1 Payroll Tax Legislation 2. 2 Severance Payments Legislation 3

Transcription

1 Web Appendix Contents 1 Payroll Tax Legislation 2 2 Severance Payments Legislation 3 3 Difference-in-Difference Results Senior Workers, 1997 Change Young Workers, 2001 Change Results for Men and Women 10 5 Additional Evidence on Substitution After the 30th Threshold /1971 cohort vs. older cohorts No rebound in employment of cohorts older than Graphical evidence around each policy change Age Distribution of Hires Before and After 1997 Change Wage Incidence Exploiting Changes in Generosity 31 7 Self-Employed and Public Workers around Age of Age Distribution of Self-Employment and Public Jobs 35 1

2 1 Payroll Tax Legislation The following is a list of Spain s legislation concerning payroll taxes: Real-Decreto Ley 9/1997 de 16 de mayo. Ley 64/1997 de 26 de diciembre. Ley 50/1998 Ley 55/1999 de 29 de diciembre. Ley 14/2000. Real-Decreto Ley 5/2001 de 2 de marzo. Ley 12/2001 de 9 de julio. Ley 24/2001 Ley 53/2002 Ley 62/2003 Ley 2/2004 Ley 30/2005 Real-Decreto Ley 5/2006 de 9 de julio. Ley 43/2006 de 29 de diciembre. 2

3 Table 1: Payroll Tax Cuts Programs Unemployed people 1/1/1984-8/4/1992 8/5/1992-5/16/ /17/ /16/ /17/ /31/1999 1/1/2000-3/3/2001 3/4/2001-6/31/2006 7/1/2006-9/18/2010 < 25 y, and 1 y registered as un- Less 30 years More than a year unemployed More than 6 months unemployed Older 44 years years years Those still entitled to 1 year of UI or UA Less 45 years women Women Only workers 1 y registered as unemployed. Subsidy of 2400 e. 50% all contract. employed,; or y if had not worked for more than 3 m e subsidy. Only workers 1 y registered as unemployed. Subsidy of 3000 e. 50% all contract. 40%, 24 months. 40%, 24 months (60% for women). 60%, first 2 years, 50% after. 35% first year, 25% second year (+5% women). 40% first year, 30% second year (+5% women). 45% first year, 40% after (+5% women). 20%, 24 months. 50%, first year, 45% second year. 50% first year, 45% after. 20%, 24 months. 50% first year, 45% after. 55% first year, 50% after. 50% first year, 45% second year. 25% two years. 800 euros/year, 4 years. 600 euros/year, 4 years euros/year, all years. 600 euros/year, 4 years. Notes: The table is a summary of the main payroll tax credits available in each year. The sources are the laws enumerated in section 1 2 Severance Payments Legislation Severance payments in Spain are a function of the wage and work tenure of the worker and of whether the lay off is part of a collective dismissal, and in case it is an individual layoff, it depends on whether it is declared wrongful or not by the courts (3/4 of all layoffs are considered wrongful by the judges). For the case of collective dismissals, the employee has to be compensated for 20 days of salary for each year of work tenure in the firm, up to 12 months of salary. The same rule applies for lawful layoffs. Finally, dismissals considered wrongful are compensated with 45 days of salary for year of work tenure in the firm, up to 42 months of salary. In order to increase permanent employment, several policies have been implemented allowing the employer to sign a contract with reduced severance payments in case of wrongful dismissal. Table 2 summarizes the reductions in severance payments available for employers. For people hired under these contracts, in case of layoff, they have to be compensated with 33 days of wage for each 3

4 Table 2: Reduced Severance Payments Programs 5/17/97-6/10/ /10/ /4/2001 Less 30 years. More than a year unemployed. Less 30 years (included). Less 30 years (included). More than 6 More than 6 months unemployed. months unemployed. 45 years or older. 45 years or older. 45 years or older. Disabled. Disabled. Disabled. Women hired in sectors in which Women hired in sectors in which they are underrepresentedrepresented. they are under- If temp. workers If temp. workers If temp. workers with firm up to 5/17/1998. with firm up to 12/31/2003. with firm up to 12/31/2007. Notes: The table is a summary of the main severance payments available in each year. The sources are the laws enumerated in the previous page. year of work tenure in the firm, up to 24 months of salary, if the layoff is declared wrongful by the courts. These policies have been targeted to young people, people who had been unemployed for a long term, and individuals older than 45 years. The policy targeted also the conversion of temporary contracts into permanent ones. Employers could benefit from contracts with reduced severance payments if they converted into permanent employees people who hold a temporary contract with the firm before a certain time cutoff and whose temporary contract did not expire. 4

5 3 Difference-in-Difference Results 3.1 Senior Workers, 1997 Change 5

6 Table 3: Effects on Transitions and Employment, Vs , 1997 (1) (2) (3) (4) (5) (6) (7) Perm. Perm. Perm. Temp. Temp. Temp. Trans. Temp. Hires Lay-Offs Quits Hires Lay-Offs Quits to Perm. Treatment ** *** *** *** *** *** (0.0254) (0.0160) ( ) (0.0503) (0.0356) (0.0123) ( ) Post 0.240*** 0.494*** 0.103*** 0.394*** *** *** (0.0205) (0.0169) ( ) (0.0315) (0.0340) (0.0126) (0.0102) Treatment x Post 0.137*** *** *** * (0.0292) (0.0248) (0.0115) (0.0493) (0.0479) (0.0173) (0.0136) Observations Effects on Employment Perm. Short-Term Perm. or ST Self-Emp. Public UI Employed Treatment ** 12.28*** *** (4.003) (2.977) (3.508) (2.811) (2.197) (1.827) (2.122) Post 61.56*** *** *** 5.802*** *** (4.341) (3.231) (3.819) (3.046) (2.378) (2.034) (2.344) Treatment x Post * * (0.206) (0.152) (0.181) (0.145) (0.113) (0.0955) (0.110) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. In the upper panel, the dependent variable is a dummy indicating whether a transition in or out of a permanent contract (columns 1-3) or a temporary contract (columns 4-6) happened. In column 7th the dependent variable is a dummy that indicates that the worker was converted to permanent, from a temporary contract, within the same firm. The lower panel reports the results for employment. Each dependent variable is one of the possible employment statuses. The policy change is the expansion of payroll tax cuts in The specification is equation 2. Control variables included are calendar quarter dummies, sex, education, experience, firm size, sector, citizenship, part-time job, and workers disability. Columns 1-3 report the results for young workers, and columns 4-6 for prime-age workers. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 6

7 3.2 Young Workers, 2001 Change In 2001, the tax credit for workers younger than 30 was removed. I perform a difference-in-difference estimation to check whether the results based on the 1997 policy change still hold. In addition, the policy change in 2001 allows to test for whether the policy is still having an effect 4 years after its implementation. The results are analogous to those obtained for They are summarized in table 4. The number of new permanent hires decreases by 0.26 pp after the policy change. Likewise, the number of temporary hires decreases by 0.11 pp. The number of transitions from temporary to permanent contracts also decrease by 0.27 pp. In terms of stocks, after the 2001 policy change there is a reduction in employment of pp, almost matched by an increase in UI recipients of pp. 7

8 Figure 1: Difference-in-Difference Results for the Impact of the Removal of Payroll Tax Cuts on Transitions, Young Workers Coefficient Coefficient Coefficient Impact of Payroll Tax Cut on Permanent Hires Treatment: 25 30, Control: Quarter Distance Diff in Diff Coefficient uci (a) Impact of Payroll Tax Cut on Trans. from Temp. to Perm. March 2001 Policy Change Treatment: 25 30, Control: Quarter Distance Diff in Diff Coefficient uci (b) Impact of Payroll Tax Cut on Permanent Employment Treatment: 25 30, Control: lci 3 lci Coefficient Coefficient Impact of Payroll Tax Cut on Permanent or ST Employment Treatment: 25 30, Control: Quarter Distance Diff in Diff Coefficient uci (d) Impact of Payroll Tax Cut on Employment Treatment: 25 30, Control: Quarter Distance Diff in Diff Coefficient uci (e) lci 3 lci Quarter Distance Diff in Diff Coefficient uci lci (c) Notes: The graphs display coefficients (blue solid line) from regressions dummies indicating whether a transition occurred (hire, lay-off or quit) on time period dummies interacted with a treatment indicator (25-30 years old). Thus, the coefficients are difference-in-difference estimates in each time period. The omitted period is the quarter of the policy change. The specification is equation 3 and it includes time-period dummies, age dummies, and controls sex, education, previous wage, experience, firm size, sector, citizenship, part-time job, and workers disability. The red dashed represents the quarter when the policy was removed, and the blue dashed line are upper and lower confidence intervals at the 95% level. 8

9 Table 4: Effects on Transitions and Employment, Vs , 2001 (1) (2) (3) (4) (5) (6) (7) Perm. Perm. Perm. Temp. Temp. Temp. Trans. Temp. Hires Lay-Offs Quits Hires Lay-Offs Quits to Perm. Treatment 0.878*** *** 0.317*** 2.232*** 2.326*** 0.652*** 0.455*** (0.0365) (0.0144) (0.0128) (0.140) (0.0359) (0.0178) (0.0177) Post *** 0.109*** 0.114*** 0.253*** *** *** *** (0.0211) (0.0147) (0.0123) (0.0349) (0.0333) (0.0165) ( ) Treatment x Post *** ** ** *** *** (0.0328) (0.0199) (0.0182) (0.0449) (0.0483) (0.0245) (0.0138) Observations Effects on Employment Perm. Short-Term Perm. or ST Self-Emp. Public UI Employed Treatment *** 22.05*** * * 3.893*** (4.987) (7.919) (5.665) (3.371) (2.166) (1.446) (1.909) Post 43.66*** *** *** 16.83*** (5.661) (11.59) (8.342) (4.714) (2.772) (1.934) (2.584) Treatment x Post ** * ** * (0.186) (0.221) (0.149) (0.0870) (0.0612) (0.0589) (0.0819) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. In the upper panel, the dependent variable is a dummy indicating whether a transition in or out of a permanent contract (columns 1-3) or a temporary contract (columns 4-6) happened. In column 7th the dependent variable is a dummy that indicates that the worker was converted to permanent, from a temporary contract, within the same firm. The lower panel reports the results for employment. Each dependent variable is one of the possible employment statuses. The policy change is the removal of youth payroll tax cuts in The specification is equation 2. Control variables included are calendar quarter dummies, sex, education, experience, firm size, sector, citizenship, part-time job, and workers disability. Columns 1-3 report the results for young workers, and columns 4-6 for prime-age workers. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 9

10 4 Results for Men and Women If the null employment effect at 45 is purely driven by an inelastic demand, we expect to find a zero employment effect for any labor supply elasticity. Women are known to have more elastic supply (Eissa and Liebman, 1996; Blundell and MaCurdy, 1999). Thus, in case labor demand at 45 is not perfectly inelastic we should detect some positive employment responses for women. I test this hypothesis by splitting the sample by gender. Tables 5 and 6 report the results of the difference-in-difference analysis. As can be seen, there is a null effect on employment on both cases. Tables 7 and 8 show the estimates of the RDD. As before, there is no effect on employment (panels C and D, column 6). Note that the rest of results found for the whole sample of prime-age workers also hold both for men and women. After the 1997 policy change, there is an increase in permanent hires, and a decrease in temporary hires (tables 5 and 6, columns 1 and 4, upper panel). And the jump in hiring at 45 is larger in the long-run (tables 7 and 8, panel A and B). For the case of young workers, tables 9 and 10 report the difference-in-difference estimates after the policy change in The employment effect is larger for women than for men. However, note that the effect is larger basically because the subsidized permanent contracts stimulate also temporary employment for women. In the long-run, the estimate is larger for men since the subsidy for permanent contracts does not stimulate temporary employment any more. Tables 11 and 12 show the estimates in the long-run (based on the regression discontinuity evidence). 10

11 Table 5: Effects on Transitions and Employment for Men, Vs , 1997 (1) (2) (3) (4) (5) (6) (7) Perm. Perm. Perm. Temp. Temp. Temp. Trans. Temp. Hires Lay-Offs Quits Hires Lay-Offs Quits to Perm. Treatment (0.0374) (0.0221) (0.0109) (0.0619) (0.0498) (0.0178) (0.0110) Post (0.0272) (0.0234) (0.0122) (0.0470) (0.0483) (0.0186) (0.0143) Treatment x Post (0.0399) (0.0339) (0.0170) (0.0649) (0.0669) (0.0251) (0.0183) Observations Effects on Employment Perm. Short-Term Perm. or ST Self-Emp. Public UI Employed Treatment (5.539) (4.154) (4.709) (3.875) (2.568) (2.499) (2.836) Post (6.028) (4.524) (5.141) (4.213) (2.778) (2.789) (3.144) Treatment x Post (0.286) (0.212) (0.244) (0.200) (0.132) (0.131) (0.148) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. The sample is restricted to men. In the upper panel, the dependent variable is a dummy indicating whether a transition in or out of a permanent contract (columns 1-3) or a temporary contract (columns 4-6) happened. In column 7th the dependent variable is a dummy that indicates that the worker was converted to permanent, from a temporary contract, within the same firm. The lower panel reports the results for employment. Each dependent variable is one of the possible employment statuses. The policy change is the expansion of payroll tax cuts in The specification is equation 2. Control variables included are calendar quarter dummies, education, experience, firm size, sector, citizenship, part-time job, and workers disability. Columns 1-3 report the results for young workers, and columns 4-6 for prime-age workers. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 11

12 Table 6: Effects on Transitions and Employment for Women, Vs , 1997 (1) (2) (3) (4) (5) (6) (7) Perm. Perm. Perm. Temp. Temp. Temp. Trans. Temp. Hires Lay-Offs Quits Hires Lay-Offs Quits to Perm. Treatment (0.0330) (0.0228) ( ) (0.0608) (0.0497) (0.0160) (0.0130) Post (0.0292) (0.0242) ( ) (0.0465) (0.0467) (0.0159) (0.0132) Treatment x Post (0.0417) (0.0359) (0.0145) (0.0734) (0.0668) (0.0226) (0.0193) Observations Effects on Employment Perm. Short-Term Perm. or ST Self-Emp. Public UI Employed Treatment (5.629) (4.072) (5.186) (3.948) (3.797) (2.645) (3.177) Post (6.049) (4.379) (5.611) (4.239) (4.103) (2.936) (3.489) Treatment x Post (0.290) (0.206) (0.268) (0.204) (0.195) (0.138) (0.165) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. The sample is restricted to women. In the upper panel, the dependent variable is a dummy indicating whether a transition in or out of a permanent contract (columns 1-3) or a temporary contract (columns 4-6) happened. In column 7th the dependent variable is a dummy that indicates that the worker was converted to permanent, from a temporary contract, within the same firm. The lower panel reports the results for employment. Each dependent variable is one of the possible employment statuses. The policy change is the expansion of payroll tax cuts in The specification is equation 2. Control variables included are calendar quarter dummies, education, experience, firm size, sector, citizenship, part-time job, and workers disability. Columns 1-3 report the results for young workers, and columns 4-6 for prime-age workers. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 12

13 Table 7: Effects on Transitions and Employment of Crossing the 45th Birthday Threshold, Men (1) (2) (3) (4) (5) (6) Perm. Short-Term Self-Employed UI Public Worker Panel A: Effects on Transitions, Short-Run, RDD Treatment (0.0872) (0.109) (0.0370) (0.0796) (0.0323) Observations Panel B: Effects on Transitions, Long-Run, RDD Perm. Short-Term Self-Employed UI Public Worker Treatment (0.0270) (0.0372) (0.0113) (0.0245) ( ) Observations Perm. Short-Term Perm. or ST Self-Employed Public Worker Non-Employed Panel C: Effects on Employment, RDD Treatment (0.0549) (0.0726) (0.0709) (0.0306) (0.0118) (0.0720) Slope Before ( ) ( ) ( ) ( ) ( ) ( ) Slope After (0.0107) (0.0137) (0.0137) ( ) ( ) (0.0137) Observations Panel D: Effects on Employment, Donut RDD Treatment (0.196) (0.236) (0.241) (0.119) (0.0419) (0.242) Slope Before ( ) ( ) ( ) ( ) ( ) ( ) Slope After (0.0105) (0.0125) (0.0127) ( ) ( ) (0.0126) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. The sample is restricted to men. Panel A and B report the effects on entries into permanent, short-term, self-employed, or public jobs, as well as into UI, around the 45th birthday thresholds. Panel A focuses on the short-run effects (first 12 months after policy). Panel B focuses on the long-run effects (from the 12th month until policy is removed). Panels C and D display the effects on the probability of being employed in any of these contracts and the overall probability of working. The specification is equation 1. The dependent variables are dummies indicating each type of transition (panels A and B), or the employment status (panels C and D). The estimation window is of 24 months on each side of the threshold. Panel D performs a donut RDD omitting the 12 months before and after the 45th birthday. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 13

14 Table 8: Effects on Transitions and Employment of Crossing the 45th Birthday Threshold, Women (1) (2) (3) (4) (5) (6) Perm. Short-Term Self-Employed UI Public Worker Panel A: Effects on Transitions, Short-Run, RDD Treatment (0.0659) (0.0835) (0.0346) (0.0697) (0.0786) Observations Panel B: Effects on Transitions, Long-Run, RDD Perm. Short-Term Self-Employed UI Public Worker Treatment (0.0258) (0.0484) ( ) (0.0291) (0.0167) Observations Perm. Short-Term Perm. or ST Self-Employed Public Worker Non-Employed Panel C: Effects on Employment, RDD Treatment (0.0617) (0.0837) (0.0884) (0.0312) (0.0200) (0.0917) Slope Before ( ) ( ) (0.0102) ( ) ( ) (0.0105) Slope After (0.0117) (0.0151) (0.0159) ( ) ( ) (0.0164) Observations Panel D: Effects on Employment, Donut RDD Treatment (0.217) (0.271) (0.290) (0.118) (0.0749) (0.301) Slope Before ( ) (0.0103) (0.0110) ( ) ( ) (0.0114) Slope After (0.0116) (0.0142) (0.0151) ( ) ( ) (0.0156) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. The sample is restricted to women. Panel A and B report the effects on entries into permanent, short-term, self-employed, or public jobs, as well as into UI, around the 45th birthday thresholds. Panel A focuses on the short-run effects (first 12 months after policy). Panel B focuses on the long-run effects (from the 12th month until policy is removed). Panels C and D display the effects on the probability of being employed in any of these contracts and the overall probability of working. The specification is equation 1. The dependent variables are dummies indicating each type of transition (panels A and B), or the employment status (panels C and D). The estimation window is of 24 months on each side of the threshold. Panel D performs a donut RDD omitting the 12 months before and after the 45th birthday. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 14

15 Table 9: Effects on Transitions and Employment for Men, Vs , 1997 (1) (2) (3) (4) (5) (6) (7) Perm. Perm. Perm. Temp. Temp. Temp. Trans. Temp. Hires Lay-Offs Quits Hires Lay-Offs Quits to Perm. Treatment (0.0311) (0.0179) (0.0103) (0.0956) (0.0500) (0.0206) (0.0156) Post (0.0273) (0.0204) (0.0125) (0.0457) (0.0474) (0.0199) (0.0132) Treatment x Post (0.0410) (0.0265) (0.0172) (0.0776) (0.0676) (0.0294) (0.0217) Observations Effects on Employment Perm. Short-Term Perm. or ST Self-Emp. Public UI Employed Treatment (0.0797) (0.0756) (0.0692) (0.0508) (0.0297) (0.0392) (0.0485) Post (0.0822) (0.0727) (0.0704) (0.0537) (0.0334) (0.0378) (0.0460) Treatment x Post (0.107) (0.102) (0.0921) (0.0686) (0.0396) (0.0494) (0.0632) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. The sample is restricted to men. In the upper panel, the dependent variable is a dummy indicating whether a transition in or out of a permanent contract (columns 1-3) or a temporary contract (columns 4-6) happened. In column 7th the dependent variable is a dummy that indicates that the worker was converted to permanent, from a temporary contract, within the same firm. The lower panel reports the results for employment. Each dependent variable is one of the possible employment statuses. The policy change is the expansion of payroll tax cuts in The specification is equation 2. Control variables included are calendar quarter dummies, education, experience, firm size, sector, citizenship, part-time job, and workers disability. Columns 1-3 report the results for young workers, and columns 4-6 for prime-age workers. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 15

16 Table 10: Effects on Transitions and Employment for Women, Vs , 1997 (1) (2) (3) (4) (5) (6) (7) Perm. Perm. Perm. Temp. Temp. Temp. Trans. Temp. Hires Lay-Offs Quits Hires Lay-Offs Quits to Perm. Treatment (0.0318) (0.0181) ( ) (0.116) (0.0471) (0.0169) (0.0116) Post (0.0277) (0.0203) ( ) (0.0459) (0.0442) (0.0159) (0.0104) Treatment x Post (0.0382) (0.0275) (0.0137) (0.0687) (0.0650) (0.0246) (0.0183) Observations Effects on Employment Perm. Short-Term Perm. or ST Self-Emp. Public UI Employed Treatment (0.0732) (0.0697) (0.0704) (0.0423) (0.0399) (0.0449) (0.0544) Post (0.0749) (0.0665) (0.0720) (0.0454) (0.0447) (0.0443) (0.0530) Treatment x Post (0.100) (0.0960) (0.0947) (0.0573) (0.0531) (0.0588) (0.0726) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. The sample is restricted to women. In the upper panel, the dependent variable is a dummy indicating whether a transition in or out of a permanent contract (columns 1-3) or a temporary contract (columns 4-6) happened. In column 7th the dependent variable is a dummy that indicates that the worker was converted to permanent, from a temporary contract, within the same firm. The lower panel reports the results for employment. Each dependent variable is one of the possible employment statuses. The policy change is the expansion of payroll tax cuts in The specification is equation 2. Control variables included are calendar quarter dummies, education, experience, firm size, sector, citizenship, part-time job, and workers disability. Columns 1-3 report the results for young workers, and columns 4-6 for prime-age workers. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 16

17 Table 11: Effects on Transitions and Employment of Crossing the 30th Birthday Threshold, Men (1) (2) (3) (4) (5) (6) Perm. Short-Term Self-Employed UI Public Worker Panel A: Hires, Short-Run, 05/ /1998 Treatment (0.0777) (0.110) (0.0298) (0.0756) (0.0211) Observations Panel B: Hires, Long-Run, 06/ /2001 Treatment (0.0560) (0.0590) (0.0207) (0.0393) (0.0145) Observations Perm. Short-Term Self-Emp. UI Public Employed Panel C: Emp., Treated Cohorts Born 05/ /1971 Treatment (0.130) (0.192) (0.0551) (0.0779) (0.0193) (0.162) Slope Before (0.0113) (0.0137) ( ) ( ) ( ) ( ) Slope After (0.0192) (0.0183) ( ) ( ) ( ) (0.0106) Observations Panel D: Emp., Placebo Cohorts Born 03/ /1976 Treatment (0.142) (0.278) (0.0578) (0.108) (0.0206) (0.314) Slope Before (0.0106) (0.0169) ( ) ( ) ( ) (0.0200) Slope After (0.0186) (0.0259) ( ) ( ) ( ) (0.0328) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. The sample is restricted to men. Panel A and B report the effects on entries into permanent, short-term, self-employed, or public jobs, as well as into UI, around the 30th birthday thresholds. Panel A focuses on the short-run effects (first 12 months after policy). Panel B focuses on the long-run effects (from the 12th month until policy is removed). Panels C displays the effects on the probability of being employed in any of these contracts, and the overall probability of working, for treated cohorts born between 05/ /1971. Panels D repeats the exercise in panel C but for a placebo cohort: workers born between 03/ /1976. The specification is equation 1. The estimation window is of 12 months on each side of the threshold. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 17

18 Table 12: Effects on Transitions and Employment of Crossing the 30th Birthday Threshold, Women (1) (2) (3) (4) (5) (6) Perm. Short-Term Self-Employed UI Public Worker Panel A: Hires, Short-Run, 05/ /1998 Treatment (0.0719) (0.0942) (0.0249) (0.0780) (0.0621) Observations Panel B: Hires, Long-Run, 06/ /2001 Treatment (0.0416) (0.0753) (0.0165) (0.0427) (0.0300) Observations Perm. Short-Term Self-Emp. UI Public Employed Panel C: Emp., Treated Cohorts Born 05/ /1971 Treatment (0.160) (0.170) (0.0508) (0.0923) (0.0342) (0.172) Slope Before (0.0150) (0.0139) ( ) ( ) ( ) (0.0107) Slope After (0.0233) (0.0188) ( ) (0.0103) ( ) (0.0156) Observations Panel D: Emp., Placebo Cohorts Born 03/ /1976 Treatment (0.150) (0.261) (0.0578) (0.0971) (0.0400) (0.251) Slope Before ( ) (0.0140) ( ) ( ) ( ) (0.0148) Slope After (0.0172) (0.0202) ( ) ( ) ( ) (0.0228) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. The sample is restricted to women. Panel A and B report the effects on entries into permanent, short-term, self-employed, or public jobs, as well as into UI, around the 30th birthday thresholds. Panel A focuses on the short-run effects (first 12 months after policy). Panel B focuses on the long-run effects (from the 12th month until policy is removed). Panels C displays the effects on the probability of being employed in any of these contracts, and the overall probability of working, for treated cohorts born between 05/ /1971. Panels D repeats the exercise in panel C but for a placebo cohort: workers born between 03/ /1976. The specification is equation 1. The estimation window is of 12 months on each side of the threshold. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 18

19 Table 13: Effects on Transitions and Employment for Men, Vs , 2001 (1) (2) (3) (4) (5) (6) (7) Perm. Perm. Perm. Temp. Temp. Temp. Trans. Temp. Hires Lay-Offs Quits Hires Lay-Offs Quits to Perm. Treatment (0.0415) (0.0195) (0.0192) (0.122) (0.0502) (0.0270) (0.0196) Post (0.0283) (0.0202) (0.0187) (0.0479) (0.0467) (0.0252) (0.0104) Treatment x Post (0.0448) (0.0268) (0.0270) (0.0657) (0.0671) (0.0369) (0.0160) Observations Effects on Employment Perm. Short-Term Perm. or ST Self-Emp. Public UI Employed Treatment (6.194) (8.254) (4.842) (3.756) (1.672) (1.525) (2.221) Post (7.407) (11.34) (6.562) (4.450) (2.054) (2.203) (2.757) Treatment x Post (0.191) (0.221) (0.149) (0.0974) (0.0519) (0.0742) (0.101) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. The sample is restricted to men. In the upper panel, the dependent variable is a dummy indicating whether a transition in or out of a permanent contract (columns 1-3) or a temporary contract (columns 4-6) happened. In column 7th the dependent variable is a dummy that indicates that the worker was converted to permanent, from a temporary contract, within the same firm. The lower panel reports the results for employment. Each dependent variable is one of the possible employment statuses. The policy change is the removal of payroll tax cuts for young workers in The specification is equation 2. Control variables included are calendar quarter dummies, education, experience, firm size, sector, citizenship, part-time job, and workers disability. Columns 1-3 report the results for young workers, and columns 4-6 for prime-age workers. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 19

20 Table 14: Effects on Transitions and Employment for Women, Vs , 2001 (1) (2) (3) (4) (5) (6) (7) Perm. Perm. Perm. Temp. Temp. Temp. Trans. Temp. Hires Lay-Offs Quits Hires Lay-Offs Quits to Perm. Treatment (0.0458) (0.0213) (0.0164) (0.163) (0.0513) (0.0225) (0.0221) Post (0.0298) (0.0216) (0.0154) (0.0455) (0.0473) (0.0204) (0.0117) Treatment x Post (0.0445) (0.0295) (0.0237) (0.0651) (0.0694) (0.0314) (0.0204) Observations Effects on Employment Perm. Short-Term Perm. or ST Self-Emp. Public UI Employed Treatment (5.325) (8.074) (6.959) (3.167) (3.448) (2.712) (3.491) Post (6.330) (12.27) (10.38) (4.202) (4.235) (3.111) (4.501) Treatment x Post (0.249) (0.268) (0.201) (0.0887) (0.0914) (0.100) (0.134) Observations Notes: The estimates have been multiplied by 100 so that they can be interpreted as effects in percentage points. The sample is restricted to women. In the upper panel, the dependent variable is a dummy indicating whether a transition in or out of a permanent contract (columns 1-3) or a temporary contract (columns 4-6) happened. In column 7th the dependent variable is a dummy that indicates that the worker was converted to permanent, from a temporary contract, within the same firm. The lower panel reports the results for employment. Each dependent variable is one of the possible employment statuses. The policy change is the removal of payroll tax cuts for young workers in The specification is equation 2. Control variables included are calendar quarter dummies, education, experience, firm size, sector, citizenship, part-time job, and workers disability. Columns 1-3 report the results for young workers, and columns 4-6 for prime-age workers. Robust standard errors, clustered at the month of birth level, are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 20

21 5 Additional Evidence on Substitution After the 30th Threshold /1971 cohort vs. older cohorts I begin by looking at the changes in the post-30 slopes for the 03/1971 cohort relative to those born in 03/1970. Individuals born in March 1971 crossed their 30th birthday when the policy was removed in March Thus, those who were still unemployed at the age of 30 cannot suffer from displacement since the policy is no longer in place. Therefore, if displacement was an issue, the changes in slope that we have detected for cohorts crossing their 30th birthday between May 1997 and March 2001 should be smaller for the March 1971 cohort relative to similar cohorts born in the previous year. The test is akin to a triple-differences strategy: y it = β 1 1[ãge it > 0] + β 2 ãge it 1[ãge t > 0]+ + β 3 ãge it + β 4 Cohort i + β 5 1[ãge it > 0] Cohort i + + β 6 ãge it Cohort i + β 7 ãge it 1[ãge t > 0] Cohort i + ɛ it (1) Cohort i is a dummy indicating that the individual was born in 03/1971. β 7 captures the differential evolution of employment after the 30th birthday for the 03/1971 cohort relative to the 03/1970 one. If there was significant displacement taking place, the coefficients should be positive. I pick the 03/1970 cohort as the control group because it is the closest one birthwise for which I can observe 12 months before and after the 30th birthday, before the policy change in Results are in table 15. None of the β 7 estimates are significant. Moreover, for the evolution of employment the coefficient has a negative sign, contrary to what was expected in case displacement effects were taking place. Note also that the β 2 coefficients replicate the findings of table??. In tables 16 and 17 in the web appendix I repeat the analysis using as control groups cohorts 04/1970 and 05/1970. The results are the same. 21

22 Table 15: Displacement Test, Cohort 03/1971 Vs. Cohort 03/1970 (1) (2) (3) (4) (5) (6) Perm. Short-Term Self-Emp. Public UI Employed Treatment (0.771) (0.623) (0.426) (0.226) (0.301) (0.808) Slope Before (0.0773) (0.0644) (0.0426) (0.0225) (0.0318) (0.0828) Slope After (0.111) (0.0900) (0.0614) (0.0329) (0.0446) (0.116) Cohort Born in 03/ (0.807) (0.657) (0.430) (0.224) (0.326) (0.845) Treatment x Born in 03/ (1.082) (0.876) (0.578) (0.304) (0.425) (1.122) Slope Before 30 x Born in 03/ (0.108) (0.0907) (0.0579) (0.0301) (0.0441) (0.115) Slope After 30 x Born in 03/ (0.155) (0.127) (0.0833) (0.0441) (0.0624) (0.162) Observations Notes: The table displays the effects on being employed after the 30th birthday for cohort born in 03/1971 relative to cohort born in 03/1970. Those born in 03/1971 could benefit from the policy until their 30th birthday, but should not have experienced displacement effects after their 30th birthday because the policy was not in place anymore. Thus, in case there were displacement effects, the evolution of employment after the 30th birthday for those born in 03/1971 should be positive relative to cohort 03/1970. The specification is equation 1 and the coefficient of interest is β 7 (Slope After 30 x Born in 03/1971). The dependent variables are dummies indicating whether the worker was employed as permanent, short-term, receiving UI, employed, self-employed, or public worker. The estimation window is of 12 months on each side of the threshold. Robust standard error are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 22

23 Table 16: Displacement Test, Cohort 03/1971 Vs. Cohort 04/1970 (1) (2) (3) (4) (5) (6) Perm. Short-Term Self-Emp. Public UI Employed Treatment (0.781) (0.643) (0.447) (0.238) (0.311) (0.823) Slope Before (0.0850) (0.0733) (0.0485) (0.0250) (0.0352) (0.0920) Slope After (0.123) (0.102) (0.0703) (0.0375) (0.0498) (0.129) Cohort Born in 03/ (0.833) (0.687) (0.452) (0.234) (0.341) (0.876) Treatment x Born in 03/ (1.114) (0.909) (0.605) (0.319) (0.442) (1.158) Slope Before 30 x Born in 03/ (0.122) (0.103) (0.0659) (0.0340) (0.0494) (0.129) Slope After 30 x Born in 03/ (0.175) (0.144) (0.0950) (0.0503) (0.0703) (0.182) Observations Notes: The table displays the effects on being employed of the policy removal in 2001 for young workers. Cohort born in 03/1971 could benefit from the policy until their 30th birthday, but should not have experienced displacement effects after their 30th birthday because the policy was not in place anymore. Cohort born in 03/1970 was not benefiting from the policy anymore in May 2001 because they had crossed the 30th threshold. Thus, in case they were being displaced, the policy removal should alleviate the displacement. The specification is equation 1. The dependent variables are dummies indicating whether the worker was employed as permanent, short-term, receiving UI, employed, self-employed, or public worker. The estimation window is of 12 months on each side of the threshold. Robust standard error are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 23

24 Table 17: Displacement Test, Cohort 03/1971 Vs. Cohort 05/1970 (1) (2) (3) (4) (5) (6) Perm. Short-Term Self-Emp. Public UI Employed Treatment (0.837) (0.673) (0.467) (0.252) (0.343) (0.872) Slope Before (0.101) (0.0831) (0.0562) (0.0303) (0.0415) (0.107) Slope After (0.145) (0.116) (0.0806) (0.0437) (0.0596) (0.151) Cohort Born in 03/ (0.892) (0.724) (0.478) (0.252) (0.365) (0.928) Treatment x Born in 03/ (1.182) (0.953) (0.634) (0.337) (0.476) (1.221) Slope Before 30 x Born in 03/ (0.143) (0.118) (0.0763) (0.0404) (0.0576) (0.150) Slope After 30 x Born in 03/ (0.204) (0.165) (0.109) (0.0585) (0.0826) (0.211) Observations Notes: The table displays the effects on being employed of the policy removal in 2001 for young workers. Cohort born in 03/1971 could benefit from the policy until their 30th birthday, but should not have experienced displacement effects after their 30th birthday because the policy was not in place anymore. Cohort born in 03/1970 was not benefiting from the policy anymore in May 2001 because they had crossed the 30th threshold. Thus, in case they were being displaced, the policy removal should alleviate the displacement. The specification is equation 1. The dependent variables are dummies indicating whether the worker was employed as permanent, short-term, receiving UI, employed, self-employed, or public worker. The estimation window is of 12 months on each side of the threshold. Robust standard error are shown in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1% 24

25 5.2 No rebound in employment of cohorts older than 30 The second test looks at whether there is a rebound in the evolution of employment of cohorts that had already crossed the 30th birthday in 03/2001. Specification is as in equation 1, but now the slopes are measured with respect to the time that has gone by since the change in March I choose as my benchmark cohort those born in 03/1970 since for them I have a full year of job market history after being 30 in 03/2001. As can be seen in table 18, there are no positive and significative employment effects of the discontinuation of the youth employment credit. Again, this is the opposite that would be expected if there had been any important displacement effect. As a robustness check, I repeat the analysis for cohorts born between 10/ /1970. The results are shown in the middle and lower part of table 18 and in table 19 in the web appendix. As before, there is no evidence consistent with displacement effects. 1 The specification is y it = η1[ month it > 0] + λ month it 1[ month t > 0] + β month it + ɛ it. 1[ month it > 0] is a dummy indicating a month after March 2001, and month it is the month-distance with respect to March

26 Table 18: Displacement Test, 2 Year Window (1) (2) (3) (4) (5) (6) Perm. Short-Term Self-Employed Public Worker UI Employed Born 03/1970 Treatment (0.717) (0.590) (0.414) (0.251) (0.308) (0.761) Slope Before (0.0668) (0.0551) (0.0387) (0.0231) (0.0287) (0.0720) Slope After (0.0955) (0.0779) (0.0553) (0.0333) (0.0406) (0.101) Observations Born 02/1970 Treatment (0.779) (0.626) (0.452) (0.284) (0.301) (0.796) Slope Before (0.0726) (0.0588) (0.0414) (0.0260) (0.0285) (0.0756) Slope After (0.103) (0.0830) (0.0603) (0.0377) (0.0402) (0.106) Observations Born 01/1970 Treatment (0.675) (0.554) (0.364) (0.205) (0.274) (0.716) Slope Before (0.0628) (0.0519) (0.0338) (0.0186) (0.0254) (0.0677) Slope After (0.0897) (0.0736) (0.0486) (0.0274) (0.0362) (0.0952) Observations Notes: The table displays the effects on being employed of the policy removal in 2001 for several cohorts that were not benefiting from the policy anymore in May 2001 because they had crossed the 30th threshold. Thus, in case they were being displaced, the policy removal should alleviate the displacement. The specification is equation 1 and the parameter of interest is λ (Slope after 30). The dependent variables are dummies indicating whether the worker was employed as permanent, short-term, receiving UI, employed, self-employed, or public worker. The estimation window is of 12 months on each side of the threshold. Robust standard errors are shown in parentheses. p < 0.10, p < 0.05, p <

27 Table 19: Displacement Test, 2 Year Window (1) (2) (3) (4) (5) (6) Perm. Short-Term Self-Employed Public Worker UI Employed Born 12/1969 Treatment (0.764) (0.581) (0.405) (0.274) (0.294) (0.787) Slope Before (0.0713) (0.0544) (0.0379) (0.0251) (0.0280) (0.0744) Slope After (0.102) (0.0771) (0.0543) (0.0362) (0.0398) (0.105) Observations Born 11/1969 Treatment (0.785) (0.634) (0.458) (0.273) (0.346) (0.806) Slope Before (0.0730) (0.0602) (0.0424) (0.0250) (0.0320) (0.0763) Slope After (0.104) (0.0842) (0.0609) (0.0365) (0.0464) (0.107) Observations Born 10/1969 Treatment (0.761) (0.594) (0.447) (0.271) (0.317) (0.783) Slope Before (0.0711) (0.0556) (0.0416) (0.0244) (0.0297) (0.0742) Slope After (0.101) (0.0791) (0.0594) (0.0360) (0.0416) (0.104) Observations Notes: The table displays the effects on being employed of the policy removal in 2001 for several cohorts that were not benefiting from the policy anymore in May 2001 because they had crossed the 30th threshold. Thus, in case they were being displaced, the policy removal should alleviate the displacement. The specification is equation 1 and the parameter of interest is λ (Slope after 30). The dependent variables are dummies indicating whether the worker was employed as permanent, short-term, receiving UI, employed, self-employed, or public worker. The estimation window is of 12 months on each side of the threshold. Robust standard errors are shown in parentheses. p < 0.10, p < 0.05, p <