## _wp1264

## Source details

**Canonical URL:** [_wp1264](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1264.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1264.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1264.pdf.json)

---

### I. Introduction — scope and contribution
- Sample and period:
  - Panel for 97 countries over the period 1985-2008.
- Main contribution:
  - Uses a broad sample (97 countries) to control for unobserved country-specific characteristics and assess evolution over time.
  - Focuses on a composite indicator of labor market flexibility (and its six sub-components).
  - Analyzes both static and dynamic effects and addresses possible endogeneity and reverse causality from unemployment to labor market institutions.
- Context from prior literature:
  - Prior cross-country work cited includes Botero et al. (2004) and Feldmann (2009).
  - Previous evidence generally finds that more rigid labor market institutions associate with higher unemployment.

### II. Data and indicators
- Labor market flexibility data:
  - Source: Fraser Institute’s Economic Freedom of the World (EFW) database.
  - All indicators standardized on a 0–10 scale (higher = more flexible labor market).
  - Composite index range (sample): minimum 1.8, maximum 9.5, total observations for composite index = 1214.
- Six sub-components of labor market flexibility (EFW area5b):
  - Minimum wage (area5bi) — based on World Bank’s Difficulty of Hiring.
  - Hiring and firing regulation (area5bii) — based on WEF’s Global Competitiveness Report.
  - Centralized collective wage bargaining (area5biii) — based on WEF’s Global Competitiveness Report.
  - Mandated cost of hiring (area5biv) — based on World Bank Doing Business data.
  - Mandated cost of worker dismissal (area5bv) — based on World Bank Doing Business data.
  - Conscription (area5bvi) — based on The Military Balance and War Resisters International data.
- Other data sources:
  - IMF’s World Economic Outlook (WEO), World Bank’s World Development Indicators (WDI), Penn World Table version 7.0, Laeven and Valencia (2010) financial crises database.
- Descriptive correlations:
  - Composite labor market flexibility indicator is negatively correlated with all unemployment outcomes.
  - Sub-indicators with stronger negative correlations on average: minimum wage and hiring and firing regulation.
  - Strongest negative correlation observed between mandated hiring costs and youth unemployment.
  - Unemployment is mostly concentrated among young people aged 15–24.

### III. Empirical methodology
- Two-part empirical strategy:
  - Static reduced-form specification regressing unemployment (total, youth, long-term) on the composite labor market flexibility index and controls.
  - Dynamic specification to test whether labor market flexibility affects the change in unemployment over time.
- Static model specification (equation structure preserved in source):
  - Unemployment rate regressed on labor market flexibility, vector of controls (output gap, ln government consumption to GDP (lncg), ln openness (lnopen), ln urban population (lnurbpop), ln population density (lnpopdens), crisis dummy, lagged unemployment U_{t-1}), and country fixed effects.
- Key control and behavioral findings:
  - Output gap included to control business cycle fluctuations.
  - Government size proxied by lncg (log of ratio of government consumption to GDP).
  - Crisis dummy equals 1 for occurrence of a financial crisis and 0 otherwise.
  - High persistence of unemployment is modeled via the lagged unemployment term.

### IV. Main empirical findings — static and dynamic effects
- Aggregate effects (total unemployment):
  - Increasing the rating of the composite labor market index by one standard deviation decreases, on average, the total unemployment rate by about 0.3–1.3 percentage points.
  - Baseline specification (only country fixed effects and output gap): a one standard deviation increase in the labor market index is associated with a 0.9 percentage point decrease in the unemployment rate.
  - Five-year averaged data: a one standard deviation increase in the labor market index decreases the unemployment rate by 1.3 percentage points (column IX).
- OECD vs non-OECD comparisons:
  - Non-OECD: a one standard deviation increase in the labor market institutions index is associated with a 0.96 percentage point decrease in unemployment.
  - OECD: the effect is about 0.86 percentage points.
  - The effect and persistence are not statistically different between OECD and non-OECD countries (95 percent confidence bands overlap).
- Persistence of unemployment:
  - A one percentage point increase in previous unemployment translates into a 0.83 percentage point higher unemployment in the current period (0.82 when government size is included).
- Time evolution:
  - Recursive re-estimation over subperiods 1980–1990 through 1980–2008 shows the effect of labor market flexibility on the level of unemployment has increased over time.
  - The effect becomes sizeable and statistically significant only after 2002 (caveat: unbalanced panel with more observations after 2000).
- Youth and long-term unemployment (static results):
  - Youth unemployment: a one standard deviation increase in the composite labor markets index decreases youth unemployment by 1.41 percentage points.
  - Long-term unemployment: a one standard deviation improvement in the composite index reduces long-term unemployment by 3.56 percentage points.
  - Sub-component static effects on youth unemployment (Table 7):
    - area5bii (hiring and firing regulation): coefficient -0.494.
    - area5bvi (conscription): coefficient -0.385.
    - area5biii (degree of decentralization in wage bargaining): coefficient 0.369 (positive, not significantly different from zero).
  - Sub-component static effects on long-term unemployment (Table 8):
    - area5biv (mandated cost of hiring): -1.415 (selected entry in Table 8); area5bvi: -0.992 (selected entry in Table 8).

- Dynamic relationship: changes in unemployment
  - Estimation: two-step GMM-system estimator to address endogeneity and reverse causality.
  - Composite-index dynamic effects:
    - Increasing the rating of the composite labor market index by one standard deviation decreases, on average, the change in the total unemployment rate by about 0.35‒0.49 percentage points.
    - Baseline specification suggests a one standard deviation increase is associated with a half percentage point decrease in the annual change in unemployment.
  - Table 10 dynamic results (selected exact coefficients):
    - Unemployment (area5b): -0.321 (significant at ***).
    - Youth unemployment (area5b): -0.346 (significant at *).
    - Long-term unemployment (area5b): -1.296 (significant at **).
  - Dynamic sub-component effects (Table 11, selected exact coefficients):
    - area5bii (hiring and firing regulations): unemployment -0.214 (significant at *); youth unemployment -0.490 (significant at **).
    - area5bv (mandated cost of worker dismissal/hiring costs as reported): unemployment -0.136 (significant at **); long-term unemployment -1.281 (significant at **).
    - area5bvi (conscription): unemployment -0.057 (selected); youth unemployment -0.278 (significant at ***); long-term unemployment -0.884 (significant at **).

### V. Robustness and diagnostic tests
- Robustness checks:
  - Results persist when controlling for macroeconomic and demographic variables.
  - Results hold for OECD and non-OECD subsamples.
  - Five-year averaging amplifies estimated effects, suggesting larger medium-term impacts.
- GMM diagnostics (selected exact p-values reported):
  - Hansen J-test p-value ranges reported: 0.2 to 0.4.
  - Arellano–Bond AR(2) test p-values reported: 0.6 to 0.9.
  - Specific Table 9 Hansen J-test p-values: 0.20, 0.31, 0.32, 0.35, 0.43, 0.32, 0.31, 0.25.
  - Specific Table 9 Arellano–Bond AR(2) p-values: 0.58, 0.57, 0.63, 0.57, 0.60, 0.63, 0.57, 0.89.
- OECD vs non-OECD dynamic comparison (Table 9 full-sample dynamic area5b):
  - Full-sample area5b coefficients (columns I–VI): -0.321, -0.303, -0.309, -0.231, -0.289, -0.289.
  - OECD (column VII) area5b: -0.108.
  - Non-OECD (column VIII) area5b: -0.643.

### VI. Interpretation and policy implications
- Overall inference:
  - Increases in the flexibility of labor market regulations and institutions have a statistically significant negative impact on both levels and changes of unemployment outcomes (total, youth, and long-term unemployment).
- Key channels and indicators:
  - Hiring and firing regulations and hiring costs are found to have the strongest effects on unemployment outcomes.
  - Military conscription (area5bvi) shows substantial negative associations with long-term and youth unemployment.
- Policy considerations:
  - Policies that enhance labor market flexibility should reduce unemployment.
  - Design and sequence of reforms matter, and complementary policies are needed to improve quality of employment and minimize possible negative short-term effects on inequality and job destruction (not investigated in this paper).
  - Protecting workers rather than jobs is important—coupling unemployment benefits with pressure on unemployed to take jobs and measures to help them (citing Blanchard, 2006).
  - Employment protection should internalize social costs and not inhibit job creation and labor reallocation.
  - Artificial restrictions on individual employment contracts should be avoided.

### VII. Key exact figures and statistics reported
- Selected exact magnitudes and statistics preserved from the source:
  - 1.58 percentage point reduction in youth unemployment (text header).
  - Composite index effects (static/dynamic and selected table coefficients):
    - Composite index (area5b) coefficients in Table 3: -0.569, -0.549, -0.472, -0.532, -0.569, -0.575, -0.193, -0.185, -0.835.
    - Table 6 sub-component static coefficients (selected): area5bi -0.182; area5bii -0.289; area5biii 0.139; area5biv -0.272; area5bv 0.002; area5bvi -0.053.
    - Table 7 youth unemployment: area5b -0.955; area5bii -0.494; area5biii 0.369; area5biv -0.124; area5bv -0.230; area5bvi -0.385.
    - Table 8 long-term unemployment: area5b -2.355; area5biv -1.415; area5bvi -0.992.
    - Table 9 dynamic full-sample area5b coefficients: -0.321, -0.303, -0.309, -0.231, -0.289, -0.289; OECD -0.108; Non-OECD -0.643.
    - Table 10 dynamic (selected): Unemployment area5b -0.321; Youth unemployment area5b -0.346; Long-term unemployment area5b -1.296.
    - Table 11 dynamic sub-components (selected): area5bii -0.214 (unemployment), -0.490 (youth unemployment); area5bv -0.136 (unemployment), -1.281 (long-term unemployment); area5bvi -0.057 (unemployment), -0.278 (youth unemployment), -0.884 (long-term unemployment).
  - Static magnitude examples:
    - A one standard deviation improvement in the composite labor market institutions index reduces long-term unemployment by 3.56 percentage points.
    - A one standard deviation improvement in the mandated cost of hiring index (area5biv) yields a 2.78 percentage point decrease in long-term unemployment.
    - A one standard deviation improvement in the conscription index (area5bvi) results in a 3.85 percentage point decrease in long-term unemployment.

### VIII. Annex — data description and sample (selected)
- Dependent and control variables categories:
  - Unemployment: Unemployment rate (lur, from WEO); Youth unemployment rate (unempyouth, from WDI); Long-term unemployment (unemplong, from WDI).
  - Macroeconomic variables: GDP per capita (rgdpl, from WEO); Demand pressure (gap_growth_n); Government size (lncg, from PWT); Openness (lnopenk, from PWT).
  - Demographic variables: Population size (lnpop, from PWT); Urbanization (lnurbpop, from WDI); Density (lnpopdens, from WDI).
  - Financial crisis: crisis dummy from Laeven and Valencia (2010).
- Country sample time coverage (selected entries preserved exactly):
  - Albania: 2002-2008
  - Algeria: 2002-2009
  - Argentina: 1985,1990,1995,2000-2008
  - Australia: 1990,1995,2000-2008
  - Austria: 1985,1990,1995,2000-2008
  - Belgium: 1985,1990,1995,2000-2008
  - Brazil: 1990,1995,2000-2008
  - Canada: 1985,1990,1995,2000-2008
  - China: [not listed in provided excerpt]
  - France: 1985,1990,1995,2000-2008
  - Germany: 1985,1990,1995,2000-2008
  - Japan: 1985,1990,1995,2000-2008
  - Korea: 1990-2008
  - Mexico: 1990,1995,2000-2008
  - Russia: 2000-2008
  - United Kingdom: 1985,1990,1995,2000-2008
  - United States: 1985,1990,1995,2000-2008
  - Vietnam: 2003-2008

*Source: _wp1264 (IMF working paper content provided).*

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

### _wp1264 - References

### I. Introduction — scope and contribution
- Sample and period:
  - Panel for 97 countries over the period 1985-2008.
- Main contribution:
  - Uses a broad sample (97 countries) to control for unobserved country-specific characteristics and assess evolution over time.
  - Focuses on a composite indicator of labor market flexibility (and its six sub-components).
  - Analyzes both static and dynamic effects and addresses possible endogeneity and reverse causality from unemployment to labor market institutions.
- Context from prior literature:
  - Prior cross-country work cited includes Botero et al. (2004) and Feldmann (2009).
  - Previous evidence generally finds that more rigid labor market institutions associate with higher unemployment.

### II. Data and indicators
- Labor market flexibility data:
  - Source: Fraser Institute’s Economic Freedom of the World (EFW) database.
  - All indicators standardized on a 0–10 scale (higher = more flexible labor market).
  - Composite index range (sample): minimum 1.8, maximum 9.5, total observations for composite index = 1214.
- Six sub-components of labor market flexibility (EFW area5b):
  - Minimum wage (area5bi) — based on World Bank’s Difficulty of Hiring.
  - Hiring and firing regulation (area5bii) — based on WEF’s Global Competitiveness Report.
  - Centralized collective wage bargaining (area5biii) — based on WEF’s Global Competitiveness Report.
  - Mandated cost of hiring (area5biv) — based on World Bank Doing Business data.
  - Mandated cost of worker dismissal (area5bv) — based on World Bank Doing Business data.
  - Conscription (area5bvi) — based on The Military Balance and War Resisters International data.
- Other data sources:
  - IMF’s World Economic Outlook (WEO), World Bank’s World Development Indicators (WDI), Penn World Table version 7.0, Laeven and Valencia (2010) financial crises database.
- Descriptive correlations:
  - Composite labor market flexibility indicator is negatively correlated with all unemployment outcomes.
  - Sub-indicators with stronger negative correlations on average: minimum wage and hiring and firing regulation.
  - Strongest negative correlation observed between mandated hiring costs and youth unemployment.
  - Unemployment is mostly concentrated among young people aged 15–24.

### III. Empirical methodology
- Two-part empirical strategy:
  - Static reduced-form specification regressing unemployment (total, youth, long-term) on the composite labor market flexibility index and controls.
  - Dynamic specification to test whether labor market flexibility affects the change in unemployment over time.
- Static model specification (equation structure preserved in source):
  - Unemployment rate regressed on labor market flexibility, vector of controls (output gap, ln government consumption to GDP (lncg), ln openness (lnopen), ln urban population (lnurbpop), ln population density (lnpopdens), crisis dummy, lagged unemployment U_{t-1}), and country fixed effects.
- Key control and behavioral findings:
  - Output gap included to control business cycle fluctuations.
  - Government size proxied by lncg (log of ratio of government consumption to GDP).
  - Crisis dummy equals 1 for occurrence of a financial crisis and 0 otherwise.
  - High persistence of unemployment is modeled via the lagged unemployment term.

### IV. Main empirical findings — static and dynamic effects
- Aggregate effects (total unemployment):
  - Increasing the rating of the composite labor market index by one standard deviation decreases, on average, the total unemployment rate by about 0.3–1.3 percentage points.
  - Baseline specification (only country fixed effects and output gap): a one standard deviation increase in the labor market index is associated with a 0.9 percentage point decrease in the unemployment rate.
  - Five-year averaged data: a one standard deviation increase in the labor market index decreases the unemployment rate by 1.3 percentage points (column IX).
- OECD vs non-OECD comparisons:
  - Non-OECD: a one standard deviation increase in the labor market institutions index is associated with a 0.96 percentage point decrease in unemployment.
  - OECD: the effect is about 0.86 percentage points.
  - The effect and persistence are not statistically different between OECD and non-OECD countries (95 percent confidence bands overlap).
- Persistence (“momentum”) of unemployment:
  - A one percentage point increase in previous unemployment translates into a 0.83 percentage point higher unemployment in the current period (0.82 when government size is included).
- Time evolution:
  - Recursive re-estimation over subperiods 1980–1990 through 1980–2008 (detailed in source) shows the effect of labor market flexibility on the level of unemployment has increased over time.
  - The effect becomes sizeable and statistically significant only after 2002 (noting caution due to an unbalanced panel with more observations after 2000).
- Youth and long-term unemployment:
  - Labor market flexibility has a strong and negative effect on youth and long-term unemployment, consistent with results for overall unemployment.
  - Youth unemployment: a one standard deviation increase in the composite labor markets index decreases youth unemployment by 1.41 percentage points.
  - Among sub-components:
    - Hiring and firing regulations index (area5bii): a one standard deviation increase results in a 0.78 percentage point decrease in youth unemployment.
    - Military conscription index (area5bvi): improvement (less conscription) yields a statistically significant negative effect on youth unemployment (magnitude reported in source beyond excerpt cut-off).

### V. Robustness and interpretation
- Robustness checks:
  - Results persist when controlling for macroeconomic and demographic variables.
  - Results hold for OECD and non-OECD subsamples.
  - Longer averaging (5-year) amplifies estimated effects, suggesting larger medium-term impacts.
- Interpretation and caveats:
  - Results suggest policies increasing labor market flexibility can substantially reduce unemployment levels and changes.
  - Caution on reverse causality: reforms often occur during weak economic performance and high unemployment, which could bias OLS estimates downward; addressed via dynamic specification and controls for endogeneity in the paper.
  - Effects on quality of employment, inequality, and short-term job destruction are not investigated in this analysis and should be considered when designing reforms.

### VI. Policy implications (as discussed in source)
- General implication:
  - Policies aimed at increasing labor market flexibility may have an important effect in reducing unemployment.
- Design considerations:
  - Reforms should be properly designed to improve quality of employment.
  - Reforms should seek to minimize possible negative short-term effects on inequality and job destruction (not investigated here).

*Source: _wp1264 - References (content from the supplied IMF PDF excerpt).*

### 1.58 percentage point reduction in youth unemployment. This may be interpreted as evidence

### _wp1264 - 1.58 percentage point reduction in youth unemployment. This may be interpreted as evidence

### Main findings on levels of unemployment and labor market institutions
- A one standard deviation improvement in the composite labor market institutions index reduces long-term unemployment by 3.56 percentage points.
- A one standard deviation improvement in the mandated cost of hiring index (area5biv) yields a 2.78 percentage point decrease in long-term unemployment.
- A one standard deviation improvement in the conscription index (area5bvi) results in a 3.85 percentage point decrease in long-term unemployment, implying that military conscription may have long lasting effects.
- For youth unemployment, the static results indicate:
  - A reported "1.58 percentage point reduction in youth unemployment" (text header).
  - In Table 7, the composite index (area5b) coefficient on youth unemployment is -0.955 (t-statistic indicates significance at **), implying a negative association between labor market flexibility and youth unemployment.
  - Sub-component effects on youth unemployment (Table 7):
    - area5bii (hiring and firing regulation) coefficient: -0.494 (significant at **).
    - area5bvi (conscription) coefficient: -0.385 (significant at **).
    - area5biii (degree of decentralization in wage bargaining) coefficient: 0.369 (positive, not significantly different from zero).

### Dynamic relationship: changes in unemployment
- Dynamic reduced-form specification estimated with two-step GMM-system estimator to address endogeneity and reverse causality.
- Improvements in labor market regulations that allow for higher flexibility have a statistically significant negative effect on changes in unemployment.
- Point estimates and interpretations (Table 9 and Table 10):
  - Increasing the rating of the composite labor market index by one standard deviation decreases, on average, the change in the total unemployment rate by about 0.35‒0.49 percentage points.
  - Baseline specification suggests a one standard deviation increase in the labor market index is associated with a half percentage point decrease in the annual change in unemployment.
  - Table 10 dynamic results:
    - Unemployment (area5b): -0.321 (significant at ***).
    - Youth unemployment (area5b): -0.346 (significant at *) — interpreted as about half percentage point decrease in youth unemployment per one standard deviation increase in the composite index.
    - Long-term unemployment (area5b): -1.296 (significant at **), interpreted as nearly 2 percentage points decrease in long-term unemployment per one standard deviation increase in the composite index.
- Sub-component dynamic effects on changes in unemployment (Table 11):
  - Hiring and firing regulations (area5bii) have the strongest and most statistically significant negative effect on changes in overall unemployment and youth unemployment: area5bii coefficient on unemployment: -0.214 (significant at *); on youth unemployment: -0.490 (significant at **).
  - Mandated costs of hiring (area5bv) show statistically significant effects: area5bv coefficient on unemployment change: -0.136 (significant at **); area5bv coefficient on long-term unemployment: -1.281 (significant at **).
  - Military conscription (area5bvi) has negative and statistically significant impact on change in both youth and long-term unemployment (area5bvi on youth unemployment: -0.278, significant at ***; on long-term unemployment: -0.884, significant at **), but no significant effect on changes in the total unemployment rate in some specifications.

### Robustness, tests, and subgroup results
- Consistency checks and diagnostic tests for GMM estimates:
  - Hansen J-test p-values across specifications range from 0.2 to 0.4 (cannot reject validity of instruments).
  - Arellano–Bond AR(2) test p-values across specifications range from 0.6 to 0.9 (cannot reject no second-order serial correlation).
- OECD vs non-OECD comparisons (Table 9):
  - Point estimate of the coefficient of labor market institutions is higher for non-OECD countries than for OECD countries, but estimated coefficients are not statistically different between the two groups.
  - Table 9 full-sample dynamic area5b estimates include:
    - Full-sample area5b coefficients range (columns I–VI): -0.321, -0.303, -0.309, -0.231, -0.289, -0.289 (all significant at ***/**).
    - OECD (column VII) area5b: -0.108 (significant at **).
    - Non-OECD (column VIII) area5b: -0.643 (significant at ***).

### Interpretation and policy implications
- Overall evidence indicates that increases in the flexibility of labor market regulations and institutions have a statistically significant negative impact both on the level and the change of unemployment outcomes (total, youth, and long-term unemployment).
- Among labor market flexibility indicators analyzed, hiring and firing regulations and hiring costs are found to have the strongest effects on unemployment outcomes.
- Policy considerations raised by the paper:
  - Policies that enhance labor market flexibility should reduce unemployment.
  - Design and sequence of reforms matter, and complementary policies are needed to improve quality of employment and minimize possible negative short-term effects (not investigated in this paper) on inequality and job destruction.
  - Protecting workers rather than jobs is important—coupling unemployment benefits with pressure on unemployed to take jobs and measures to help them (citing Blanchard, 2006).
  - Employment protection should internalize social costs and not inhibit job creation and labor reallocation.
  - Artificial restrictions on individual employment contracts should be avoided.

### Key exact figures and statistics reported in the paper
- Static and dynamic effect magnitudes and statistics (preserved exactly as reported):
  - 1.58 percentage point reduction in youth unemployment (text header).
  - Long-term unemployment reduction from composite index: 3.56 percentage points.
  - Mandated cost of hiring (area5biv) long-term unemployment decrease: 2.78 percentage points.
  - Conscription (area5bvi) long-term unemployment decrease: 3.85 percentage points.
  - Dynamic decrease in total unemployment change per one standard deviation increase in composite index: about 0.35‒0.49 percentage points; baseline indicates "half percentage point" decrease.
  - Table coefficients (selected, exact):
    - Table 3 area5b coefficients: -0.569, -0.549, -0.472, -0.532, -0.569, -0.575, -0.193, -0.185, -0.835 (t-statistics shown and significance stars in table).
    - Table 6 sub-component static coefficients (selected): area5bi -0.182; area5bii -0.289; area5biii 0.139; area5biv -0.272; area5bv 0.002; area5bvi -0.053.
    - Table 7 youth unemployment: area5b -0.955; area5bii -0.494; area5biii 0.369; area5biv -0.124; area5bv -0.230; area5bvi -0.385.
    - Table 8 long-term unemployment: area5b -2.355; area5biv -1.415; area5bvi -0.992.
    - Table 9 dynamic full-sample area5b coefficients: -0.321, -0.303, -0.309, -0.231, -0.289, -0.289; OECD -0.108; Non-OECD -0.643.
    - Table 10 dynamic: Unemployment area5b -0.321; Youth unemployment area5b -0.346; Long-term unemployment area5b -1.296.
    - Table 11 dynamic sub-components (selected): area5bii -0.214 (unemployment), -0.490 (youth unemployment); area5bv -0.136 (unemployment), -1.281 (long-term unemployment); area5bvi -0.057 (unemployment), -0.278 (youth unemployment), -0.884 (long-term unemployment).
  - Diagnostic test p-values:
    - Hansen J-test p-value ranges reported: 0.2 to 0.4 (text); specific Table 9 entries: 0.20, 0.31, 0.32, 0.35, 0.43, 0.32, 0.31, 0.25.
    - Arellano–Bond AR(2) test p-values reported: 0.6 to 0.9 (text); specific Table 9 entries: 0.58, 0.57, 0.63, 0.57, 0.60, 0.63, 0.57, 0.89.

*Source: _wp1264 - 1.58 percentage point reduction in youth unemployment. This may be interpreted as evidence (IMF working paper content provided).*

### References

### _wp1264 - References

### References
- Bassanini, A. and Duval, R., 2006, ‘Employment Patterns in OECD Countries: Reassessing the Role of Policies and Institutions,” OECD Economics Department Working Papers No. 486.
- ——, 2009, “Unemployment, Institutions, and Reform Complementarities: Re-assessing the Aggregate Evidence for OECD Countries,” Oxford Review of Economic Policy No. 25, pp. 40–59.
- Blanchard, O. and Wolfers, J., 2000, “The Role of Shocks and Institutions in the Rise of European Unemployment: The Aggregate Evidence,” Economic Journal No. 110, Vol. 462, pp. 1–33.
- Blanchard, O., 2006, “European Unemployment: The Evolution of Facts and Ideas,” Economic Policy No. 21 Vol. 45, pp. 5–59.
- Belot, M. and van Ours, J.C., 2004, “Does the Recent Success of Some OECD Countries in Lowering Their Unemployment Rates Lie in the Clever Design of Their Labor Market Reforms?,” Oxford Economic Papers No. 56, Vol. 4, pp. 621–42.
- Bernal-Verdugo, L.E., Furceri, D. and Guillaume, D., 2012, “Crises, Reforms, and Unemployment,” IMF Working Papers (Forthcoming).
- Botero, J. C., Djankov, S., La Porta, R., et al., 2004, The Regulation of Labor,” Quarterly Journal of Economics No. 119 Vol. 4, pp. 1339–82.
- Elmeskov, J., Martin, J.P., and Scarpetta, S., 1998, Key Lessons for Labour Market Reforms: Evidence from OECD Countries’ Experiences,” Swedish Economic Policy, Review No. 5 Vol. 2, pp. 205–52.
- Feldmann, H., 2006, Government Size and Unemployment: Evidence from Industrial Countries, Public Choice, No. 127 Vol. 3–4, pp. 451–67.
- Feldmann, H., 2009, The Unemployment Effects of Labor Regulation around the World, Journal of Comparative Economics No. 37 Vol. 1, pp. 76–90.
- Heston, A., Summers, R. and Aten, B., 2011, Penn World Table Version 7.0, Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania.
- Laeven, L. and Valencia, F., 2010, “Resolution of Banking Crises: The Good, the Bad, and the Ugly,” International Monetary Fund Working Paper No 146.
- Nickell, S., 1997, “Unemployment and Labor Market Rigidities: Europe versus North America,” Journal of Economic Perspectives No. 11 Vol. 3, pp. 55–74.
- Nickell, S., 1998, “Unemployment: Questions and Some Answers,” Economic Journal No. 108, pp. 802–16.
- Nunziata, L., 2002, “Unemployment, Labour Market Institutions and Shocks,” Nuffield College Working Papers in Economics No. 16.
- Roodman, D., 2009a, “How to do xtabond2: An introduction to difference and system GMM in Stata,” Stata Journal, No. 9 Vol. 1, pp. 86–136.
- Roodman, D., 2009b, “A Note on the Theme of Too Many Instruments, Oxford Bulletin of Economics and Statistics, No. 71, Vol. 1, pp. 135–58.
- Scarpetta, S., 1996, “Assessing the Role of Labour Market Policies and Institutional Settings on Unemployment: A Cross-country Study,” OECD Economic Studies, No. 26, pp. 43–98.

### Annex I: Data Description
- The dependent and control variables included in the analysis belong to one of several categories, namely:
  1) Unemployment
    - Unemployment rate (lur, from WEO): Percentage of the total labor force that is currently unemployed.
    - Youth unemployment rate (unempyouth, from WDI): Percentage of the total labor force of ages 15 to 24 that is currently unemployed.
    - Long-term unemployment (unemplong, from WDI): Fraction (in percent) of the unemployment rate that is of long-term.
  2) Macroeconomic variables
    - GDP per capita (rgdpl, from WEO): Purchasing power parity (PPP) converted GDP per capita (with the Laspeyres methodology), derived from growth rates of private consumption, government expenditures, and investment at 2005 constant prices.
    - Demand pressure (gap_growth_n): Gap in the current real GDP per capita growth with respect to a moving average of n years, centered at the current period.
    - Government size (lncg, from PWT): (log) Government consumption share of PPP converted GDP per capita at current prices, in percent.
    - Openness (lnopenk, from PWT): (log) Openness at 2005 constant prices, in percent.
  3) Demographic variables
    - Population size (lnpop, from PWT): (log) Total population (in thousands).
    - Urbanization (lnurbpop, from WDI): (log) Urban population, as percent of total population.
    - Density (lnpopdens, from WDI): (log) Population density, measured by the number of people per square kilometer of land area.
  4) Financial crisis
    - Financial crisis indicator (crisis): This dummy variable assigns a value of 1 to years in which a country was going through a financial crisis according to Laeven and Valencia (2010), and 0 otherwise.

### Annex—Data Sample
- Country samples with time coverage (as listed):
  - Albania: 2002-2008
  - Algeria: 2002-2009
  - Argentina: 1985,1990,1995,2000-2008
  - Armenia: 2004-2008
  - Australia: 1990,1995,2000-2008
  - Austria: 1985,1990,1995,2000-2008
  - Azerbaijan: 2004-2008
  - Barbados: 2005-2008
  - Barbados entry appears once; refer to original list for placement.
  - Belarus entry not present in source.
  - Belgium: 1985,1990,1995,2000-2008
  - Belize: 2004-2008
  - Bosnia & Herz.: 2004-2008
  - Brazil: 1990,1995,2000-2008
  - Bulgaria: 2000-2008
  - Canada: 1985,1990,1995,2000-2008
  - Chile: 1995,2000-2008
  - The Bahamas: 2006-2008
  - Colombia: 1995,2000-2008
  - Costa Rica: 2000-2008
  - Croatia: 2000-2008
  - Cyprus: 2003-2008
  - Czech Rep.: 2000-2008
  - Denmark: 1985,1990,1995,2000-2008
  - Dominican Rep.: 1985,1990,1995,2000-2008
  - Ecuador: 2000-2008
  - El Salvador: 2000-2008
  - Estonia: 2000-2008
  - Finland: 1985,1990,1995,2000-2008
  - Fiji: 2002-2008
  - France: 1985,1990,1995,2000-2008
  - Georgia: 1985,1990,1995,2000-2008
  - Germany: 1985,1990,1995,2000-2008
  - Hong Kong: 1990,1995,2000-2008
  - Hungary: 1995,2000-2008
  - Iceland: 1990,1995,2000-2008
  - Indonesia: 1995,2000-2008
  - Iran: 2000-2008
  - Israel: 2000-2008
  - Italy: 1985,1990,1995,2000-2008
  - Jamaica: 2000-2008
  - Japan: 1985,1990,1995,2000-2008
  - Jordan: 1990,1995,2000-2008
  - Kazakhstan: 2005-2008
  - Korea: 1990-2008
  - Kuwait: 2002-2008
  - Kyrgyz Rep.: 2005-2008
  - Latvia: 2000-2008
  - Lithuania: 2000-2008
  - Luxembourg: 1990,1995,2000-2008
  - Malaysia: 1990,1995,2000-2008
  - Malta: 2002-2008
  - Mauritius: 2000-2008
  - Mexico: 1990,1995,2000-2008
  - Moldova: 2005-2008
  - Mongolia: 2000-2008
  - Montenegro: 2005-2008
  - Morocco: 2001-2008
  - Netherlands: 1985,1990,1995,2000-2008
  - New Zealand: 1985,1990,1995,2000-2008
  - Nicaragua: 2000-2008
  - Nigeria: 2000-2008
  - Norway: 1985,1990,1995,2000-2008
  - Panama: 2000-2008
  - Paraguay: 2000-2008
  - Peru: 1995,2000-2008
  - Philippines: 1990,1995,2000-2008
  - Poland: 1990,1995,2000-2008
  - Portugal: 1985,1990,1995,2000-2008
  - Romania: 2000-2008
  - Russia: 2000-2008
  - Serbia: 2006-2008
  - Singapore: 1990,1995,2000-2008
  - Slovenia: 2000-2008
  - Slovak Republic: 1995,2000-2008
  - South Africa: 1990,1995,2000-2008
  - Spain: 1985,1990,1995,2000-2008
  - Sri Lanka: 2000-2008
  - Sweden: 1985,1990,1995,2000-2008
  - Switzerland: 1985,1990,1995,2000-2008
  - Taiwan: 1990,1995,2000-2008
  - Thailand: 1990,1995,2000-2008
  - Tunisia: 2001-2008
  - Turkey: 1990,1995,2000-2008
  - Trinidad & Tobago: 2000-2008
  - Ukraine: 2000-2008
  - United Kingdom: 1985,1990,1995,2000-2008
  - United States: 1985,1990,1995,2000-2008
  - Uruguay: 2000-2008
  - Venezuela: 1990,1995,2000-2008
  - Vietnam: 2003-2008

*Source: _wp1264 - References (PDF).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1264.pdf_