## wpiea2020029-print-pdf

## Source details

**Canonical URL:** [wpiea2020029-print-pdf](https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020029-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2020/english/wpiea2020029-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2020/english/wpiea2020029-print-pdf.pdf.json)

---

### I. Introduction — context, approach, and goals
- Context and problem
  - A dramatic fall in employment following the global financial crisis worsened Spain’s income distribution and social inclusion.
  - Spanish income inequality, measured by the Gini coefficient, deteriorated by about 2 percentage points between 2008 and 2013.
  - The risk of poverty increased by 4 percentage points between 2008 and 2013.
  - Job losses disproportionately affected low-skilled and young workers, particularly those on temporary contracts.
- Policy response
  - A comprehensive structural reform package was adopted in response to the crisis, with emphasis on labor market reforms.
  - The 2012 labor reforms involved changes to the collective bargaining framework, employment protection regulation, and active labor market policies.
- Empirical objective
  - Assess the distributional impact of Spain’s 2012 labor market reforms on absolute measures of living standards (employment, average hours worked) and relative measures of income distribution.
- Two-step empirical strategy
  - Step 1: Adjust original data in a panel regression for time-invariant country-specific and time fixed effects, then detrend using a country-specific time trend for the pre-reform period to obtain adjusted series capturing policy and idiosyncratic shocks.
  - Step 2: Apply the synthetic control method to the adjusted data to construct a counterfactual (possible evolution of outcomes in the absence of reforms) and estimate reform impact.
- Methodological advantage
  - The approach provides direct insights on aggregate effects of labor reforms on several dimensions of social inclusion and increases the potential donor pool for constructing counterfactuals.

### II. Main empirical findings
- Employment and inequality
  - The 2012 labor market reforms improved employment outcomes: employment growth was substantially higher after the 2012 reforms than the estimated employment growth in the absence of the reforms.
  - The reforms had a positive distributional impact as evidenced by an estimated negative impact on the Gini coefficient.
  - No substantial impact was found on the overall share of population at risk of poverty.
- Hours worked and part-time employment
  - Evidence suggests a reduction in average hours worked induced by the reforms.
  - The reforms were associated with a deterioration of involuntary part-time employment (results statistically insignificant).
  - Enhanced flexibility in using part-time contracts may have incentivized employers to offer more part-time contracts despite employees’ desire to work more hours.
  - Structural shift from construction toward services (a sector with much higher rate of part-time employment) may have contributed to fewer hours worked.
- In-work poverty
  - A negative impact of the reforms on in-work poverty was observed, which likely reflects the reduction in hours worked.
- Heterogeneity and caveats
  - Estimated reform impacts could still be influenced by other idiosyncratic shocks that happened simultaneously with the labor market reforms (examples include the severe macroeconomic crisis, other structural reforms, or the 2012–14 Agreement for Employment and Collective Bargaining signed by Spanish social agents in early 2012).
  - Residual components of common shocks may remain in adjusted series due to potentially heterogeneous country responses.

### III. Box 1 — Spain: Key Elements of the 2012 Labor Market Reforms
- Key reform measures
  - Reform package enacted in 2012 with some adjustments until 2013.
  - Measures aimed at: (i) enhancing firms’ flexibility; (ii) reducing duality; and (iii) improving active labor market policies.
  - Some policies applied only to new contracts or to future years of service on existing contracts.
- Firms’ flexibility (selected measures)
  - Firm-level agreements prioritized over sector or region-level agreements.
  - Opt-outs from collective agreements were eased.
  - Period of ultra-activity of collective agreements was reduced.
  - Part-time contracts made more flexible, including allowing overtime/complementary hours.
  - Use of fair dismissals based on objective business criteria was extended and firm-level agreements on dismissals facilitated.
  - Requirement for administrative authorization of collective dismissals was eliminated.
  - Training requirement for dismissed workers was introduced.
  - A tax was imposed on profitable employers that implemented collective dismissals.
- Duality (selected measures)
  - Cost for unfair dismissals of permanent workers was reduced from 45 to 33 days' wages per year of seniority up to 24 months (down from 42 months).
  - A special contract for small firms was introduced to incentivize hiring of permanent workers, particularly of unemployed youth.
  - The 2-year limit for extension of temporary contracts was re-instated.
- Active labor market policies (selected measures)
  - Training and apprenticeship contracts were introduced, providing firms with incentives to train and retrain uneducated youth.
  - A new activation program for the long-term unemployed was launched.
  - A greater role was given to private agencies in training and job placement.
  - A system of evaluation and results-oriented allocation of funds was introduced in 2013.

### IV. Context and stylized facts (selected statistics)
- Crisis impact (2008–13)
  - Real GDP per capita in Spain was down by 10 percent in 2013 compared to 2008.
  - Employment dropped on average 3.5 percent annually during 2009–13.
  - Overall unemployment rate jumped to 27 percent during 2013.
  - Youth unemployment rate surpassed 55 percent.
  - Gini coefficient increased by about 2 percentage points by 2013.
  - Share of population at risk of poverty rose by about 4 percentage points by 2013.
  - In-work poverty rate jumped by 2 percentage points in 2014.
  - Real monthly wage earnings of workers in the first and second lowest deciles adjusted most sharply downwards during the crisis; distribution of real hourly wages was relatively stable.
- Recovery and structural change (post-2013)
  - In 2018 real GDP per capita exceeded its 2007 level by 2 percent.
  - Employment growth averaged about 2.5 percent during 2014–18.
  - Unemployment rate in 2018 was down by more than 10 percentage points from its 2013 27.
  - Share of construction in employment was down by more than 6 percentage points compared to 2008.
  - Out of three percentage points increase in the rate of part-time employment during 2008–18, one percentage point was due to the structural change in the economy.

### V. Methodology (two-stage empirical approach)
- Stage 1 — Panel regression adjustments
  - Remove time-invariant country-specific fixed effects and common time effects to obtain residuals v_it, then detrend residuals by fitting a linear time trend up to the reform date (2012) and extending it post-reform to obtain ε_it (time-variant country-specific component).
  - Equations used:
    - y_it = a + μ_i + δ_i + v_it
    - v_it = β_i * T12_i + ε_it
  - Purpose: isolate policy shocks (such as labor market reforms) and country-specific idiosyncratic shocks.
- Stage 2 — Synthetic control on time-variant component
  - Apply synthetic control method to ε_it to construct counterfactuals as weighted averages of control countries minimizing pre-reform differences.
  - Placebo experiments performed for inference; only placebo countries with pre-reform root mean squared predictive error not more than twice that of Spain were used for inference.
- Data and sample notes
  - Cross-country panel covers around 2000 to either 2017 or 2018 depending on data availability.
  - Outcome variables: Gini coefficient, income ratio of top 20 percent to bottom 20 percent, employment growth, youth unemployment rate, involuntary part-time employment, average hours worked, risk of poverty, in-work poverty.
  - Covariates vary by model and include: employment growth, overall and youth unemployment rate, wage growth, old age dependency ratio, share of population by education, part-time employment, average hours worked, trade union density, GDP per capita, real GDP growth, social spending to GDP ratio.
  - Comparable control group selected from EU and non-EU OECD advanced countries; countries that implemented major post-crisis labor market reforms were excluded.
- Methodological limitations noted
  - Cannot fully control for other idiosyncratic shocks or reverse causality.
  - Some control units may have undertaken minor reforms biasing estimates downward.
  - Results could be sensitive to choice of linear trend and heterogeneous responses to common shocks.
  - Share of temporary employment remains an outlier for Spain, limiting appropriate control units for that variable.

### VI. Main results (summary of empirical findings)
- Employment — extensive margin
  - Positive impact on job creation: Spain’s time-variant country-specific component of employment growth was close to zero before 2012 but increased substantially after the reforms.
  - Synthetic control mimicked pre-reform behavior well; post-2012 Spain shows a substantial positive gap relative to synthetic control.
  - Placebo experiment indicates estimated impact statistically significant at the 5 percent level for employment growth.
  - Youth unemployment rate: similar positive picture; pre-reform fit reasonable and post-reform divergence large; placebo p-value: 0.06 for youth unemployment rate five years after reforms.
- Employment — intensive margin
  - Negative impact on intensive-margin indicators:
    - Average hours worked: significant divergence between Spain’s actual and synthetic measures after 2012; placebo experiment p-value close to zero.
    - Involuntary part-time employment: time-variant component shows a declining trend after 2012 but remains above synthetic measure, suggesting reforms might have increased involuntary part-time employment; placebo experiment suggests this impact is not statistically significant and result is not robust to covariate selection.
  - Robustness checks for average hours worked (comparison with Euro Area average; excluding Italy and/or Ireland from control sample) confirm substantial negative deviation in Spain’s average hours worked after 2012.
- Income inequality
  - Gini coefficient: time-variant country-specific component exhibits a substantial decline since 2012; synthetic control tracks pre-2012 well and diverges after 2012.
  - Placebo experiment: estimated negative impact on Gini statistically significant at the 10 percent level.
  - Income ratio of top 20 percent to bottom 20 percent: results do not suggest any systematic impact of the reforms on this indicator.
  - Interpretation: improved labor utilization (job creation) likely benefited lower half of income distribution.
- Poverty measures
  - Population at risk of poverty: no systematic deterioration due to reforms; actual outcome tends to be below synthetic measure but placebo analysis indicates deviation not statistically different from other countries (placebo p-value: 0.43).
  - In-work poverty: considerable divergence after 2012 with Spain’s actual outcome worsening relative to synthetic control; placebo p-value: 0.00. Suggests 2012 reforms might have contributed to an increase in in-work poverty, potentially reflecting increased involuntary part-time employment and lower working hours. Pre-reform data series for in-work poverty is shorter, so these results require caution.

### VII. Conclusions and interpretation (key takeaways)
- Positive effects
  - The 2012 reforms improved employment outcomes, including for youth. Post-reform growth of employment was systematically higher and youth unemployment lower compared with the counterfactual without reforms.
  - The reforms contributed to a significant reduction in the Gini coefficient five years after the reforms.
- Negative or mixed effects
  - Reforms appear to have had negative effects on the intensive margin of employment: reduction in average hours worked and an increase in involuntary part-time employment (the latter not conclusive statistically).
  - No systematic impact on overall population at risk of poverty, but evidence points to a deterioration in in-work poverty following the reforms.
- Possible channels and caveats
  - Enhanced flexibility of part-time contracts and structural shift from construction to services (sectors with higher part-time shares) are plausible contributors to increased part-time employment and lower hours.
  - The analysis does not isolate the role of specific reform measures; other idiosyncratic shocks or contemporaneous structural policies could influence estimated impacts.
  - Results should be interpreted with caution given methodological limitations and potential sensitivity to choice of controls and trends.
- Policy-relevant points emphasized
  - Labor market reforms can raise employment and reduce measured income inequality while having ambiguous effects on in-work poverty and hours worked.
  - Complementary policies—such as well-targeted active labor market policies and measures addressing working-time adequacy—may be important to address adverse effects on hours worked and in-work poverty.
  - Caution is required when attributing observed outcomes solely to reforms because of concurrent macroeconomic shocks and other reforms.

### VIII. Annex I — Spain’s Comparison with Advanced European Countries (method and robustness notes)
- Data sources: Eurostat, Haver Analytics, OECD, WDI, and IMF staff estimates.
- Advanced Europe defined as: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, and the United Kingdom.
- Variables and sample timing (as labeled in figures): employment growth (2000–2018), employment rate (2000–2017), average hours worked (2000–2018), Gini coefficient (2001–2017), population at risk of poverty (2000–2018), in-work poverty (2003–2017).
- Synthetic control compositions (examples from Figure I.2)
  - Employment Growth control unit: France (49 percent), Latvia (29  percent), Netherlands (14 percent), Belgium (9 percent).
  - Employment rate control unit: France (47 percent), Slovakia (32  percent), Germany (19 percent), Ireland (2 percent).
  - Average Hours Worked control unit: Italy (67 percent), Slovakia (33 percent).
  - Gini Coefficient control unit: France (42 percent), Latvia (39  percen t), Italy (19 percent).
  - Risk of Poverty control unit: Italy (80 percent), Latvia (20 percen t).
  - In-work Poverty control unit: Italy (66 percent), Greece (34 percen t).
- Synthetic control setup and optimization criterion summarized:
  - Treatment effect: a_i,t = (Y_i,t^I − Y_i,t^N).
  - Counterfactual linear factor model and weight optimization minimizing pre-intervention mean square prediction error described in formal notation.
- Covariates and data sources
  - Dependent variables and sources listed (Employment Growth: OECD; Youth Unemployment Rate: OECD; Involuntary Part-Time Employment: Eurostat; Average Hours Worked: OECD; Gini Coefficient: World Bank WDI; Ratio of Top 20 Percent to Bottom 20 percent: Eurostat; At Risk-of-Poverty Rate: Eurostat; In-Work Poverty: Eurostat; Real GDP per Capita: WEO; Overall Unemployment Rate: OECD; Youth In-Work Poverty: Eurostat).
  - Selected covariates include GDP per capita, old age dependency ratio, rate of underemployment, youth unemployment rate, employment growth, share of population with secondary or primary education, part-time employment, trade union density, social spending to GDP ratio, GDP growth, wage growth, average hours worked.
- Robustness and auxiliary findings
  - Unemployment-rate robustness check control unit: Italy (73 percent), Latvia (27 percent); Placebo Units p-value: 0.05.
  - Average Hours Worked robustness: alternative synthetic compositions when excluding Italy and/or Ireland detailed with exact percentage weights.
  - EU-15 defined as simple average for listed EU countries.

*Source: wpiea2020029-print-pdf (excerpts of Introduction, Box 1, Main results, Annex I).*

### References .............................................................................................................

### wpiea2020029-print-pdf - References .....................................................................................................

### I. Introduction — context, approach, and goals
- Context and problem:
  - A dramatic fall in employment following the global financial crisis worsened Spain’s income distribution and social inclusion.
  - Spanish income inequality, measured by the Gini coefficient, deteriorated by about 2 percentage points between 2008 and 2013.
  - The risk of poverty increased by 4 percentage points between 2008 and 2013.
  - Job losses disproportionately affected low-skilled and young workers, particularly those on temporary contracts.
- Policy response:
  - A comprehensive structural reform package was adopted in response to the crisis, with emphasis on labor market reforms.
  - The 2012 labor reforms involved changes to the collective bargaining framework, employment protection regulation, and active labor market policies.
- Empirical objective:
  - Assess the distributional impact of Spain’s 2012 labor market reforms on absolute measures of living standards (employment, average hours worked) and relative measures of income distribution.
- Two-step empirical strategy:
  - Step 1: Adjust original data in a panel regression for time-invariant country-specific and time fixed effects, then detrend using a country-specific time trend for the pre-reform period to obtain adjusted series capturing policy and idiosyncratic shocks.
  - Step 2: Apply the synthetic control method to the adjusted data to construct a counterfactual (possible evolution of outcomes in the absence of reforms) and estimate reform impact.
- Methodological advantage:
  - The approach provides direct insights on aggregate effects of labor reforms on several dimensions of social inclusion and increases the potential donor pool for constructing counterfactuals.

### II. Main empirical findings
- Employment and inequality:
  - The 2012 labor market reforms improved employment outcomes: employment growth was substantially higher after the 2012 reforms than the estimated employment growth in the absence of the reforms.
  - The reforms had a positive distributional impact as evidenced by an estimated negative impact on the Gini coefficient.
  - No substantial impact was found on the overall share of population at risk of poverty.
- Hours worked and part-time employment:
  - Evidence suggests a reduction in average hours worked induced by the reforms.
  - The reforms were associated with a deterioration of involuntary part-time employment (results statistically insignificant).
  - Enhanced flexibility in using part-time contracts may have incentivized employers to offer more part-time contracts despite employees’ desire to work more hours.
  - Structural shift from construction toward services (a sector with much higher rate of part-time employment) may have contributed to fewer hours worked.
- In-work poverty:
  - A negative impact of the reforms on in-work poverty was observed, which likely reflects the reduction in hours worked.
- Heterogeneity and caveats:
  - Findings must be interpreted with caution: estimated reform impacts could still be influenced by other idiosyncratic shocks that happened simultaneously with the labor market reforms (examples include the severe macroeconomic crisis, other structural reforms, or the 2012–14 Agreement for Employment and Collective Bargaining signed by Spanish social agents in early 2012).
  - Residual components of common shocks may remain in adjusted series due to potentially heterogeneous country responses.

### III. Relation to existing literature
- Cross-country and Spain-specific evidence summarized:
  - Causa et al. (2016): labor productivity growth alone is not inclusive; labor utilization growth is. Some labor market reforms tend to raise income inequality (examples: reducing job protection for regular contracts, limiting automatic extension of collective agreements to non-union members, reducing unionization rates), while well-targeted active labor market policies can enhance inclusiveness.
  - Other cross-country findings:
    - Deeper labor flexibility associated with rising income inequality in OECD countries (Tridico, 2018).
    - Labor market reforms may lead to suboptimal distributional effects in presence of product market rigidities (Roeger et al., 2019).
    - Ostry et al. (2018): inconclusive evidence on collective bargaining reforms’ impact on income inequality.
  - Spain-specific findings:
    - Anghel et al. (2018): increase in wage income inequality during the crisis largely explained by drops in jobs and hours worked at lower-income deciles; inability of firms to change working conditions prior to 2012 led to adjustment through quantities rather than prices.
    - OECD (2014), IMF (2015), Garcia-Perez (2016), Doménech et al. (2018): 2012 reforms supported wage flexibility and contributed to job creation.
    - Doménech et al. (2018) illustrative counterfactual: if wage flexibility facilitated by the reforms had started in early 2008, destruction of nearly 1.8 million full-time equivalent jobs could have been avoided over the following decade, and the unemployment rate at end-2017 would have been around 9 percent instead of 17 percent.
    - Mixed evidence on other margins: limited decline in coverage of collective negotiations (Bentolila and Jansen, 2016); negligible impact on individual dismissals (OECD, 2014); persistent weaknesses in spending and efficiency in active labor market policies (IMF, 2017a).
  - Macro studies:
    - IMF (2016): positive medium-term effects of labor and product market reforms on output and employment in advanced economies; employment protection reforms can be contractionary in periods of slack (Duval et al., 2017).
    - Andrés et al. (2017): in a recessionary deleveraging shock, labor and product market reforms can stimulate output and employment even in short run due to competitiveness effects.
    - Salas (2018): 2010 and 2012 labor reforms’ competitiveness gains estimated to account for nearly one-tenth to above one-quarter of Spain’s real export growth rate over 2010–13.

### IV. Interpretation and policy implications highlighted in the text
- Trade-offs:
  - The paper’s findings align with literature documenting growth-equity trade-offs associated with structural reforms (see e.g. Ostry et al., 2018).
  - Positive employment effects can offset potential inequality-increasing channels of some labor reforms; faster job creation tends to benefit households in the bottom half of the income distribution (Causa et al., 2016).
  - Grandfathering of existing job contracts under some reform measures in Spain may have mitigated potential negative distributional impacts.
- Policy-relevant points emphasized:
  - Labor market reforms can raise employment and reduce measured income inequality while having ambiguous effects on in-work poverty and hours worked.
  - Complementary policies—such as well-targeted active labor market policies and measures addressing working-time adequacy—may be important to address adverse effects on hours worked and in-work poverty.
  - Caution is required when attributing observed outcomes solely to reforms because of concurrent macroeconomic shocks and other reforms.

*Source: wpiea2020029-print-pdf - References (excerpts of Introduction and Literature Review).*

### Box 1. Spain: Key Elements of the 2012 Labor Market Reforms

### Box 1. Spain: Key Elements of the 2012 Labor Market Reforms

### Key reform measures
- Reform objectives and timing
  - Reform package enacted in 2012 with some adjustments until 2013.
  - Measures aimed at: (i) enhancing firms’ flexibility; (ii) reducing duality; and (iii) improving active labor market policies.
  - Some policies applied only to new contracts or to future years of service on existing contracts.
- Firms’ flexibility (selected measures)
  - Firm-level agreements were prioritized over sector or region-level agreements in collective bargaining framework.
  - Opt-outs from collective agreements were eased.
  - Period of ultra-activity of collective agreements was reduced.
  - Part-time contracts were made more flexible, including allowing overtime/complementary hours.
  - Use of fair dismissals based on objective business criteria was extended and firm-level agreements on dismissals facilitated.
  - Requirement for administrative authorization of collective dismissals was eliminated.
  - Training requirement for dismissed workers was introduced.
  - A tax was imposed on profitable employers that implemented collective dismissals.
- Duality (selected measures)
  - Cost for unfair dismissals of permanent workers was reduced from 45 to 33 days' wages per year of seniority up to 24 months (down from 42 months).
  - A special contract for small firms was introduced to incentivize hiring of permanent workers, particularly of unemployed youth.
  - The 2-year limit for extension of temporary contracts was re-instated.
- Active labor market policies (selected measures)
  - Training and apprenticeship contracts were introduced, providing firms with incentives to train and retrain uneducated youth.
  - A new activation program for the long-term unemployed was launched.
  - A greater role was given to private agencies in training and job placement.
  - A system of evaluation and results-oriented allocation of funds was introduced in 2013.

### Context and stylized facts (selected statistics)
- Crisis impact (2008–13)
  - Real GDP per capita in Spain was down by 10 percent in 2013 compared to 2008.
  - Employment dropped on average 3.5 percent annually during 2009–13.
  - Overall unemployment rate jumped to 27 percent during 2013.
  - Youth unemployment rate surpassed 55 percent.
  - Gini coefficient increased by about 2 percentage points by 2013.
  - Share of population at risk of poverty rose by about 4 percentage points by 2013.
  - In-work poverty rate jumped by 2 percentage points in 2014.
  - Real monthly wage earnings of workers in the first and second lowest deciles adjusted most sharply downwards during the crisis; distribution of real hourly wages was relatively stable.
- Recovery and structural change (post-2013)
  - In 2018 real GDP per capita exceeded its 2007 level by 2 percent.
  - Employment growth averaged about 2.5 percent during 2014–18.
  - Unemployment rate in 2018 was down by more than 10 percentage points from its 2013 27.
  - Share of construction in employment was down by more than 6 percentage points compared to 2008.
  - Out of three percentage points increase in the rate of part-time employment during 2008–18, one percentage point was due to the structural change in the economy.

### Methodology (two-stage empirical approach)
- Stage 1 — Panel regression adjustments
  - Remove time-invariant country-specific fixed effects and common time effects to obtain residuals v_it, then detrend residuals by fitting a linear time trend up to the reform date (2012) and extending it post-reform to obtain ε_it (time-variant country-specific component).
  - Equations used:
    - y_it = a + μ_i + δ_i + v_it
    - v_it = β_i * T12_i + ε_it
  - Purpose: isolate policy shocks (such as labor market reforms) and country-specific idiosyncratic shocks.
- Stage 2 — Synthetic control on time-variant component
  - Apply synthetic control method to ε_it to construct counterfactuals as weighted averages of control countries minimizing pre-reform differences.
  - Placebo experiments performed for inference; only placebo countries with pre-reform root mean squared predictive error not more than twice that of Spain were used for inference.
- Data and sample notes
  - Cross-country panel covers around 2000 to either 2017 or 2018 depending on data availability.
  - Outcome variables: Gini coefficient, income ratio of top 20 percent to bottom 20 percent, employment growth, youth unemployment rate, involuntary part-time employment, average hours worked, risk of poverty, in-work poverty.
  - Covariates vary by model and include: employment growth, overall and youth unemployment rate, wage growth, old age dependency ratio, share of population by education, part-time employment, average hours worked, trade union density, GDP per capita, real GDP growth, social spending to GDP ratio.
  - Comparable control group selected from EU and non-EU OECD advanced countries; countries that implemented major post-crisis labor market reforms were excluded.
- Methodological limitations noted
  - Cannot fully control for other idiosyncratic shocks or reverse causality.
  - Some control units may have undertaken minor reforms biasing estimates downward.
  - Results could be sensitive to choice of linear trend and heterogeneous responses to common shocks.
  - Share of temporary employment remains an outlier for Spain, limiting appropriate control units for that variable.

### Main results (summary of empirical findings)
- Employment — extensive margin
  - Positive impact on job creation: Spain’s time-variant country-specific component of employment growth was close to zero before 2012 but increased substantially after the reforms.
  - Synthetic control mimicked pre-reform behavior well; post-2012 Spain shows a substantial positive gap relative to synthetic control.
  - Placebo experiment indicates estimated impact statistically significant at the 5 percent level for employment growth.
  - Youth unemployment rate: similar positive picture; pre-reform fit reasonable and post-reform divergence large; placebo p-value: 0.06 for youth unemployment rate five years after reforms.
- Employment — intensive margin
  - Negative impact on intensive-margin indicators:
    - Average hours worked: significant divergence between Spain’s actual and synthetic measures after 2012; placebo experiment p-value close to zero.
    - Involuntary part-time employment: time-variant component shows a declining trend after 2012 but remains above synthetic measure, suggesting reforms might have increased involuntary part-time employment; placebo experiment suggests this impact is not statistically significant and result is not robust to covariate selection.
  - Robustness checks for average hours worked (comparison with Euro Area average; excluding Italy and/or Ireland from control sample) confirm substantial negative deviation in Spain’s average hours worked after 2012.
- Income inequality
  - Gini coefficient: time-variant country-specific component exhibits a substantial decline since 2012; synthetic control tracks pre-2012 well and diverges after 2012.
  - Placebo experiment: estimated negative impact on Gini statistically significant at the 10 percent level.
  - Income ratio of top 20 percent to bottom 20 percent: results do not suggest any systematic impact of the reforms on this indicator.
  - Interpretation: improved labor utilization (job creation) likely benefited lower half of income distribution.
- Poverty measures
  - Population at risk of poverty: no systematic deterioration due to reforms; actual outcome tends to be below synthetic measure but placebo analysis indicates deviation not statistically different from other countries (placebo p-value: 0.43).
  - In-work poverty: considerable divergence after 2012 with Spain’s actual outcome worsening relative to synthetic control; placebo p-value: 0.00. Suggests 2012 reforms might have contributed to an increase in in-work poverty, potentially reflecting increased involuntary part-time employment and lower working hours. Pre-reform data series for in-work poverty is shorter, so these results require caution.

### Conclusions and interpretation (key takeaways)
- Positive effects
  - The 2012 reforms improved employment outcomes, including for youth. Post-reform growth of employment was systematically higher and youth unemployment lower compared with the counterfactual without reforms.
  - The reforms contributed to a significant reduction in the Gini coefficient five years after the reforms.
- Negative or mixed effects
  - Reforms appear to have had negative effects on the intensive margin of employment: reduction in average hours worked and an increase in involuntary part-time employment (the latter not conclusive statistically).
  - No systematic impact on overall population at risk of poverty, but evidence points to a deterioration in in-work poverty following the reforms.
- Possible channels and caveats
  - Enhanced flexibility of part-time contracts and structural shift from construction to services (sectors with higher part-time shares) are plausible contributors to increased part-time employment and lower hours.
  - The analysis does not isolate the role of specific reform measures; other idiosyncratic shocks or contemporaneous structural policies could influence estimated impacts.
  - Results should be interpreted with caution given methodological limitations and potential sensitivity to choice of controls and trends.

*Source: Box 1. Spain: Key Elements of the 2012 Labor Market Reforms (wpiea2020029-print-pdf)*

### Annex I. Spain’s Comparison with Advanced European Countries

### Annex I. Spain’s Comparison with Advanced European Countries

### Absolute and Relative Measures of Living Standards (Figure I.1)
- Sources: Eurostat, Haver Analytics, OECD, WDI, and IMF staff estimates.
- Advanced Europe includes Austria, Belgium, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, and the United Kingdom.
- Variables and depicted measures (as labeled in the figure):
  - Employment Growth (Percent): time series covering 2000–2018 with Spain shown relative to the 25-75th percentile and median of Advanced European Countries.
  - Employment Rate (Percent): time series covering 2000–2017 with Spain shown relative to the 25-75th percentile and median of Advanced European Countries.
  - Average Hours Worked (Hours per week): time series covering 2000–2018 with Spain shown relative to the 25-75th percentile and median of Advanced European Countries.
  - Gini Coefficient: time series covering 2001–2017 with Spain shown relative to the 25-75th percentile and median of Advanced European Countries.
  - Population at Risk of Poverty (Percent): time series covering 2000–2018 with Spain shown relative to the 25-75th percentile and median of Advanced European Countries.
  - In-Work Poverty (Percent): time series covering 2003–2017 with Spain shown relative to the 25-75th percentile and median of Advanced European Countries.

### Results of Synthetic Control Analysis Based on Raw Data (Figure I.2)
- Sources: Eurostat, Haver Analytics, OECD, WDI, and IMF staff estimates.
- Employment Growth (1998–2017)
  - Spain vs. Control Unit.
  - Note: Control unit comprises of France (49 percent), Latvia (29  percent), Netherlands (14 percent), and Belgium (9 percent).
- Employment rate (1998–2017)
  - Spain vs. Control Unit.
  - Note: Control unit comprises of France (47 percent), Slovakia (32  percent), Germany (19 percent), and Ireland (2 percent).
- Average Hours Worked (2000–2017)
  - Spain vs. Control Unit.
  - Note: Control unit comprises of Italy (67 percent), and Slovakia (33 percent).
- Gini Coefficient (2002–2017)
  - Spain vs. Control Unit.
  - Note: Control unit comprises of France (42 percent), Latvia (39  percen t), and Italy (19 percent).
- Risk of Poverty (1999–2017)
  - Spain vs. Control Unit.
  - Note: Control unit comprises of Italy (80 percent) and Latvia (20 percen t).
- In-work Poverty (2004–2017)
  - Spain vs. Control Unit.
  - Note: Control unit comprises of Italy (66 percent) and Greece (34 percen t).

### Key Methodological Notes (from Annex II and Annex III material included in the unit)
- Synthetic control setup:
  - Consider J+1 countries; only country 1 is treated at time T0+1.
  - Observed outcome for untreated country: Y_i,t^N.
  - Observed outcome for treated country: Y_i,t^I.
  - Treatment effect for country i at time t: a_i,t = (Y_i,t^I − Y_i,t^N).
  - Counterfactual linear factor model: Y_i,t^N = δ_i + θ_i Z_i + φ_i μ_i + ε_i,t.
  - Synthetic control weights W = (e_2, ..., e_{J+1})' with e_j ≥ 0 and sum(e_j)=1.
  - Estimated treatment effect: â_1,t = Y_1,t − Σ_{j=2}^{J+1} e_j^* Y_j,t.
- Optimization criterion:
  - W* chosen to minimize ‖X_1 − X_0 W‖_V = (X_1 − X_0 W)' V (X_1 − X_0 W), where X_1 is the vector of pre-reform characteristics for the treated country, X_0 contains the same for control countries, and V is symmetric, positive semidefinite and diagonal chosen to minimize pre-intervention mean square prediction error.
- Covariates and data sources (as listed in Table III.1 and Table III.2):
  - Dependent variables and sources include Employment Growth (OECD), Youth Unemployment Rate (OECD), Involuntary Part-Time Employment (Eurostat), Average Hours Worked (OECD), Gini Coefficient (World Bank WDI), Ratio of Top 20 Percent Earners to Bottom 20 percent (Eurostat), At Risk-of-Poverty Rate (Eurostat), In-Work Poverty (Eurostat), Real GDP per Capita (WEO), Overall Unemployment Rate (OECD), Youth In-Work Poverty (Eurostat).
  - Selected covariates used in synthetic control specifications include GDP per capita, old age dependency ratio, rate of underemployment, youth unemployment rate, employment growth, share of population with secondary or primary education, part-time employment, trade union density, social spending to GDP ratio, GDP growth, wage growth, and average hours worked (as applicable to each dependent variable).

### Robustness and Auxiliary Findings (Annex IV material present)
- Unemployment Rate robustness (time-variant country-specific component, hours per week):
  - Sources: Eurostat and IMF staff estimates.
  - Note: Control unit for the unemployment-rate robustness check comprises of Italy (73 percent) and Latvia (27 percent).
  - Placebo analysis reported with Placebo Units p-value: 0.05.
- Average Hours Worked robustness (time-variant country-specific component, hours per week):
  - Sources: OECD and IMF staff estimations.
  - Notes on synthetic compositions in alternative specifications:
    - Excluding Italy: Synthetic Spain comprises of Ireland (44 percent), Netherlands (25 percent), Latvia (18 percent), Sweden (12 percent), and France (1 percent).
    - Excluding Ireland: Synthetic Spain comprises of Italy (60 percent), UK (39 percent), and Latvia (1 percent).
    - Excluding Italy and Ireland: Synthetic Spain comprises of UK (87 percent), Latvia (9 percent), and Denmark (4 percent).
  - EU-15 is defined as the simple average for: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxemburg, Netherlands, Portugal, Spain, Sweden, and UK.

*Sources: Figures and tables as labeled in the content unit; Eurostat, Haver Analytics, OECD, WDI, IMF staff estimates.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020029-print-pdf.pdf_