## Section 3 — Regulatory environment in Latin America: EPL, MW, payroll taxes, UI, and collective bargaining

## Source details

**Canonical URL:** [Section 3 — Regulatory environment in Latin America: EPL, MW, payroll taxes, UI, and collective bargaining](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019155.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wpiea2019155.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019155.pdf.json)

---

### Overview of labor markets in Latin America
- Average overall unemployment rate in the region: 8.2%, women: 9.9%, men: 7.1%.
- Average unemployment in Latin America is almost 3 percentage points higher than in the U.S. and Canada.
- U.S. current unemployment referenced as 4.6%.
- Country-level unemployment examples:
  - Brazil: overall 12.9%; women 15.1%; men 11.2%.
  - Cuba: overall 2.6%; women 3%; men 2.3%.
  - Guatemala and Cuba have unemployment rates just above 2%.
- Many low-unemployment countries likely have high informality, implying under-employment or hidden unemployment.
- Countries with relatively lower informality (Brazil, Costa Rica, Argentina, Venezuela) have higher unemployment rates.
- Colombia noted as an anomaly: high unemployment and high shares of informal sector workers.

### Informality and its measurement
- Definition components for informal workers include:
  - workers hired by employers that do not comply with labor legislation;
  - workers without written contracts or with contracts that do not provide mandatory minimum wages or benefits;
  - workers employed by entities/quasi-corporations without formal accounts;
  - workers in family businesses, domestic workers, and the self-employed;
  - workers employed by formal-sector employers through intermediaries or ad-hoc arrangements that do not comply with labor legislation.
- Average shares of informality in LAC:
  - Women in informal sector: 55.8%.
  - Men in informal sector: 51.2%.
- Highest national shares:
  - Honduras: women 78.5%; men 73%.
  - Bolivia, Peru, Paraguay, Nicaragua: shares above 60%.
- Brazil (lowest shares in region): female 40.3%; male 39.2%.
- Implications:
  - Employers may avoid costs associated with labor market regulations.
  - Informal firms tend to remain small and cannot take advantage of economies of scale, contributing to lower productivity.

### Labor productivity in Latin America
- Measurement: output per worker (GDP in 2011 constant U.S. dollars, PPP-adjusted).
- Country-level output per worker (constant 2011 international $ PPP):
  - Honduras: 9,841
  - Nicaragua: 11,544
  - Bolivia: 15,081
  - Guatemala: 18,702
  - El Salvador: 18,726
  - Paraguay: 19,002
  - Ecuador: 21,912
  - Peru: 22,181
  - Colombia: 26,400
  - LAC Average: 28,629
  - Venezuela: 29,555
  - Brazil: 31,681
  - Dominican Republic: 32,253
  - Costa Rica: 36,417
  - Mexico: 38,390
  - Uruguay: 42,787
  - Argentina: 45,109
  - Panama: 47,143
  - Chile: 48,603
  - Canada: 89,682
  - United States: 119,208
- Relative productivity:
  - Chile is highest in region but only 40% of U.S. productivity and 54% of Canada’s.
  - Regional averages: 24% of U.S. labor productivity; 32% of Canada.
  - Central American countries (large informality) around 10% of U.S./Canada productivity.
- Reasons for lower productivity:
  - High shares of informal employment;
  - Informal sector employs less skilled workers, is less capital-intensive, and produces at smaller scale.

### Inequality trends
- Gini coefficient changes:
  - Fell by 0.4 from 2002 to 2008.
  - Fell by an additional 0.27 since 2008.
  - Gini fell from 0.56 in 1996 to 0.52 in 2002 and to 0.49 in 2016.
- Labor income Gini (Messina and Silva, 2018): declines in all countries except no change in Honduras and an increase of 0.03 in Costa Rica.
- Earnings distribution changes since 2002:
  - Lower tail rose by 50%, middle rose by 32%, upper tail rose by 15%.
  - Gap between college-educated and primary schooling fell from 330% to 240% since 2003.
- Country-level Gini reductions (reported magnitudes):
  - Brazil: close to 15.
  - Argentina, Mexico, Chile, Bolivia: 7.
  - Nicaragua, Panama, Honduras, El Salvador: close to 5.
  - Peru, Uruguay, Ecuador: less than 5.

### Roles and trade-offs of labor market institutions (LMI)
- Purposes:
  - Employment Protection Legislation (EPL): protect workers against unfair dismissals.
  - Minimum Wages (MW): ensure workers are paid a ‘living’ wage.
  - Payroll taxes: finance health insurance, pensions, workers’ compensation.
  - Unemployment Insurance (UI): provide income support for those who lose their jobs.
- Alternative rationales:
  - EPL can act as a commitment device to encourage on-the-job training (OJT).
  - UI may prolong job searches but improve match quality.
- Criticisms:
  - EPL, MW, payroll taxes, and collective bargaining may increase employer costs, reduce demand for formal labor, increase unemployment, and increase informality.

### Employment Protection Legislation (EPL): evidence and mechanisms
- EPL index observations:
  - Most Latin American countries: EPL index for regular contracts close to OECD average of 2.28.
  - EPL index higher than OECD average in Venezuela, Argentina, Mexico, Honduras, and Colombia.
  - Costs of dismissing temporary workers much higher in Latin America compared to the U.S., Canada, and the OECD.
- Time trends (examples 2008–2014):
  - OECD: temporary contracts increased from 2.06 to 2.08; permanent contracts declined from 2.36 to 2.28.
  - Mexico: regular contracts declined from 2.71 to 2.62; temporary contracts declined from 4.08 to 2.29.
  - Brazil: regular contracts increased from 1.25 to 1.75.
- Theoretical effects:
  - EPL reduces dismissals but increases expected hiring costs, reducing new hires; net employment effect ambiguous.
  - Productivity impact ambiguous:
    - Positive: greater OJT and more selective hiring.
    - Negative: retention of less productive workers.
- Micro-evidence highlights:
  - Stricter EPL reduces turnover; relaxing EPL increases transitions both into and out of employment.
  - Mixed effects on employment levels across studies.
  - Net conclusions in excerpt:
    - Effects on turnover are large and negative when EPL is stricter.
    - Effects on productivity are consistently negative but small.
    - Effects on employment are ambiguous; EPL can sometimes increase employment.

### Minimum wages (MW): prevalence, rationale, and empirical evidence
- Ratios of MW to mean wage (selected values):
  - Honduras: 0.99
  - Guatemala: 0.97
  - Mexico: 0.29
  - Peru: 0.16
  - LAC average: 0.6
  - Country series (Figure 11): Peru 0.16; Mexico 0.29; Uruguay 0.39; Brazil 0.43; Nicaragua 0.47; Chile 0.47; Bolivia 0.53; Costa Rica 0.56; Panama 0.57; Colombia 0.57; LAC average 0.60; El Salvador 0.65; Venezuela 0.67; Argentina 0.68; Ecuador 0.77; Paraguay 0.79; Dominican Republic 0.87; Guatemala 0.97; Honduras 0.99.
- Ratios of MW to median wage (selected values, Figure 12):
  - Costa Rica: 0.03
  - Mexico: 0.34
  - Uruguay: 0.45
  - Ecuador: 0.49
  - Bolivia: 0.51
  - Peru: 0.61
  - LAC average: 0.67
  - Brazil: 0.70
  - Venezuela: 0.72
  - Panama: 0.73
  - Argentina: 0.87
- Rationale:
  - Increase earnings and living conditions;
  - Counteract employer monopsony and act as efficiency wage.
- Wage effects:
  - Direct increases for those below MW and indirect “numeraire” and “lighthouse” effects observed in multiple studies.
  - Evidence of positive wage effects between 0.7–0.9 of the MW in some studies.
- Employment effects (micro-data: mixed):
  - Examples: 9% MW increase in Colombia reduced unemployment probability by 1.4%; South Africa and Turkey studies found no employment impact; Indonesia 10% MW increase reduced employment by 0.34%; U.S. meta-study finds most results hover around zero.
- Synthesis:
  - MWs clearly increase wages and reduce inequality.
  - Evidence on employment tends to show small or no effects.

### Payroll taxes: levels, transmission, and impacts
- Reported payroll tax rates:
  - Canada: 65%
  - OECD average: 41%
  - LAC average: 25%
  - Brazil: 91%
  - Argentina: 48%
  - Mexico: 39%
  - Costa Rica: 32%
  - Colombia: 31%
  - Paraguay: 26%
  - Honduras: 13%
  - Dominican Republic: 15%
  - Guatemala: 17%
  - El Salvador: 17%
- Theoretical effects:
  - Finance valuable services but increase labor costs, can reduce formal employment and increase informality.
  - If workers fully value services, cost may be passed to wages (full pass-through) with no employment effect; pass-through may fail if services not valued or MW bind.
- Empirical evidence in Latin America:
  - Heckman and Pages (2004): 36% pass-through and 4.5% employment reduction.
  - Colombia (Kugler and Kugler, 2011): 10% payroll tax increase lowered formal wages by 1.4%–2.3% (pass-through 14%–23%) and reduced formal employment 4%–5%.
  - 2013 Colombian reform reduced payroll taxes for low earners and some self-employed; formal employment increased across multiple data sources.
  - Argentina (Cruces et al., 2010): 55% shifting to wages but no employment effect.
  - Chile (Gruber, 2001): full pass-through and no employment effect for pension tax reduction.
- Productivity and misallocation:
  - Payroll taxes can bias firms toward smaller, less capital-intensive, less productive firms, reducing productivity.
  - Evidence suggests reductions in payroll taxes in the region should have contributed to productivity increases.

### Unemployment Insurance (UI): coverage, replacement rates, and effects
- UI coverage in Latin America: available only in Venezuela, Uruguay, Chile, Argentina and Brazil.
- Gross replacement rates reported:
  - Venezuela: 30%
  - Uruguay: 25%
  - Chile: 23%
  - Argentina: 22%
  - Brazil: 15%
  - U.S.: 57%
  - Canada: 59%
- Theoretical effects:
  - UI can improve match quality and post-unemployment wages but may prolong unemployment spells via moral hazard.
- Empirical evidence:
  - Argentina (2006 reform): UI lengthened unemployment spells and had positive impact on post-unemployment wages.
  - Brazil: UI prolongs job search, increases self-employment and entrepreneurship, no wage effect.
  - Chile: no effect on employment or wages.
  - Central Europe evidence: mixed; some countries show prolonged unemployment spells, others show higher exit rates from unemployment for UI recipients and reductions in poverty.
- Synthesis:
  - More generous UI in Latin America and Central Europe generally prolongs unemployment spells and has either a positive effect or no effect on wages.

### Unionization and collective bargaining: coverage and labor-market effects
- Unionization rates (Figure 15 selected values):
  - Bolivia: 39.1
  - Uruguay: 30.1
  - Canada: 28.4
  - LAC Average: 16.3
  - U.S.: 10.3
  - Selected country series: Nicaragua 5.3; Peru 5.7; Paraguay 6.7; Colombia 9.5; Dominican Republic 11.0; Panama 11.9; Mexico 12.5; Brazil 18.9; El Salvador 19.0; Costa Rica 19.4; Chile 19.6; Argentina 27.7; Uruguay 30.1; Bolivia 39.1.
- Collective bargaining coverage rates (Figure 16 selected values):
  - Uruguay: 94.6
  - Cuba: 81.4
  - Brazil: 70.5
  - Argentina: 51.8
  - LAC Average: 24.8
  - Canada: 30.3
  - U.S.: 11.5
  - Low coverage examples: Panama 1.0; Venezuela 2.5; Peru 4.8; El Salvador 5.0; Honduras 5.7; Nicaragua 7.6; Mexico 9.9; Costa Rica 10.6.
- Key contrasts:
  - In Uruguay and Argentina coverage far exceeds unionization rates due to sector/national bargaining.
  - In Canada and U.S. coverage and unionization are similar because bargaining is often at establishment level.
- Theoretical and empirical effects:
  - Unions raise wages and compress wage differentials for unionized workers.
  - In Latin America, unionized workers tend to be more educated; unionization is likely to increase overall wage inequality.
  - Unions can improve morale/productivity via “voice” but may increase employer costs and reduce employment.
- Micro-evidence:
  - Uruguay (Cassoni et al., 2005): shift from centralized to enterprise-level bargaining led to positive wage impacts, greater investments, reduced profits, and no employment effect.
  - Brazil: union density has an inverted U-shape effect on employment and productivity; at current Brazilian union density effects positive on employment and productivity but negative on profitability.
  - Peru: unions reduce profits and have no effect on productivity.
- Synthesis:
  - Unions and collective bargaining raise wages and affect investment and profits; employment impacts are mixed and context-dependent.
  - Net implication: effects positive if unionization moderate, nil or negative if very low or very high.

### Demographic factors — education, age, and gender
- Education and experience:
  - Average years of education for adults increased from 5.8 to 8.3 between 1990 and 2010.
  - Fraction of high school graduates rose from 21% to 36% between 1990 and 2010.
  - Share with primary education declined from 54% to 39% between 1990 and 2010.
  - Share of college-educated doubled between 1990 and 2010.
  - Average years of schooling: U.S. overall 11 years; Latin America 5 years on average.
  - Experience (age) has been increasing; population is aging.
- Implications:
  - More educated and older workers more likely formal and employed; contributed to declines in informality and unemployment.
  - Effects on inequality ambiguous—could reduce returns to skill or increase within-group inequality.
- Returns and earnings distribution:
  - Since 2000s, earnings of unskilled workers grew faster than skilled.
  - Return to skill declined in all countries except Costa Rica and the Dominican Republic.
  - Experience premium (21–30 years vs 0–5 years) declined from 56% in 1993 to 33% in 2013.
  - Lopez-Calva and Lustig (2010): 50% of decline in earnings inequality explained by narrowing skilled vs unskilled earnings.
- Gender:
  - Female labor force participation has grown over past two decades.
  - Gender wage gap levels:
    - 20% in Chile and Peru.
    - 15% in Brazil.
    - Between 5% and 10% in Uruguay, Dominican Republic, Mexico, Venezuela and Bolivia.
    - Between 5% and 10% (duplicate range in source) and positive but below 5% in Panama, Nicaragua, Guatemala and Colombia.
    - Women earn more on average than men in El Salvador, Costa Rica, Argentina, Ecuador and Paraguay.
  - Time pattern: gap stable 1993–1999, large decline 1999–2003, smaller decline since 2003.
  - Conclusion: women’s entry into labor force contributed to decline in inequality.

### Institutional factors — impacts on employment, informality, wages, inequality, and productivity
- Employment and informality:
  - Reduced EPL strictness and reduced payroll taxes likely boosted formal employment.
  - Centralized collective bargaining in countries with medium unionization may have increased formal employment but can create insider-outsider problems.
  - Higher minimum wages probably had little or small effect in reducing formal employment.
- Inequality:
  - Growth in skilled workers and female participation contributed to inequality decline.
  - Institutional actions that raise low-end earnings (higher MW, more generous UI) reduce inequality provided they do not lower employment.
  - Unionization in Latin America tends to increase wages for higher-income workers, limiting equalizing effects.
  - Declines in payroll taxes likely contributed to lower inequality if less-skilled workers benefited from ensuing wage increases.
- Productivity:
  - Evidence limited and mixed.
  - Unions may increase productivity for unionized jobs.
  - EPL appears to decrease productivity in the region.
  - No direct evidence in excerpt on MW and UI impacts on productivity in the region; other regions show potential positive impacts.
  - Payroll taxes and regulation-induced informality generate misallocation toward smaller, less capital-intensive, less productive firms.
  - Reductions in payroll taxes in past decades should have contributed to productivity increases.

### Conclusion — overall labor-market performance in Latin America
- Improvements:
  - Unemployment and informality have declined.
  - Inequality has decreased substantially.
- Remaining challenges:
  - Unemployment, informality, and inequality remain higher in Latin America than elsewhere.
  - Productivity has remained roughly constant and is lower in LAC compared to the U.S. and Canada.
- Drivers of improvements:
  - Demographic: more educated and older workforce; increased female labor force participation.
  - Institutional: reduced EPL strictness, lower payroll taxes, centralized bargaining in some countries, higher MW and more generous UI (with conditional employment effects).
- Net effect: combination of demographic and institutional changes contributed to declines in unemployment, informality, and inequality, with mixed evidence on productivity effects.

*Source: wpiea2019155 - Section 3 (IMF working paper PDF).*

### Section 3, I will describe the regulatory environment in Latin America with regards to EPL, MW,

### Section 3 — Regulatory environment in Latin America: EPL, MW, payroll taxes, UI, and collective bargaining

### Overview of labor markets in Latin America
- Average overall unemployment rate in the region is 8.2%, 9.9% for women and 7.1% for men.
- Average unemployment in Latin America is almost 3 percentage points higher than in the U.S. and Canada.
- U.S. current unemployment referenced as 4.6%.
- Country-level unemployment examples:
  - Brazil: overall 12.9%; women 15.1%; men 11.2%.
  - Cuba: overall 2.6%; women 3%; men 2.3%.
  - Guatemala and Cuba have unemployment rates just above 2%.
- Many countries with lower reported unemployment likely have high informality, implying under-employment or hidden unemployment.
- Countries with relatively lower informality (Brazil, Costa Rica, Argentina, Venezuela) have higher unemployment rates.
- Colombia noted as an anomaly: high unemployment and high shares of informal sector workers.

### Informality and its measurement
- Informal workers include:
  - workers hired by employers that do not comply with labor legislation;
  - workers without written contracts or with contracts that do not provide mandatory minimum wages or benefits;
  - workers employed by entities/quasi-corporations without formal accounts;
  - workers in family businesses, domestic workers, and the self-employed;
  - workers employed by formal-sector employers through intermediaries or ad-hoc arrangements that do not comply with labor legislation.
- Average shares of informality in LAC:
  - Women in informal sector: 55.8%.
  - Men in informal sector: 51.2%.
- Highest national shares:
  - Honduras: women 78.5%; men 73%.
  - Bolivia, Peru, Paraguay, Nicaragua: shares above 60%.
- Brazil (country with lowest shares of informal workers in region): female 40.3%; male 39.2%.
- Implications of informality:
  - Employers may avoid costs associated with labor market regulations.
  - Informal firms tend to remain small and cannot take advantage of economies of scale, contributing to lower productivity.

### Labor productivity in Latin America
- Labor productivity measured as output per worker (GDP in 2011 constant U.S. dollars, PPP-adjusted).
- Chile has the highest labor productivity in the region but is only 40% of productivity in the U.S. and 54% of that in Canada.
- Average productivity in the Latin American region is:
  - 24% of labor productivity in the U.S.
  - 32% of labor productivity in Canada.
- Central American countries (with large informality shares) have productivity around 10% of labor productivity in the U.S. and Canada.
- Reasons for lower productivity in the region:
  - High shares of informal employment;
  - Informal sector employs less skilled workers, is less capital-intensive, and produces at smaller scale.

### Inequality trends
- Latin American countries historically had high levels of wealth and income inequality.
- Since 2002 inequality has fallen substantially in most Latin American countries.
- Gini coefficient changes:
  - Fell by 0.4 from 2002 to 2008.
  - Fell by an additional 0.27 since 2008.
  - Gini fell from 0.56 in 1996 to 0.52 in 2002 and to 0.49 in 2016 (IDB, 2016).
- Labor income Gini for South and Central American countries (Messina and Silva, 2018):
  - Declines in labor income Gini in all countries except:
    - No change in Honduras.
    - Increase of 0.03 in Costa Rica.
- Some improvements likely associated with labor market legislation such as increases in minimum wages.

### Roles and trade-offs of labor market institutions (LMI)
- Purpose of LMIs:
  - Employment Protection Legislation (EPL): protect workers against unfair dismissals.
  - Minimum Wages (MW): ensure workers are paid a ‘living’ wage.
  - Payroll taxes: pay for health insurance, pensions, workers’ compensation.
  - Unemployment Insurance (UI): provide income support for those who lose their jobs.
- Alternative rationales:
  - EPL can create a commitment device to encourage on-the-job training (OJT) investment (Wasmer, 2006).
  - UI may prolong job searches but improve match quality between workers and employers.
- Criticisms and unintended consequences:
  - EPL, MW, payroll taxes, and collective bargaining may increase costs for employers, reduce demand for formal labor, increase unemployment, and increase informal employment.

### Employment Protection Legislation (EPL): empirical evidence and mechanisms
- EPL index:
  - In most Latin American countries the EPL index for regular contracts is close to the OECD average of 2.28.
  - EPL index higher than OECD average in Venezuela, Argentina, Mexico, Honduras, and Colombia.
  - Costs of dismissing temporary workers are much higher in Latin America compared to the U.S., Canada, and even compared to the OECD.
- Time trends (2008 to 2014 examples):
  - OECD EPL index: temporary contracts increased from 2.06 to 2.08; permanent contracts declined from 2.36 to 2.28.
  - Mexico: regular contracts declined from 2.71 to 2.62; temporary contracts declined from 4.08 to 2.29.
  - Brazil: regular contracts increased from 1.25 to 1.75.
- Theoretical effects of EPL:
  - EPL reduces dismissals but increases the expected cost of hiring, reducing new hires.
  - EPL reduces turnover by reducing both dismissals and hires.
  - Ambiguous overall impact on employment (depends on which effect dominates).
  - Ambiguous impact on productivity:
    - Positive channel: increased OJT raising productivity; greater selectivity in hiring more productive workers.
    - Negative channel: employers forced to retain less productive workers, lowering productivity.
- Micro-evidence on turnover and employment:
  - Consistent finding: increased turnover after EPL becomes less strict.
    - Kugler (1999, 2004): after severance payments fell in Colombia in 1991, both transitions unemployment→employment and employment→unemployment increased.
    - Saavedra and Torero (2004) and Paes de Barros and Corseuil (2004): increased separations after EPL reductions in Peru and Brazil.
    - Micco and Pages (2007): stricter EPL reduces turnover in more volatile sectors.
    - Autor, Kerr and Kugler (2007): stricter wrongful discharge laws in the U.S. reduced employment adjustments and firm entry.
  - Mixed evidence on EPL effects on employment levels:
    - Kugler (2004): found increased employment in Colombia after EPL became less strict.
    - Besley and Burgess (2004): pro-worker labor market reforms reduced employment in India.
    - Ahmed and Pages (2009): stricter EPL reduces employment.
    - Micco and Pages (2007): no evidence of stricter EPL effect on average employment per plant.
    - Avdagic (2015): no effects of EPL on overall youth employment in Central Europe once controls included.
    - Autor, Kerr and Kugler (2007): evidence of increased employment after increased strictness of wrongful discharge laws in the U.S.
  - Net conclusion on EPL:
    - Effects on turnover are large and negative when EPL is stricter.
    - Effects on productivity are consistently negative but small in magnitude.
    - Effects on employment are ambiguous; EPL can sometimes increase employment.

### Minimum wages (MW): prevalence and potential binding effects
- Ratios of MW to mean wage (examples and interpretation):
  - Honduras: 0.99 (MW at 99% of mean wage).
  - Guatemala: 0.97 (MW at 97% of mean wage).
  - Mexico: 0.29 (MW at 29% of mean wage).
  - Peru: 0.16 (MW at 16% of mean wage).
  - Average ratio for all of LAC: 0.6 (MW likely binding in most countries).
- Ratios of MW to median wage (better measure of MW coverage given inequality):
  - Overall ratio of MW to median wage in LAC: 0.67.
  - Overall ratio of MW to mean wage in LAC: 0.6.
  - Country comparisons (MW to median vs MW to mean):
    - Argentina: 0.87 vs. 0.68.
    - Panama: 0.72 vs. 0.57.
    - Venezuela: 0.72 vs. 0.67.
    - Brazil: 0.7 vs. 0.43.
    - Peru: 0.61 vs. 0.16.
    - Uruguay: 0.45 vs. 0.39.
    - Mexico: (ratio truncated in source text).
- Implication: In many LAC countries the MW is likely binding, especially relative to the median wage, and may affect a large share of workers in the labor market.

*Italic source: wpiea2019155 - Section 3 (IMF working paper PDF).*

### 0.34 vs. 0.29). Thus, given that there are so many informal sector workers, comparing the MW to

### wpiea2019155 - 0.34 vs. 0.29). Thus, given that there are so many informal sector workers, comparing the MW to

### Minimum wages (MW): rationale, effects, and empirical evidence
- Rationale for MW:
  - Increase earnings and living conditions for workers.
  - Address inefficiently low employment generated by employer monopsony power; MW can improve efficiency by increasing employment.
  - Act as an efficiency wage to overcome information asymmetries (moral hazard and adverse selection), increasing productivity (references: Shapiro and Stiglitz, 1984; Akerlof and Yellen, 1986).
- Direct and indirect wage effects:
  - Direct: increase wages of those earning below the legal minimum.
  - Indirect: “numeraire effect” — wages above the minimum may increase at multiples of the MW because MW serves as reference.
  - Indirect: “lighthouse effect” — informal sector employers may pay more as MW signals a higher ‘living’ wage.
- Empirical findings on wage effects:
  - Maloney and Nunez (2004) and Kristensen and Cunningham (2006) find large positive effects for those between 0.7-0.9 of the MW and evidence of both lighthouse and numeraire effects.
  - Similar positive impacts documented for South Africa (Dinkelman and Ranchod, 2011; Bhorat, 2013), Thailand (Lathapipat and Poggi, 2016), Turkey (Papps, 2012), China (Fang and Ling, 2013; Lin and Yun, 2016).
- Empirical findings on employment effects (micro-data: mixed):
  - Maloney and Nunez (2004): a 9% increase in the MW in 1999 in Colombia reduces the probability of becoming unemployed by 1.4% (note: coincided with economic downturn).
  - Dinkelman and Ranchod (2011) and Bhorat et al. (2003): no impact on employment in South Africa.
  - Papps (2012): no effect on employment in Turkey.
  - Del Carpio et al. (2015): a 10% increase in the MW in Indonesia reduces employment by 0.34%.
  - Lathapipat and Poggi (2016) and del Carpio, Messina and Sanz-de-Galeano (2018): small negative effects on employment in Thailand; del Carpio, Messina and Sanz-de-Galeano (2018) find small effects using difference-in-difference across provinces.
  - U.S. meta-study (Wolfson and Belman, 2016): most results on MW impact on employment hover around zero.
- Synthesis:
  - MWs clearly increase wages along the distribution and reduce inequality.
  - Evidence on employment tends to show small or no effects.

### Payroll taxes: levels, transmission, and impacts
- Role and prevalence:
  - Employers and employees pay payroll taxes as a share of employees’ wages to finance health insurance, pensions, and sometimes training, childcare, workers’ compensation.
  - Mandatory for formal sector employers.
- Comparative levels (reported figures):
  - Canada payroll tax rate: 65%.
  - OECD average payroll tax rate: 41%.
  - LAC average payroll tax rate: 25%.
  - Brazil payroll tax rate: 91%.
  - Argentina payroll tax rate: 48%.
  - Mexico payroll tax rate: 39%.
  - Costa Rica payroll tax rate: 32%.
  - Colombia payroll tax rate: 31%.
  - Paraguay payroll tax rate: 26%.
  - Honduras payroll tax rate: 13%.
  - Dominican Republic payroll tax rate: 15%.
  - Guatemala payroll tax rate: 17%.
  - El Salvador payroll tax rate: 17%.
- Theoretical effects:
  - Payroll taxes can provide valuable services and overcome market failures (insurance).
  - They can increase labor costs for employers, reduce demand for formal employment, and increase informality.
  - If workers fully value services funded, costs may be passed on to workers as lower wages (full pass-through) with no effect on formal employment; full pass-through may fail if services not fully valued or if minimum wages bind.
- Empirical evidence for Latin America:
  - Heckman and Pages (2004): 36% pass-through and 4.5% reduction in employment in Latin America.
  - Kugler and Kugler (2011) for Colombia (post-1994 reform): a 10% increase in payroll taxes lowered formal wages by between 1.4% and 2.3% (pass-through between 14% and 23%) and reduced formal employment between 4% and 5%.
  - Kugler, Kugler and Herrera (2017) on 2013 reform: reduced payroll taxes for workers earning less than 10 MW and self-employed hiring more than 3 employees; formal employment increases across household surveys, social security records, and monthly manufacturing surveys.
  - Cruces et al. (2010) for Argentina: 55% shifting to wages but no effect on employment.
  - Gruber (2001) for Chile: full pass-through and no effect on employment for a pension tax reduction.
- Productivity and misallocation:
  - Payroll taxes can bias firms toward smaller, less capital-intensive, less productive firms — reducing productivity.
  - Levy (2008) discusses adverse effects on firm size and capital investments.
  - Eslava et al. (2009) show misallocation effects and reductions in productivity due to stricter labor market regulations in Colombia.
- Policy implication (implied from evidence):
  - Reductions in payroll taxes in the region during past decades should contribute toward increases in productivity.

### Unemployment Insurance (UI): coverage, replacement rates, and effects
- Coverage in Latin America:
  - UI available only in Venezuela, Uruguay, Chile, Argentina and Brazil.
- Reported gross replacement rates:
  - Venezuela: 30%.
  - Uruguay: 25%.
  - Chile: 23%.
  - Argentina: 22%.
  - Brazil: 15%.
  - U.S. replacement rate: 57%.
  - Canada replacement rate: 59%.
- Theoretical effects:
  - UI provides income to unemployed workers and can permit longer search for better skill matches, improving post-unemployment wages and productivity.
  - UI can also prolong unemployment spells and potentially reduce search effort via moral hazard, raising reservation wages without improving matches.
- Empirical evidence:
  - Gonzales-Rozada and Ruffo (2016) for Argentina (2006 reform expanding eligibility): UI lengthened unemployment spells and had a positive impact on post-unemployment wages.
  - Cunningham (2000) for Brazil: UI prolongs job search, increases self-employment and entrepreneurship, no effect on wages.
  - Huneeus (2012) for Chile: no effect on employment or wages.
  - Evidence for Central Europe: prolonged unemployment spells in Hungary, Bulgaria, Poland (Micklewright and Nagy, 1998; Kotzeva, 1996; Steiner and Kwisikowski, 1995; Adamchik, 1999); Romania shows higher likelihood of leaving unemployment for UI recipients (Earle and Pauna, 1998); UI reduces poverty by 4.1% in Hungary and 3.2% in Poland (Vodopivec et al., 2003).
- Synthesis:
  - More generous UI in Latin America and Central Europe generally prolongs unemployment spells and has either a positive effect or no effect on wages.

### Unionization and collective bargaining: coverage and labor-market effects
- Unionization levels (reported figures):
  - Bolivia unionization rate: 39.1%.
  - Uruguay unionization rate: 30.1%.
  - Canada unionization rate: 28.4%.
  - Regional average unionization rate in LAC: 16.3%.
  - U.S. unionization rate: 10.3%.
  - Several Southern Cone countries (Argentina, Chile, Brazil), plus Costa Rica and El Salvador, have unionization rates above the regional mean.
  - Andean and Central American countries typically have much lower unionization rates near 10% or lower.
- Collective bargaining coverage (reported figures and contrasts):
  - Uruguay collective bargaining coverage: 94.6%.
  - Cuba collective bargaining coverage: 81.4%.
  - Brazil collective bargaining coverage: 70.5%.
  - Argentina covers half of its workforce through collective bargaining.
  - Average collective bargaining coverage in LAC: around 24.8%.
  - Median collective bargaining coverage: only one tenth of all workers.
  - Coverage vs. unionization contrast: Uruguay and Argentina have coverage much higher than unionization rates; in Canada and the U.S. coverage and unionization rates are similar (Canada: 28.4% unionization vs. 30.3% coverage; U.S.: 10.3% vs. 11.5%) because bargaining in Canada and the U.S. typically occurs at establishment level, whereas in most of Latin America bargaining is at sectorial or national levels.
- Theoretical and empirical effects:
  - Unions typically raise wages and compress wage differentials for unionized workers.
  - In Latin America, unionized workers tend to be more educated; unionization is likely to increase overall wage inequality in the region (Kuhn and Marquez, 2005).
  - Unions can offer “voice,” potentially improving morale and productivity (Freeman and Medoff, 1984).
  - Unions may increase costs for employers and reduce employment.
- Micro-data evidence:
  - Cassoni et al. (2005) for Uruguay (panel of establishments 1988–1995) — movement from centralized to enterprise-level bargaining: positive impacts on wages and greater investments, reduction in profits, and no effect on employment; suggests movement toward less union coverage had positive impacts on wages and investments, not employment.
  - (Text ends mid-sentence regarding Menezes-Filho et al. (2005) for Brazilian manufacturing firms from 1988 — further results not included in supplied excerpt.)
- Synthesis:
  - Unions and collective bargaining in the region raise wages and affect investment and profit outcomes; employment effects are mixed and can be absent in some settings.

*Source: wpiea2019155 - excerpt from provided PDF content.*

### 1998. They find that union density has an inverted U-shape effect on employment and productivity.

### wpiea2019155 - 1998. They find that union density has an inverted U-shape effect on employment and productivity.

### Unions: employment, productivity, and profitability
- Findings from Brazil: union density has an inverted U-shape effect on employment and productivity.
  - "At the current level of union density in Brazil the effects are positive on both employment and productivity."
  - There are negative impacts on profitability at the observed union density levels.
- Findings from Peru (Saavedra and Torero (2005), manufacturing firms 1994–1996):
  - Unions reduce profits.
  - Unions have no effect on productivity.
- Synthesis for Latin America:
  - Overall results show negative effects on profits.
  - Effects on productivity and employment are mixed:
    - No effects on employment in Uruguay.
    - Positive effects on employment in Brazil (with an inverted U-shape).
    - No effect on productivity in Peru.
    - Positive effect on productivity in Brazil (with an inverted U-shape).
  - Net implication: "the effects of unions overall appear to be positive if unionization is moderate in Latin American countries, but nil or negative if unionization rates are very low or very high."

### Demographic factors — Skill: education and age
- Trends in education and experience:
  - Average years of education for adults increased from 5.8 to 8.3 between 1990 and 2010 (Messina and Silva (2017)).
  - Fraction of high school graduates rose from 21% to 36% between 1990 and 2010.
  - Share of individuals with primary education declined from 54% to 39% between 1990 and 2010.
  - Share of college-educated doubled between 1990 and 2010.
  - Average years of schooling: U.S. overall is 11 years; Latin America is 5 years on average.
  - Experience (measured by age) has been increasing; the population is aging, increasing the share of older workers relative to younger workers.
- Implications for productivity, employment, informality:
  - More educated and older workers are more skilled and more likely to be employed, particularly in the formal sector.
  - These trends likely contributed to declines in informality and unemployment.
- Ambiguous effects on inequality:
  - Increased supply of more educated and older workers could:
    - Reduce returns to skill (lower inequality) if demand unchanged.
    - Increase within-group inequality (higher inequality) because within-group variance is greater for more educated and older workers.
- Empirical changes in returns and earnings distribution:
  - Since the 2000s, earnings of unskilled workers have grown faster than earnings of the skilled (Messina and Silva (2017)).
  - Earnings changes since 2002: lower tail rose by 50%, middle rose by 32%, upper tail rose by 15% (Rodriguez-Castellan et al. (2016)).
  - Gap between college-educated and primary schooling fell from 330% to 240% since 2003.
  - Return to skill declined in all countries in the region except Costa Rica and the Dominican Republic (Messina and Silva (2017)).
  - Experience premium (21–30 years vs 0–5 years) declined from 56% in 1993 to 33% in 2013 (Messina and Silva (2017)).
  - Lopez-Calva and Lustig (2010): 50% of the decline in earnings inequality explained by narrowing in skilled vs unskilled earnings.
  - Gini reductions reported:
    - Brazil: close to 15.
    - Argentina, Mexico, Chile, Bolivia: 7.
    - Nicaragua, Panama, Honduras, El Salvador: close to 5.
    - Peru, Uruguay, Ecuador: reductions less than 5.

### Demographic factors — Gender
- Labor force participation of women has grown over the past two decades.
  - Potential effects: could reduce employment and increase informality (women more likely to be unemployed and work in informal sector); could mechanically increase inequality since women typically earn less than men.
- Gender gap levels across countries:
  - 20% in Chile and Peru.
  - 15% in Brazil.
  - Between 5% and 10% in Uruguay, Dominican Republic, Mexico, Venezuela and Bolivia.
  - Positive but below 5% in Panama, Nicaragua, Guatemala and Colombia.
  - Women earn on average more than men in El Salvador, Costa Rica, Argentina, Ecuador and Paraguay.
- Time pattern:
  - Gender gap stable from 1993 to 1999.
  - Large decline between 1999 and 2003.
  - Smaller decline since 2003.
  - Conclusion: entry of women into the labor force has contributed to the decline in inequality.

### Institutional factors and labor-market policies — impacts on employment, informality, wages, inequality, and productivity
- Employment and informality:
  - Reduced strictness in Employment Protection Legislation (EPL) and reduced payroll taxes likely boosted formal employment.
  - Centralized collective bargaining in countries with medium unionization rates could have contributed to higher formal employment, but may create insider-outsider problems for non-formal workers.
  - Higher minimum wages probably had little effect or a small effect in reducing formal employment.
- Inequality:
  - Increase in skilled workers and female labor force participation contributed to decline in inequality over the past two decades.
  - Institutional factors that tend to increase wages reduce inequality if they do not lower employment:
    - Higher minimum wages and more generous unemployment insurance increase earnings at the lower end and reduce inequality, "insofar as they do not lower employment creation as the evidence seems to suggest."
  - Unionization in Latin America is more likely to increase wages for those higher up in the income distribution, thus not necessarily reducing inequality.
  - Payroll taxes reduce wages for the formal sector; since payroll taxes have been declining in the region, these changes likely contributed to lower inequality if less-skilled workers benefited more from resulting wage increases.
- Productivity:
  - Evidence is more scant on productivity effects.
  - Reviewed evidence suggests unions increase productivity for those who get unionized jobs.
  - Evidence on EPL indicates EPL decreases productivity in the region.
  - No evidence on the impact of minimum wages (MW) and unemployment insurance (UI) on productivity in the region; evidence from other regions points to positive impacts of these institutional factors on productivity.
  - Payroll taxes and other factors that push employers away from hiring formal workers toward informality generate misallocation toward smaller and less capital-intensive firms that are less productive.
  - Reductions in payroll taxes in the region during the past few decades should have contributed to increases in productivity (Eslava et al., 2009).

### Conclusion — overall labor-market performance in Latin America
- Improvements over the past few decades:
  - Unemployment and informality have declined.
  - Inequality has decreased substantially.
- Remaining challenges:
  - Unemployment, informality, and inequality remain higher in Latin America than in the rest of the world.
  - Productivity has remained roughly constant over the past few decades and remains lower in LAC compared to the U.S. and Canada.
- Drivers of improvements:
  - Demographic factors: growth in more educated and older workers; increased female labor force participation.
  - Institutional changes: reduced EPL strictness, lower payroll taxes, centralized collective bargaining in some countries, higher minimum wages and more generous UI (with conditional effects on employment).
  - Net effect: combination of demographic and institutional factors contributed to declines in unemployment, informality, and inequality, with mixed and more limited evidence on productivity effects.

*Source: wpiea2019155 - 1998. They find that union density has an inverted U-shape effect on employment and productivity.*

### 8682. London: Center for Economic Policy Research.

### 8682. London: Center for Economic Policy Research.

### Labor market outcomes and unemployment
- Unemployment rate, total as a percentage of the labor force (national estimates) reported for 1997 to 2016 (World Bank Databank source referenced).
- Figure 1: Unemployment Rate Latin America & the Caribbean — chart spans 1997–2016 and distinguishes Latin America & the Caribbean, OECD, Canada, United States (no single summary value provided in text).
- Figure 2: Unemployment in Latin America (ILO estimates for 2017) — selected numeric estimates shown:
  - 2.6
  - 8.2
  - 12.9
  - 2.3
  - 7.1
  - 11.2
  - 3
  - 9.9
  - 15.1

### Informality
- Informal employment and informal sector as a percent of employment presented for 1997–2016 (ILO Statistics).
- Figure 3: Informal Employment in Latin America — time series 2003–2017 (no single summary value provided).
- Figure 4: Share of Workers in Informal Sector, 2012 — sex disaggregation with numeric values:
  - Men: 39.2
  - Women: 51.2
  - Additional figure value: 73
  - Additional figure value: 40.3
  - Additional figure value: 55.8
  - Additional figure value: 78.5

### Labor productivity
- ILO estimates (Nov. 2017) of output per worker (GDP constant 2011 international $ in PPP) for 1997–2017.
- Figure 5: Labor Productivity in Latin America — comparative series include Latin America and the Caribbean and Northern America (no single summary value provided).
- Figure 6: Labor Productivity in Latin America — country-level values (output per worker in constant 2011 international $ PPP):
  - Honduras: 9,841
  - Nicaragua: 11,544
  - Bolivia: 15,081
  - Guatemala: 18,702
  - El Salvador: 18,726
  - Paraguay: 19,002
  - Ecuador: 21,912
  - Peru: 22,181
  - Colombia: 26,400
  - LAC Average: 28,629
  - Venezuela: 29,555
  - Brazil: 31,681
  - Dominican Republic: 32,253
  - Costa Rica: 36,417
  - Mexico: 38,390
  - Uruguay: 42,787
  - Argentina: 45,109
  - Panama: 47,143
  - Chile: 48,603
  - Canada: 89,682
  - United States: 119,208

### Inequality and wages
- Figure 7: Levels of the Gini Coefficient (Milanovic dataset covering 1950–2015) — regional series for 1997–2014 (no single summary value provided).
- Figure 8: Change in Labor Income Gini in Latin America 1995–2013 — country changes shown as numeric series ranging approximately between -0.6 and 0.4 for listed countries (plot tick labels: -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.1 0.2 0.3 0.4). Countries listed include Bolivia, Brazil, Ecuador, Chile, Nicaragua, Argentina, Peru, Uruguay, El Salvador, Mexico, Panama, Dominican Republic, Honduras, Costa Rica.
- Figure 18: Change in Return to Skill (completed tertiary vs. primary or less) in Latin America 1995–2013 — numeric axis spans -15 to 15 with listed countries: Brazil, Bolivia, Chile, Mexico, Argentina, Nicaragua, Panama, Honduras, El Salvador, Peru, Uruguay, Ecuador, Dominican Republic, Costa Rica.
- Figure 19: Gender Wage Gap in Latin America — country values shown on axis from -10 to 25; countries listed include El Salvador, Costa Rica, Argentina, Ecuador, Paraguay, Colombia, Guatemala, Nicaragua, LAC Average, Panama, Bolivia, Venezuela, Mexico, Dominican Republic, Uruguay, Brazil, Peru, Chile.

### Labor market institutions, regulations, and taxes
- Employment Protection and contract structure:
  - Figure 9: Employment Protection Legislation Index in Latin America — compares regular contracts and temporary employment across countries (numerical axis 0–6; specific country values not provided in text).
  - Figure 10: Changes in EPL in Latin America — chart reports paired values for Temporary employment and Regular contracts with sample numeric ticks including 0.21, 0.33, 1.17, 1.51, 2.29, 2.36, 2.42, 2.62, 3.04, 4.08.
- Minimum wage ratios:
  - Figure 11: Ratio of minimum to mean wages in Latin America — country values:
    - Peru: 0.16
    - Mexico: 0.29
    - Uruguay: 0.39
    - Brazil: 0.43
    - Nicaragua: 0.47
    - Chile: 0.47
    - Bolivia: 0.53
    - Costa Rica: 0.56
    - Panama: 0.57
    - Colombia: 0.57
    - LAC average: 0.60
    - El Salvador: 0.65
    - Venezuela: 0.67
    - Argentina: 0.68
    - Ecuador: 0.77
    - Paraguay: 0.79
    - Dominican Republic: 0.87
    - Guatemala: 0.97
    - Honduras: 0.99
  - Figure 12: Ratio of minimum to median wages in Latin America — country values:
    - Costa Rica: 0.03
    - Mexico: 0.34
    - Uruguay: 0.45
    - Ecuador: 0.49
    - Bolivia: 0.51
    - Peru: 0.61
    - LAC average: 0.67
    - Brazil: 0.70
    - Venezuela: 0.72
    - Panama: 0.73
    - Argentina: 0.87
- Payroll taxes and compensation costs:
  - Figure 13: Payroll Taxes in Latin America (expressed as a percentage of GDP for averages 1985–2009 where applicable) — country values:
    - Honduras: 0.13
    - Dominican Republic: 0.15
    - Guatemala: 0.17
    - El Salvador: 0.17
    - Venezuela: 0.18
    - Nicaragua: 0.19
    - Ecuador: 0.19
    - Chile: 0.20
    - Bolivia: 0.23
    - Peru: 0.24
    - LAC Average: 0.25
    - Paraguay: 0.26
    - Colombia: 0.31
    - Costa Rica: 0.32
    - Uruguay: 0.34
    - Mexico: 0.39
    - OECD Average: 0.41
    - Argentina: 0.48
    - Canada: 0.65
    - Brazil: 0.91
  - Figure 14: Gross Replacement Rate in Latin America — values shown include:
    - Latin America & The Caribbean: 0.05
    - Brazil: 0.15
    - Argentina: 0.22
    - Chile: 0.23
    - Uruguay: 0.25
    - Venezuela: 0.30
    - United States: 0.57
    - Canada: 0.59

### Collective bargaining and unionization
- Figure 15: Trade Union Density Rate in Latin America — country values (percentage of employees who are union members):
  - Nicaragua: 5.3
  - Peru: 5.7
  - Paraguay: 6.7
  - Colombia: 9.5
  - United States: 10.3
  - Dominican Republic: 11.0
  - Panama: 11.9
  - Mexico: 12.5
  - LAC Average: 16.3
  - Brazil: 18.9
  - El Salvador: 19.0
  - Costa Rica: 19.4
  - Chile: 19.6
  - Argentina: 27.7
  - Canada: 28.4
  - Uruguay: 30.1
  - Bolivia: 39.1
- Figure 16: Collective Bargaining Coverage Rate in Latin America — country values (percentage of employees covered by collective agreements):
  - Panama: 1.0
  - Venezuela, Bolivarian Republic of: 2.5
  - Peru: 4.8
  - El Salvador: 5.0
  - Honduras: 5.7
  - Nicaragua: 7.6
  - Mexico: 9.9
  - Costa Rica: 10.6
  - United States: 11.5
  - Colombia: 15.7
  - Chile: 17.9
  - LAC Average: 24.8
  - Canada: 30.3
  - Argentina: 51.8
  - Brazil: 70.5
  - Cuba: 81.4
  - Uruguay: 94.6

### Human capital and returns to education
- Figure 17: Average Years of Schooling in Latin America (Barro & Lee averages 1985–2010 for individuals 25+):
  - Selected country/group values shown on axis include 2.07, 4.45, 11.24, 2.75, 5.02, 11.27, 2.39, 4.72, 11.22 (breakdowns by Total, Male, Female for listed countries; no single summary provided).

### Data sources and notes included in the unit
- Cited datasets and sources within the unit include:
  - World Bank Databank (unemployment series SL.UEM.TOTL.NE.ZS)
  - International Labor Organization (ILOSTAT) for multiple indicators (informal employment, productivity, employment protection, wages, payroll taxes, replacement rates, union density, collective bargaining coverage)
  - Branko Milanovic Gini dataset (World Bank, Research Department October 2016)
  - Messina, Julián, and Joana Silva. 2018. Wage Inequality in Latin America (World Bank)
  - Barro & Lee Educational Attainment Dataset (Quality of Government Dataset 2017)
- The unit contains figure notes explaining ILO definitions (e.g., informal sector and informal employment definitions) and datasets' coverage years and country-specific data vintage information where applicable.

*Source: wpiea2019155 - 8682. London: Center for Economic Policy Research.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019155.pdf_