces labor demand; productivity fluctuations can reflect labor hoarding or changes in informality.
  - Informality acted as a crucial countercyclical margin (example: Colombia: informality rose in late 1990s, fell in 2000s, rose again in GFC).
  - Brazil and Mexico show less adjustment along informality margin in recent periods.
- Implication:
  - To assess labor market slack in Latin America, combine unemployment rate with formal employment growth and informality measures.

### Revisiting Okun’s law (section IV) — estimation and main results
- Estimation details:
  - CCE estimator applied to an unbalanced panel of 127 countries, annual 1990–2017; alternative gap specification via HP filter (smoothing parameter 6.25) yields similar results.
- Baseline estimates (Table 1):
  - Specification 1 (no lags): contemporaneous coefficient ߚ ≈ -0.125*** (standard error 0.0152).
  - Including up to four lags (specifications 2–5): contemporaneous coefficient around -0.115 to -0.126.
  - Only the first lag of ΔGDP is robustly significant in multi-lag specifications (ΔGDPt-1 ≈ -0.0330*** in spec. 2).
  - Observations: 3,399 (spec.1) down to 3,307 (spec.5); Countries: 127.
- Cross-country patterns:
  - AE average Okun coefficient larger in absolute value (more negative) than for EMs/EMDEs.
  - LAC coefficients somewhat larger in absolute value than EMs broadly.
- Institutional links and informality:
  - No clear linear relationship between Okun’s coefficient and WEF hiring and firing practices indicator; possible weak inverted-U pattern.
  - Strong negative relationship between Okun’s coefficient and informality: higher informality → weaker unemployment response to output; informality explains 36 percent of cross-country variation.
  - Redundancy costs and wage flexibility show bivariate relationships:
    - Higher redundancy costs linked to lower responsiveness of unemployment to cyclical fluctuations.
    - Higher wage flexibility dampens output fluctuations’ impact on unemployment.
- Outlier-robust regressions (Table 2):
  - Controlling for informality, most labor institution indicators not statistically significant except wage flexibility.
  - Economic magnitudes (spec.6): one standard deviation increase in informality increases Okun’s coefficient by 0.10 point; one standard deviation increase in wage flexibility increases coefficient by 0.03 point.
- Caveat:
  - Greater sensitivity of unemployment to growth is not unambiguously welfare-improving absent adequate unemployment insurance or social safety nets.

### Links between informality and labor market institutions (section V) — cross-country regression evidence
- Dependent variable:
  - Informality = informal employment / total non-agricultural employment (ILO cross-section, latest available year; Uruguay supplement from IDB where ILO unavailable).
- Controls and institutional variables:
  - GDP per capita, years of schooling; institutional variables include hiring & firing practices, wage flexibility, redundancy costs (weeks of salary), minimum wage/value-added, dummy for third-party approval to dismiss one worker, dummy for fixed-term contracts prohibited for permanent tasks.
- Main regression findings (Table 3):
  - GDP per capita: strongly negative and significant (example: coefficient -11.87*** in spec.1).
  - Schooling (years): strongly negative and significant (example: -4.398*** in spec.1).
  - Redundancy costs: statistically significant positive association with informality (example: 0.393*** in spec.6).
  - Third-party dismissal approval dummy: positive and significant in several specifications (example: 8.391** in spec.2).
  - Wage flexibility, hiring and firing composite, fixed-term contract dummy, minimum wage/value-added: generally not robustly significant in multivariate regressions.
  - Bayesian model averaging (BMA): redundancy costs and dismissal approval have high posterior inclusion probabilities (PIP) among institutional variables.
- Interpretation:
  - L&Rs that increase expected costs or administrative burden of firing (higher redundancy costs, third-party approval) associated with higher informality.
  - Suggests an indirect channel from institutions → informality → weaker Okun response.

### Case studies using synthetic control (section VI)
- Method summary:
  - Synthetic control method; donor pools exclude countries with major reforms within ±10 years; predictors include pre-treatment averages of outcomes, GDP growth, informality, labor force growth, redundancy costs.
- Mexico (2012 labor reform — Nov 2012):
  - Reform features: lower hiring costs, new flexible contract types, simplified dispute settlement, ceilings on unjustified dismissal compensation, emphasis on productivity/skills.
  - OECD indicators: relaxation for temporary contracts but overall employment protection relatively unchanged.
  - Synthetic control results (sample 2001–2017): no significant favorable impact on unemployment rate or employment rate; differences small; placebo tests confirm non-significance.
  - Interpretation: muted impact possibly due to limited scope and largely unchanged overall employment protection.
- Chile (2001 labor reform):
  - Reform features: increased firing costs for unjustified or anti-union dismissals, increased penalties, higher nightly overtime costs, restrictions on replacing striking workers, gradual reduction in work week.
  - Context: modest GDP growth and elevated unemployment.
  - Synthetic control results (sample 1991–2006): modest adverse effects on employment rate (synthetic employment rate over 2 percentage points higher by end of period); unemployment differences ≤ 1.2 percentage points; placebo tests generally do not show significant impact (employment rate p-value significant only in initial year).
  - Interpretation: limited/significant deleterious effects not strongly evidenced.
- Uruguay (2005 labor reform — Decree 105/05 expanding sectoral wage councils):
  - Reform effect: private sector collective bargaining coverage rose from ~28 percent in 2000 to over 97 percent in 2005.
  - Context: post-2002 recovery, single-digit inflation, favorable external conditions.
  - Synthetic control results (sample 1995–2010): unemployment rates fell and employment rates increased relative to synthetic control; unemployment effects small; employment rate effects marginally significant at longer horizons; placebo tests: unemployment effects insignificant, employment effects marginally significant.
  - Interpretation: strengthened coordination through collective bargaining possibly supported wage moderation and employment gains in a favorable growth context.
- Overall lesson:
  - Reform outcomes depend critically on reform design, scope, and macroeconomic context; similar reforms can have muted, adverse, or positive effects.

### Conclusions and policy recommendations
- Summary conclusions:
  - Employment growth strong in the 2000s but moderated post-terms-of-trade boom; Latin American labor market performance broadly comparable to other EMDEs.
  - Informality plays a central countercyclical role, dampening unemployment responses to output fluctuations; unemployment rate alone insufficient as a slack indicator in Latin America.
  - Main determinants of informality: GDP per capita and education.
  - Institutional contributors to higher informality: higher redundancy costs and cumbersome dismissal regulations (third-party approval requirements).
- Policy recommendations (consistent with Duval and Loungani, 2018):
  - Reduce expected costs of firing procedures and make dismissal procedures more transparent, predictable, and less administratively burdensome to tackle informality and improve labor market functioning.
  - Simultaneously build up unemployment insurance and other worker protections to preserve welfare when increasing labor market flexibility.
- Case-evidence caveat:
  - Mexico 2012: limited impact—likely due to limited scope and unchanged overall protection.
  - Chile 2001: modest adverse effect on employment rate but not robustly significant on unemployment.
  - Uruguay 2005: positive employment effects consistent with coordination improvements in favorable growth context.
- Final takeaway:
  - Addressing specific binding elements of EPL (redundancy costs, dismissal procedures) while strengthening social safety nets offers a promising route to reduce informality and improve labor market outcomes; success depends on precise institutional design and macroeconomic context.

*Source: wpiea2019055 — Section V and related sections from the IMF working paper PDF provided.*

### Section V, we use simple cross-country regressions to explore the effect of labor market

### Section V, we use simple cross-country regressions to explore the effect of labor market institutions such as employment protection legislation on informality.

### Methods
- Use of simple cross-country regressions to explore the effect of labor market institutions (example: employment protection legislation) on informality.
- Final part of the paper examines the impact of a small number of labor market reforms in Latin America since 2000 using the synthetic control method to build appropriate counterfactuals (Abadie and Gardeazabal, 2003).
- Appendix A contains a detailed list of available data, including sources, country and time coverage and a comparative analysis.

### Reforms studied (synthetic control)
- Countries and reform years analyzed:
  - Chile (2001)
  - Mexico (2012)
  - Uruguay (2005)
- Results: mixed findings regarding the impact of structural changes in labor market institutions on key outcome variables.

### Data on labor market institutions
- Minimum wage data:
  - Available from several reliable sources for a large set of countries.
  - Main difficulty: choosing an adequate denominator to scale the data to gauge how binding the minimum wage is.
- Employment Protection Legislation (EPL) data:
  - ‘Hard’ Laws and Regulations data (L&Rs) available from the World Bank (WB), the OECD and the International Labor Organization (ILO).
  - Perceptions-based indicators (PBIs) constructed by the World Economic Forum (WEF) from its executive survey.
- The paper focuses on minimum wage policy and employment protection legislations and only very briefly touches upon other aspects such as payroll taxes, employment benefits, unemployment insurance and collective bargaining.

### Perceptions-based indicators (PBIs) versus Laws & Regulations (L&Rs)
- The relationship between different L&R datasets, and between PBIs and L&Rs, is studied.
- Key empirical observations:
  - Complicated hiring and firing procedures have the strongest influence on perceptions of labor market rigidity, more than explicit redundancy costs.
  - Other EPL dimensions measured by the OECD, ILO and WB datasets (such as notice periods) do not impact perceptions at all.
- Recommendation from findings:
  - Combining perceptions-based indicators with ‘hard’ measures of laws and regulations is likely to provide the most comprehensive view of labor markets.
  - Care should be taken when using aggregate indexes for either PBIs or L&Rs, since aggregates may mask the role of individual components.

### Key findings on Latin American labor market rigidities
- Evidence suggests Latin American labor markets exhibit noticeable rigidities in some key dimensions:
  - Redundancy costs are higher than in advanced economies (AEs) or other EMDEs.
  - Permanent contracts are mandatory for permanent tasks in many countries.
  - Dismissal of even 1 worker often requires third-party approval.

*Source: wpiea2019055 - Section V, IMF working paper content provided.*

### references include the review of the impact of labor market institutions in EMDEs by

### wpiea2019055 - references include the review of the impact of labor market institutions in EMDEs by

### Stylized facts: labor markets in Latin America
- Labor markets are highly informal, marked by low labor productivity and a relatively rigid regulatory environment, creating a duality between well-protected formal sector jobs and unprotected informal sector jobs.
- Employment/unemployment
  - Total employment expanded at over 2 percent per year in the period prior to the financial crisis.
  - Argentina and Brazil recorded negative employment growth for at least one year after 2012.
  - Mexico, Peru: very low and relatively stable unemployment rates.
- Labor productivity
  - Median output per worker in South America in 2017 was about 30 percent of the level for the median advanced economy (20 percent in Central America).
  - Productivity growth volatile: all countries except Peru and Chile achieved positive productivity growth in only two out of the four periods considered.
- Informality
  - Informality accounts for over 50 percent of total employment in Latin America.
  - Informality in South and Central America is broadly comparable to other EMDEs; AEs have significantly lower informality.
  - Large heterogeneity within Latin America: informality ranges from around 30 to 70 percent among the largest economies.
  - Peru and Mexico (and Argentina to some degree) have higher informality than expected for their level of development.
  - Countries with lowest and most stable unemployment rates (Peru and Mexico) have the highest informality relative to development level, suggesting formal ↔ informal acts as a substitute margin for employment ↔ unemployment.
- Labor market institutions
  - WEF perceptions: wage flexibility similar across Latin America and other groups; hiring and firing practices in South America perceived as substantially more rigid.
  - Aggregate OECD and ILO indices do not show South America as having stronger de jure employment protection; possible explanations: de facto implementation issues or index aggregation hiding specific binding subcomponents.
  - Main drivers of perceived rigidity: cumbersome hiring and especially firing procedures rather than severance pay per se.
  - Specific indicators: redundancy costs higher than in AEs or other EMDEs; permanent contracts mandatory for permanent tasks in many countries; dismissal of even 1 worker often requires third-party approval.
  - Ratio of minimum wage to value added: Central America stands out as having a very high ratio; no clear evidence that minimum wage is more binding in South America than elsewhere.

### Decomposing unemployment dynamics (sections II–III)
- Decomposition framework
  - Unemployment changes decomposed into labor demand (output + labor productivity) and labor supply (participation rate + working-age population).
  - Alternative decomposition highlights formal employment and labor formality (formal/total employment ratio) as margins.
- Key patterns across subperiods (1997–2000; 2000–07; 2007–11; 2011–17)
  - Late 1990s: unemployment rose as supply outstripped demand.
  - 2000–11 commodity boom: unemployment fell as labor demand outpaced supply.
  - GFC: limited impact on unemployment in sample countries.
  - Post-2012: unemployment broadly stable except Brazil and Argentina where unemployment rose due to sharp demand drops.
- Margins of adjustment
  - Working-age population growth largely stable.
  - Labor force participation rate an active adjustment margin, mitigating unemployment fluctuations (esp. Chile, Colombia, Peru during boom years).
  - Labor productivity growth reduces labor demand by allowing same output with fewer workers; productivity fluctuations can reflect labor hoarding or changes in informality (informal sector lower productivity).
  - Labor formality/informality played a crucial countercyclical role:
    - Example: Colombia—late 1990s informality rose limiting unemployment increase; 2000s boom saw informality fall; GFC saw informality rise again.
    - Brazil and Mexico show less adjustment along informality margin in recent periods: Brazil informality fell over 2012–17 even as unemployment increased; Mexico informality increased since early 2000s despite low/falling unemployment.
- Implication
  - Changes in informality and participation helped limit unemployment rises in downturns; therefore, to assess labor market slack in Latin America, combine formal employment growth, informality rate, and unemployment.

### Revisiting Okun’s law (section IV)
- Estimation approach
  - Heterogenous panel with common correlated effects (CCE) estimator (Pesaran, 2006) for 127 countries (annual data 1990–2017; unbalanced panel; countries with <20 years excluded).
  - Specification allows country-specific slopes and accounts for cross-sectional common factors.
  - Alternative gap specification via HP filter (smoothing parameter 6.25) yields quantitatively similar results.
- Baseline estimates (Table 1)
  - Specification 1 (no lags): contemporaneous coefficient ߚ ≈ -0.125*** (standard error 0.0152).
  - Including up to four lags (specifications 2–5) leaves contemporaneous coefficient around -0.115 to -0.126; only the first lag of ΔGDP is robustly significant in multi-lag specifications (ΔGDPt-1 ≈ -0.0330*** in spec. 2).
  - Observations: 3,399 (spec.1) down to 3,307 (spec.5); Countries: 127.
- Cross-country patterns
  - Estimates for AE average larger in absolute value (more negative) than for EMs/EMDEs.
  - LAC coefficients somewhat larger in absolute value than EMs broadly.
- Okun’s coefficient and institutions
  - No clear linear relationship between Okun’s coefficient and WEF hiring and firing practices indicator; possible weak inverted-U pattern but tenuous.
  - Strong negative relationship between Okun’s coefficient and informality: higher informality → weaker unemployment response to output (informality explains 36 percent of cross-country variation).
  - Redundancy costs and wage flexibility show bivariate relationships:
    - Higher redundancy costs linked to lower responsiveness of unemployment to cyclical fluctuations.
    - Higher wage flexibility dampens output fluctuations’ impact on unemployment.
  - Outlier-robust regressions (Table 2) controlling for informality show most labor institution indicators not statistically significant, except wage flexibility.
    - Economic magnitudes (spec.6): one standard deviation increase in informality increases Okun’s coefficient by 0.10 point; one standard deviation increase in wage flexibility increases coefficient by 0.03 point.
- Caveat
  - Greater sensitivity of unemployment to growth is not unambiguously welfare-improving absent adequate unemployment insurance or social safety nets.

### Links between informality and labor market institutions (section V)
- Cross-country regressions of informality (informal employment / total non-agricultural employment) on institutions plus controls (GDP per capita, years of schooling)
  - Informality measure from ILO (latest available year per country); supplementary Uruguay IDB measure used where ILO unavailable.
  - Institutional variables: hiring & firing practices, wage flexibility, redundancy costs (weeks of salary), minimum wage/value-added, dummy for third-party approval to dismiss one worker, dummy for fixed-term contracts prohibited for permanent tasks.
- Main regression results (Table 3)
  - GDP per capita: strongly negative and significant (e.g., coefficient -11.87*** in spec.1; variations across specs but always strongly significant).
  - Schooling (years): strongly negative and significant (e.g., -4.398*** in spec.1).
  - Redundancy costs: statistically significant positive association with informality (e.g., 0.393*** in spec.6).
  - Third-party dismissal approval dummy: positive and significant in several specifications (e.g., 8.391** in spec.2).
  - Wage flexibility, hiring and firing composite, fixed-term contract dummy, minimum wage/value-added: generally not robustly significant in multivariate regressions.
  - Bayesian model averaging (BMA) confirms redundancy costs and dismissal approval have high posterior inclusion probabilities (PIP) among institutional variables.
- Interpretation
  - L&Rs aimed to protect employment (higher redundancy costs, cumbersome dismissal procedures) associated with higher informality, suggesting an indirect effect of institutions on Okun’s coefficient via informality.
  - Policy implication: reforms reducing expected cost and administrative burden of firing (making procedures transparent and predictable) could reduce informality and increase labor market responsiveness; simultaneous development of unemployment insurance and benefits recommended.

### Case studies: synthetic control assessments (section VI)
- Method: synthetic control method to construct counterfactuals; donor pools exclude countries that implemented major labor reforms within a 10-year window around treatment; predictors include pre-treatment averages of outcomes, GDP growth, informality, labor force growth, redundancy costs.
- Mexico 2012 labor reform
  - Reform (Nov 2012): lower hiring costs, stimulate formal employment, new flexible contract types, simplified dispute settlement, ceilings on unjustified dismissal compensation, productivity/skills emphasized over seniority.
  - OECD indicators: relaxation for temporary contracts but overall employment protection remained relatively unchanged.
  - Synthetic control results (sample 2001–2017): no significant favorable impact on unemployment rate or employment rate; differences small; placebo tests confirm non-significance.
  - Interpretation: muted impact possibly due to limited scope of reform and unchanged overall employment protection.
- Chile 2001 labor reform
  - Reform increased rigidities: higher firing costs for unjustified or anti-union dismissals, increased penalties, higher nightly overtime costs, restrictions on replacing striking workers, gradual reduction in work week.
  - Context: modest GDP growth and elevated unemployment.
  - Synthetic control results (sample 1991–2006): modest adverse effects on employment rate (synthetic control employment rate over 2 percentage points higher by end of period); unemployment differences ≤ 1.2 percentage points; placebo tests generally do not show significant impact (employment rate p-value significant only in initial year).
  - Interpretation: limited/significant deleterious effects not strongly evidenced.
- Uruguay 2005 labor reform
  - Reinstatement/expansion of sectoral wage councils (Decree 105/05) restoring broad private sector collective bargaining coverage (share covered rose from ~28 percent in 2000 to over 97 percent in 2005).
  - Context: post-2002 recovery, single-digit inflation, favorable external conditions.
  - Synthetic control results (sample 1995–2010): unemployment rates fell and employment rates increased relative to synthetic control; unemployment effects small; employment rate effects marginally significant at longer horizons; placebo tests: unemployment effects insignificant, employment effects marginally significant.
  - Interpretation: coordination via collective bargaining possibly improved wage moderation and stability, supporting employment gains in a favorable growth context.
- Overall case-study lesson
  - Outcomes of labor market reforms depend critically on reform design and economic context; similar reforms can have muted, adverse, or positive effects depending on scope and timing.

### Conclusion and policy implications
- Summary findings
  - Employment growth strong in 2000s but moderated post-terms-of-trade boom; Latin American labor market performance broadly comparable to other EMDEs.
  - Informality plays a central countercyclical role, dampening unemployment responses to output fluctuations; therefore unemployment rate alone is an insufficient labor market slack indicator in Latin America—combine unemployment, formal employment growth, and informality.
  - Key determinants of informality: GDP per capita (demand for formality) and education (supply of formality).
  - Institutional contributors to higher informality: higher redundancy costs and cumbersome dismissal regulations (third-party approval requirements).
- Policy recommendations (consistent with Duval and Loungani, 2018)
  - Reduce expected costs of firing procedures, make dismissal procedures more transparent, predictable and less administratively burdensome to tackle informality and improve labor market functioning.
  - Simultaneously build up unemployment insurance and other worker protections to preserve welfare when increasing labor market flexibility.
- Evidence from cases
  - Mexico 2012: limited impact on unemployment/employment—likely due to limited scope and unchanged overall protection.
  - Chile 2001: modest adverse effect on employment rate but not robustly significant on unemployment.
  - Uruguay 2005: positive employment effects consistent with improved coordination via collective bargaining in a favorable growth environment.
- Final takeaway
  - Reform success depends on precise institutional design and macroeconomic context; addressing specific binding elements of EPL (redundancy costs, dismissal procedures) while strengthening social safety nets offers a promising route to reduce informality and improve labor market outcomes.

*Source: wpiea2019055 (PDF chapter/section).*

### References

### wpiea2019055 - References (Selected methodological and data highlights)

### Measures of employment protection and labor market institutions
- OECD distinguishes four categories: (i) labor market programs, (ii) strictness of employment protection, (iii) statutory minimum wages and (iv) collective bargaining coverage.
- World Bank distinguishes three categories: (i) entering employment and working conditions, (ii) statutory minimum wages and (iii) redundancy.
- The paper focuses on employment protection legislation (EPL) and minimum wages; excludes payroll taxes, collective bargaining issues, labor market programs, and broader working-conditions measures (leave, working hours).
- Three comprehensive EPL datasets for Latin America highlighted:
  - OECD “Indicators of Employment Protection Dataset”
  - ILO “EPLex Dataset”
  - World Bank “Doing Business” labor market appendix
- Note on aggregation: OECD and ILO provide aggregate indexes (weighted averages). Correlation among individual components is often low or negative; using subcomponents is preferable in most applications.
- Cautionary notes:
  - WEF and Fraser Institute index constructions include perception-based inputs and aggregations of existing data; methodological shortcomings and changes over time warrant caution (Aleksynska and Cazes, 2014; Vasquez and Porcnik, 2018).

### Dataset attributes and timeseries/country coverage (as presented)
- WB (Doing Business): Type of data — De jure for a specific type of firm; Timeseries coverage — 2014-18 (updated annually); Country coverage — 189 countries; LAC country coverage — 32 countries.
- OECD: Type of data — De jure as well as de facto implementation of laws and regulations; Timeseries coverage — 1985-2013/14; Country coverage — Up to 45 countries, depending on the year; LAC country coverage — Limited, except in 2014 (18 countries).
- ILO (EPLex): Type of data — Purely de jure; Timeseries coverage — 2009-2013; Country coverage — 70-93 countries depending on the year; LAC country coverage — 1-11 depending on the year.
- WEF: Type of data — Mix of survey based, statistics and de jure; Timeseries coverage — 2007-2017; Country coverage — 128-139 countries depending on the year; LAC country coverage — 20-24 depending on the year.

### Variable coverage and definitions across key datasets
- WB Doing Business (Redundancy Rules and Regulations) captures five dimensions for regular contracts: (i) procedural requirements, (ii) retraining/reassignment obligation, (iii) priority rules, (iv) notice period, (v) severance pay. Many variables are indicator variables. Coverage: over 190 countries.
- OECD Indicators of EPL (regular contracts focus) capture nine dimensions: (i) notification procedures, (ii) delay before notice can start, (iii) length of notice period, (iv) severance pay, (v) definition of justified or unfair dismissal, (vi) length of trial period, (vii) compensation following unfair dismissal, (viii) possibility of reinstatement following unfair dismissal, (ix) maximum time to make claim of unfair dismissal. Higher values indicate stricter employment protection. Latin America data availability noted via OECD/IDB (2014).
- ILO EPLex covers eight de jure dimensions for regular contracts: (i) valid grounds for dismissal, (ii) prohibited grounds for dismissals, (iii) probationary period, (iv) procedural notification requirements for dismissals, (v) notice periods, (vi) severance pay, (vii) redundancy pay, (viii) avenues for redress. Higher values indicate more labor protection. Timeseries 2009-2013.
- WEF Labor Market Efficiency index: constructed from ten sub-indexes (eight perception-based from executive survey, one from World Bank Doing Business, one outcome statistic). Perception-based sub-indexes listed in footnote: (i) cooperation in labor-employer relations, (ii) flexibility of wage determination, (iii) hiring and firing practices, (iv) effect of taxation on incentives to work, (v) pay and productivity, (vi) reliance on professional management, (vii) country capacity to retain talent, (viii) country capacity to attract talent. The WB sub-index is average redundancy costs in weeks of salary; the outcome is female participation in the labor force.
- Fraser Institute index: aggregate of six sub-indexes drawn from existing datasets (World Bank Doing Business, WEF, International Institute for Strategic Studies). Because it aggregates existing series, using underlying data is generally preferable; changes in methodology in underlying series make aggregate timeseries use particularly cautious.

### Correlations: ‘hard’ EPL measures vs perception-based WEF indexes and subcomponent correlations
- Pairwise correlations among aggregate indexes:
  - WEF Hiring and Firing Index — ILO EPLex Index: -0.3837* (star indicates significance at the 1 percent level)
  - WEF Hiring and Firing Index — OECD EPL Index: -0.4312* 
  - ILO EPLex Index — OECD EPL Index: 0.6322*
- Note: Higher values of ILO and OECD indexes indicate more protection/rigidity; higher WEF values indicate more flexibility.
- Correlations between WEF Hiring and Firing perceptions and World Bank EPL indicators (selected):
  - Fixed-term contracts prohibited for permanent tasks? -0.288*
  - Maximum length of a single fixed-term contract (months) 0.3385*
  - Maximum length of fixed-term contracts (months) 0.2831*
  - Dismissal due to redundancy allowed by law? 0.2589*
  - Third-party approval if one worker is dismissed? -0.2363*
  - Third-party notification if one worker is dismissed? -0.2556*
  - Retraining or reassignment? -0.2194*
  - Priority rules for redundancies? -0.1662*
  - Priority rules for reemployment? -0.2235*
  - Severance pay for redundancy dismissal (weeks of salary) -0.1325*
  - Notice period for redundancy dismissal (weeks of salary) -0.0330
- Correlations between WEF Hiring and Firing perceptions and OECD EPL subcomponents (selected):
  - Notification procedures -0.2779
  - Delay involved before notice can start -0.1202
  - Length of the notice period at 9 months tenure -0.0674
  - Length of the notice period at 4 years tenure 0.0455
  - Length of the notice period at 20 years tenure -0.1413
  - Severance pay at 9 months tenure -0.0965
  - Severance pay at 4 years tenure -0.0574
  - Severance pay at 20 years tenure -0.0679
  - Definition of justified or unfair dismissal -0.4155*
  - Length of trial period -0.0535
  - Compensation following unfair dismissal -0.3295*
  - Possibility of reinstatement following unfair dismissal -0.1816*
  - Maximum time to make a claim of unfair dismissal -0.0158
- Correlations between WEF Hiring and Firing perceptions and ILO EPLex sub-indicators (selected):
  - "Valid grounds for dismissals" -0.3772*
  - "Prohibited grounds for dismissals" -0.1446*
  - "Maximum probationary (trial) period" -0.1399
  - "Proceedural requirements for dismissals" -0.2362*
  - "Notice periods" -0.0785
  - "Severance pay" 0.0588
  - "Redundancy pay" 0.0125
  - "Redress" -0.2796*
- Note: A star indicates significance at the 1 percent level.

### Data sources for minimum wages and other labor market variables
- Minimum wages:
  - National minimum wage data are broadly available from national authorities.
  - Cross-country comparisons require a denominator to gauge bindingness: average wages, median wages, or value-added measures are used.
  - OECD constructs minimum-wage-to-mean and minimum-wage-to-median measures for a limited set of Latin American countries (Chile, Colombia, Costa Rica and Mexico).
  - World Bank constructs minimum-wage-to-value-added as part of Doing Business.
  - IDB constructs minimum-wage-to-average-monthly-wage-in-main-occupation measures for Latin American countries.
  - The paper presents all three measures in the stylized facts section.
- Employment, Unemployment, Labor Force Participation, Working Age Population:
  - Source: ILO
- Informality:
  - Broadest coverage: cross-sectional data for 2016 from ILO “Women and Men in the Informal Economy: A Statistical Picture” (2018).
  - Timeseries on informality in Latin America: IDB’s SIMS database.

### Appendix B — Decomposing changes in unemployment (key identities and definitions)
- Notation and identity highlights (as presented):
  - U is the number of unemployed workers.
  - u is the unemployment rate.
  - LF is the labor force.
  - L is the number of persons employed.
  - F is the number of formal employees (notation in source: ܮி).
  - Y is output (notation in source: ܻ).
  - ρ is the labor force participation rate defined as LF/WAP.
  - WAP is the working age population (notation in source: ܲܣܹ).
- Two equivalent decomposition expressions are given in the source in log-difference form (symbols preserved as in source):
  - ݈݊ = ... ln(ݑ) ... (first expression as in source)
  - Alternatively, ݑൎെln( ... ) ... (second expression as in source)
- The appendix presents the algebra used to decompose changes in the number of unemployed workers into contributions from changes in the unemployment rate, labor force participation, employment, formal employment, output, and working age population (exact formulae and symbolic forms preserved in the source).

*Source: wpiea2019055 - References (excerpted methodological appendices and dataset notes from the provided PDF).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019055.pdf_