Tax-benefit Linkage and Incidence of Social Security Contributions: Evidence from France

Transcription

1 Tax-benefit Linkage and Incidence of Social Security Contributions: Evidence from France Antoine Bozio, Thomas Breda and Julien Grenet Paris School of Economics (PSE) NBER Summer Institute Boston, July 25, 2018

2 Motivation ˆ Payroll tax/social Security contributions (SSCs) compulsory payments paid to general government that confer entitlement to receive a future social benefit 1 / 68

3 Motivation ˆ Payroll tax/social Security contributions (SSCs) compulsory payments paid to general government that confer entitlement to receive a future social benefit ˆ Tax-Benefit Linkage quid pro quo tax (Musgrave, 1968) in practice wide variations in degree of linkage salience of linkage varies according to institutional designs 1 / 68

4 Motivation ˆ Payroll tax/social Security contributions (SSCs) compulsory payments paid to general government that confer entitlement to receive a future social benefit ˆ Tax-Benefit Linkage quid pro quo tax (Musgrave, 1968) in practice wide variations in degree of linkage salience of linkage varies according to institutional designs ˆ Large share of tax revenues 26% of tax revenues in OECD in 2013 Large heterogeneity across countries 1 / 68

5 Figure 1: Social Security Contributions as a % of GDP, 2015 France Austria Slovenia Czech Republic Belgium Netherlands Germany Slovak Republic Italy Finland Hungary Poland Japan Spain Estonia Greece Luxembourg Norway Sweden Portugal OECD Average Latvia Turkey Switzerland Korea United States United Kingdom Israel Canada Ireland Iceland Mexico Chile Denmark New Zealand Australia 0% 2% 4% 6% 8% 10% 12% 14% 16% 18% Source: OECD.Stat 2 / 68

6 This Paper ˆ Our research question: how does tax-benefit linkage matter for the pass-through of payroll tax analysis of three reforms of payroll taxes in France one reform with strong tax-benefit linkage, two without estimate long-run vs. short-run pass-through to workers 3 / 68

7 This Paper ˆ Our research question: how does tax-benefit linkage matter for the pass-through of payroll tax analysis of three reforms of payroll taxes in France one reform with strong tax-benefit linkage, two without estimate long-run vs. short-run pass-through to workers ˆ Efficiency argument for payroll taxation workers should incorporate future entitlement in their labor supply response (Musgrave, 1968; Summers, 1989; Gruber, 1997) criticism: equity-efficiency trade-off (Kaplow, 2008) 3 / 68

8 This Paper ˆ Our research question: how does tax-benefit linkage matter for the pass-through of payroll tax analysis of three reforms of payroll taxes in France one reform with strong tax-benefit linkage, two without estimate long-run vs. short-run pass-through to workers ˆ Efficiency argument for payroll taxation workers should incorporate future entitlement in their labor supply response (Musgrave, 1968; Summers, 1989; Gruber, 1997) criticism: equity-efficiency trade-off (Kaplow, 2008) ˆ Salience of tax-benefit linkage at given level of redistribution, perception of future entitlements crucial for labor supply responses salience of tax-benefit linkage could matter for incidence and efficiency of payroll tax (Chetty et al., 2009) 3 / 68

9 Related literature ˆ Empirical research on incidence of payroll taxes macro literature using cross-country variations OECD (1990); Alesina and Perotti (1997); Ooghe et al. (2003) relatively few well identified micro studies 4 / 68

10 Related literature ˆ Empirical research on incidence of payroll taxes macro literature using cross-country variations OECD (1990); Alesina and Perotti (1997); Ooghe et al. (2003) relatively few well identified micro studies ˆ General finding: full shifting of payroll taxes to workers payroll tax reforms in the US: workers compensation (Gruber and Kruger, 1991); mandated maternity benefits (Gruber, 1994); UI (Anderson and Meyer, 1997, 2000) privatization of 1981 Chilean pension system (Gruber, 1997) 4 / 68

11 Related literature ˆ Empirical research on incidence of payroll taxes macro literature using cross-country variations OECD (1990); Alesina and Perotti (1997); Ooghe et al. (2003) relatively few well identified micro studies ˆ General finding: full shifting of payroll taxes to workers payroll tax reforms in the US: workers compensation (Gruber and Kruger, 1991); mandated maternity benefits (Gruber, 1994); UI (Anderson and Meyer, 1997, 2000) privatization of 1981 Chilean pension system (Gruber, 1997) ˆ Recent evidence challenging received wisdom Greek reform (Saez et al.,2012); payroll tax reduction for young workers in Sweden (Saez et al., 2017) economic incidence aligned with statutory incidence 4 / 68

12 Preview of Results ˆ Full shifting of payroll tax with strong and salient tax-benefit linkage evidence of full shifting of increases in employer SSCs to workers 5 / 68

13 Preview of Results ˆ Full shifting of payroll tax with strong and salient tax-benefit linkage evidence of full shifting of increases in employer SSCs to workers ˆ No or limited shifting of payroll tax with no linkage evidence of increased labor cost, i.e., the absence of full tax shifting to workers at the individual level shifting to workers estimated between 6% and 21% 5 / 68

14 Preview of Results ˆ Full shifting of payroll tax with strong and salient tax-benefit linkage evidence of full shifting of increases in employer SSCs to workers ˆ No or limited shifting of payroll tax with no linkage evidence of increased labor cost, i.e., the absence of full tax shifting to workers at the individual level shifting to workers estimated between 6% and 21% ˆ Meta-analysis of previous literature evidence that tax-benefit linkage and its salience matter for incidence rationalize apparently contradictory results in previous literature 5 / 68

15 Outline 1. Introduction 2. Conceptual framework 3. SSC reforms in France 4. Empirical strategy and data 5. Results 6. Conclusion 7 / 68

16 Outline 1. Introduction 2. Conceptual framework 3. SSC reforms in France 4. Empirical strategy and data 5. Results 6. Conclusion 7 / 68

17 Conceptual Framework ˆ Based on Gruber (1997) and Saez et al. (2017) ˆ Labour demand CES production function with two types of workers T and C labor cost : z i = w i (1 + τ) w i : posted wage τ: payroll tax rate on employers σ: elasticity of substitution between both inputs ˆ Labour supply where n i = φ w ξ i w i = w i + q w i τ ξ: elasticity of labor supply q: extent to which employees value employer contributions 8 / 68

18 Incidence Formula ˆ Shifting parameter S to workers with possible linkage S = d ln w T w C d ln(1 + τ T ) qξ + σ ξ + σ 9 / 68

19 Incidence Formula ˆ Shifting parameter S to workers with possible linkage ˆ Three polar cases: S = d ln w T w C d ln(1 + τ T ) qξ + σ ξ + σ (1) Full linkage (q = 1) full shifting to workers (S = 1) 9 / 68

20 Incidence Formula ˆ Shifting parameter S to workers with possible linkage ˆ Three polar cases: S = d ln w T w C d ln(1 + τ T ) qξ + σ ξ + σ (1) Full linkage (q = 1) full shifting to workers (S = 1) (2) No linkage (q = 0) and σ ξ full shifting (S 1) 9 / 68

21 Incidence Formula ˆ Shifting parameter S to workers with possible linkage ˆ Three polar cases: S = d ln w T w C d ln(1 + τ T ) qξ + σ ξ + σ (1) Full linkage (q = 1) full shifting to workers (S = 1) (2) No linkage (q = 0) and σ ξ full shifting (S 1) (3) No linkage (q = 0) and ξ σ no shifting (S 0) 9 / 68

22 Incidence Formula ˆ Shifting parameter S to workers with possible linkage ˆ Three polar cases: S = d ln w T w C d ln(1 + τ T ) qξ + σ ξ + σ (1) Full linkage (q = 1) full shifting to workers (S = 1) (2) No linkage (q = 0) and σ ξ full shifting (S 1) (3) No linkage (q = 0) and ξ σ no shifting (S 0) Case (2) is the usual assumptions in the labor supply/elasticity of taxable income literature 9 / 68

23 Outline 1. Introduction 2. Conceptual framework 3. SSC reforms in France 4. Empirical strategy and data 5. Results 6. Conclusion 10 / 68

24 SSC Reforms in France ˆ SSC schedule in France SSC schedule applied to gross (posted) hourly wage Social Security Threshold (SST) is around P70 SSC schedule applied to different earnings brackets: 0 1 SST ( P70), 1 4 SST ( P98) 11 / 68

25 SSC Reforms in France ˆ SSC schedule in France SSC schedule applied to gross (posted) hourly wage Social Security Threshold (SST) is around P70 SSC schedule applied to different earnings brackets: 0 1 SST ( P70), 1 4 SST ( P98) ˆ Three major payroll tax increases above SST Pension SSC : +7.8 ppts Family SSC : +8.2 ppts Health care SSC : +9.5 ppts 11 / 68

26 Figure 2: Marginal Employer SSC Rates, Non-Executives, Under SST 1 to 3 SST Year Sources: IPP Tax and Benefit Tables (April 2016) ; TAXIPP / 68

27 Figure 2: Marginal Employer SSC Rates, Non-Executives, Reform 3 Uncapping of heath SSCs Reform 2 Uncapping of family SSCs +9.5 ppts +8.2 ppts Reform 1 Increase in pensions SSCs +7.8 ppts Under SST 1 to 3 SST Year Sources: IPP Tax and Benefit Tables (April 2016) ; TAXIPP / 68

28 Varying degree of tax-benefit linkage ˆ Reform 1 (pension) : strong tax-benefit linkage point-based system (similar to NDC system) pension P R is computed from past contributions wh t τ (with shadow prices p b,t, p s,r ) additional payroll tax leads to increased pension benefit for individuals affected P R = ( R 1 wh t p s,r p t=t b,t 0 ) τ 13 / 68

29 Varying degree of tax-benefit linkage ˆ Reform 1 (pension) : strong tax-benefit linkage point-based system (similar to NDC system) pension P R is computed from past contributions wh t τ (with shadow prices p b,t, p s,r ) additional payroll tax leads to increased pension benefit for individuals affected P R = ( R 1 wh t p s,r p t=t b,t 0 ˆ Salient tax-benefit linkage anecdotal evidence suggesting that the increase in pension benefit was well understood press reports about increase in pension benefits Quote from Le Monde, 1996 ) τ 13 / 68

30 Varying Degree of Tax-Benefit Linkage ˆ Reform 2 (Family): no tax-benefit linkage Family SSCs fund child benefit no tax-benefit linkage: universal benefit for all French families 14 / 68

31 Varying Degree of Tax-Benefit Linkage ˆ Reform 2 (Family): no tax-benefit linkage Family SSCs fund child benefit no tax-benefit linkage: universal benefit for all French families ˆ Reform 3 (Health Care): no tax-benefit linkage Health care insurance covers almost all French residents no change in benefits when increases in SSC rate 14 / 68

32 Outline 1. Introduction 2. Conceptual framework 3. SSC reforms in France 4. Empirical strategy and data 5. Results 6. Conclusion 15 / 68

33 Empirical Strategy ˆ Difference-in-differences estimation Treated: workers with gross earnings > SST before reform Control: workers with gross earnings SST before reform Before/after comparisons: up to 9 years after reforms ˆ First stage: relative change in average employer SSCs for treated vs. control ˆ Reduced-form outcomes: relative changes in hourly labor cost and hourly wage (reform 1) labor cost and gross earnings (all reforms) ˆ 2SLS: Share of employer payroll tax borne by workers 16 / 68

34 Figure 3: Empirical Strategy Average SSC rate Social Security Threshold CONTROL GROUP TREATMENT GROUP Before reform After reform Gross earnings 17 / 68

35 Data ˆ Administrative data : DADS panel 2010 employer-employee administrative data reported by employers to SS schemes 1/25 sample for years , 1/12 from 2002 onwards 1.1 million workers each year (2.2 million in recent years) some missing years: 1981, 1983, / 68

36 Data ˆ Administrative data : DADS panel 2010 employer-employee administrative data reported by employers to SS schemes 1/25 sample for years , 1/12 from 2002 onwards 1.1 million workers each year (2.2 million in recent years) some missing years: 1981, 1983, 1990 ˆ Available information start and end of job spell, firm size, sector, occupation net taxable earnings available throughout the period hours available from 1993 onwards 18 / 68

37 Data ˆ Microsimulation model TAXIPP model developed at the Institute of Public Policy (IPP) very detailed simulations of SSCs (over 50 schedules!) ˆ Simulating SSCs using TAXIPP compute gross earnings from net taxable earnings obtain labor cost by adding employer SSCs to gross earnings before 1993 our simulations are accurate only for full-time, full-year wage earners (no information on hours for part-time wage earners) 19 / 68

38 Estimation Sample ˆ Sample restrictions full-time, full-year non-executive workers observed in reference year (i.e., last pre-reform year) construct unbalanced panel around reform years ˆ Definition of treated/controls trade-off: proximity to threshold vs. treatment intensity groups defined based on gross earnings in reference year - Treated: between SST and 1.4 SST - Controls: between 0.9 SST and SST 20 / 68

39 Table 1: Summary Statistics Sample: Control Group Treatment Group Reform 1: Increase in Pensions SSCs ( ) Rank in the earnings distribution [P62 P70] [P70 P87] Mean gross earnings (euros) 30,324 36,710 Number of individuals 19,626 33,605 Reform 2: Uncapping of Family SSCs (1989 and 1990) Rank in the earnings distribution [P58 P67] [P67 P85] Mean gross earnings (euros) 26,073 31,767 Number of individuals 24,388 46,575 Reform 3: Uncapping of Health Care SSCs (1981 and 1983) Rank in the earnings distribution [P56 P65] [P65 P85] Mean gross earnings (euros) 22,418 27,452 Number of individuals 30,979 63,109 Sources: Panel DADS 2010; TAXIPP / 68

40 Estimation ˆ Specification 1: Reduced form K log(1 + τ it ) = α + θ i + θ t + β k (T i 1{t = k}) + ε it (1) k=1 K log(w it ) = µ + η i + η t + γ k (T i 1{t = k}) + ν it (2) β k, γ k : reduced-form effects of reform after k years ˆ 2SLS estimate of pass-through rate on workers: pass-through after k years = ˆγ k / ˆβ k ˆ Standard errors clustered at the individual level k=1 22 / 68

41 Estimation ˆ Specification 2 relax common-trend assumption by including individual-specific linear time trends θ i.t individual trends are fitted based on up to 5 years of pre-reform data 23 / 68

42 Estimation ˆ Specification 2 relax common-trend assumption by including individual-specific linear time trends θ i.t individual trends are fitted based on up to 5 years of pre-reform data ˆ Hourly wage vs. earnings pass-through estimated with hourly wage for reform 1 using earnings for reforms 2 and 3 (no hours) upper bound of shifting parameter estimated 23 / 68

43 Outline 1. Introduction 2. Conceptual framework 3. SSC reforms in France 4. Empirical strategy and data 5. Results 6. Conclusion 24 / 68

44 Figure 4: Reform 1 (Increase in Pensions SSCs): Gross Hourly Wage Gross Hourly Wage (100 in 1999) Reform 1 Increase in Pensions SSCs Year Treatment: 1 to 1.4 SST Control:.9 to 1 SST Sources: DADS Panel 2010; TAXIPP / 68

45 Figure 5: Reform 1 (Increase in Pensions SSCs): Hourly Labor Cost Hourly Labor Cost (100 in 1999) Reform 1 Increase in Pensions SSCs Year Treatment: 1 to 1.4 SST Control:.9 to 1 SST Sources: DADS Panel 2010; TAXIPP 0.4. Graph with earnings 26 / 68

46 Figure 6: Reform 1: log(z) vs log(w) Years since reference year Hourly labour cost Gross hourly earnings Sources: DADS Panel 2010; TAXIPP / 68

47 Figure 7: Reform 1: Pass-Through Rate on Workers (2SLS) w Years since reference year Estimate 95% CI Sources: DADS Panel 2010; TAXIPP 0.4. Graph with earnings 28 / 68

48 Figure 8: Reform 1: Pass-Through Rate on Workers w with trends Years since reference year Estimate 95% CI Sources: DADS Panel 2010; TAXIPP / 68

49 Figure 9: Reform 2 (Uncapping of Family SSCs): Gross Earnings Gross Earnings (100 in 1988) Reform 2 Uncapping of Family SSCs Year Treatment: 1 to 1.4 SST Control:.9 to 1 SST Sources: DADS Panel 2010; TAXIPP / 68

50 Figure 10: Reform 2 (Uncapping of Family SSCs): Labor Cost Labor Cost (100 in 1988) Reform 2 Uncapping of Family SSCs Year Treatment: 1 to 1.4 SST Control:.9 to 1 SST Sources: DADS Panel 2010; TAXIPP / 68

51 Figure 11: Reform 2: log(zh) vs log(wh) Years since reference year Labour cost Gross earnings Sources: DADS Panel 2010; TAXIPP / 68

52 Figure 12: Reform 2: Pass-Through Rate on Workers (2SLS) wh Estimate 95% CI Years since reference year Sources: DADS Panel 2010; TAXIPP / 68

53 Figure 13: Reform 2: Pass-Through Rate on Workers with trends Estimate 95% CI Years since reference year Sources: DADS Panel 2010; TAXIPP / 68

54 Figure 14: Reform 3 (Uncapping of Health care SSCs): Gross Earnings Gross Earnings (100 in 1980) Reform 3 Uncapping of Health SSCs Year Treatment: 1 to 1.4 SST Control:.9 to 1 SST Sources: DADS Panel 2010; TAXIPP / 68

55 Figure 15: Reform 3 (Uncapping of Health care SSCs): Labor Cost Labor Cost (100 in 1980) Reform 3 Uncapping of Health SSCs Year Treatment: 1 to 1.4 SST Control:.9 to 1 SST Sources: DADS Panel 2010; TAXIPP / 68

56 Figure 16: Reform 3: log(zh) vs log(wh) Years since reference year Labour cost Gross earnings Sources: DADS Panel 2010; TAXIPP / 68

57 Figure 17: Reform 3: Pass-Through Rate on Workers (2SLS) - wh Years since reference year Estimate 95% CI Sources: DADS Panel 2010; TAXIPP / 68

58 Figure 18: Reform 3: Pass-Through Rate on Workers with trends Years since reference year Estimate 95% CI Sources: DADS Panel 2010; TAXIPP / 68

59 Summary Table 2: Baseline estimates of pass-through rate on workers Reform: Reform 1 Reform 2 Reform 3 Dep. var.: log(hourly wage) log(earnings) log(earnings) log(earnings) Panel A. Without controlling for individual-specific trends t *** 0.812*** ** (0.303) (0.293) (0.166) (0.172) t *** 0.969*** n/a (0.327) (0.324) (0.170) n/a Panel B. Controlling for individual-specific trends t *** 1.112*** (0.318) (0.291) (0.224) (0.133) t *** 1.157*** n/a (0.335) (0.308) (0.229) n/a 40 / 68

60 Summary ˆ Markedly different estimates R1 statistically different from both R2 and R3 full shifting to employees very quickly R2 and R3 not statistically different from one another we reject full shifting to employee 6 years after the SSC increase 41 / 68

61 Summary ˆ Markedly different estimates R1 statistically different from both R2 and R3 full shifting to employees very quickly R2 and R3 not statistically different from one another we reject full shifting to employee 6 years after the SSC increase ˆ Heterogeneity men vs. women: no statistically significant difference same firm vs. other firms: inconclusive evidence 41 / 68

62 Robustness Checks ˆ Placebo reform in 1996 check common trend assumption no reform between 1992 and 1999 estimate pseudo reform in 1996 (reference year in 1995) compare evolution of labor cost/gross earnings for treated vs. control 42 / 68

63 Figure 19: Marginal Employer SSC Rates, non-executives, Reform 1 Uncapping of heath SSCs Reform 2 Uncapping of family SSCs Reform 3 Increase in pensions SSCs Placebo reform Under SST 1 to 3 SST Year 43 / 68

64 Figure 20: Placebo Reform (1996): Real Gross Earnings Gross Earnings (100 in 1995) Year Treatment: 1 to 1.4 SST Control:.9 to 1 SST Sources: DADS Panel 2010; TAXIPP / 68

65 Figure 21: Placebo Reform (1996): Labor Cost Labor Cost (100 in 1995) Year Treatment: 1 to 1.4 SST Control:.9 to 1 SST Sources: DADS Panel 2010; TAXIPP / 68

66 Figure 22: Placebo Reform: Differential log(labor cost) no trends Reform Estimate 3 Increase in Pensions SSCs 95% CI Years since reference year Sources: DADS Panel 2010; TAXIPP / 68

67 Figure 23: Placebo Reform: Differential log(labor cost) w/ trends Years since reference year Estimate 95% CI Sources: DADS Panel 2010; TAXIPP / 68

68 Robustness Checks ˆ Sensitivity to definition of treatment group closer group to SST: better identification, weak first stage further away from SST: stronger first stage, weaker identification ˆ Robustness check check sensitivity to upper bound of treatment group: variation from 1.2 to 1.6 SST check sensitivity to lower bound of control group: variation from 0.85 to 0.95 SST Graphs 48 / 68

69 Figure 24: Reform 1: Sensitivity Tests Pass-Through Rate on Workers at t (no trend) High earnings (%SST) of the treatment group 49 / 68

70 Figure 25: Reform 2: Sensitivity Tests Pass-Through Rate on Workers at t (no trend) High earnings threshold of the treatment group (% of SST) 50 / 68

71 Figure 26: Reform 3: Sensitivity Tests Pass-Through Rate on Workers at t (no trend) High earnings threshold of the treatment group (% of SST) 51 / 68

72 Discussion (1) ˆ Degree of tax-benefit linkage could explain marked difference in incidence across reforms strong and salient linkage (reform 1) vs no tax-benefit linkage (reforms 2 and 3) 52 / 68

73 Discussion (1) ˆ Degree of tax-benefit linkage could explain marked difference in incidence across reforms strong and salient linkage (reform 1) vs no tax-benefit linkage (reforms 2 and 3) ˆ Other explanations? different time period changes in labor market conditions in France 52 / 68

74 Discussion (1) ˆ Degree of tax-benefit linkage could explain marked difference in incidence across reforms strong and salient linkage (reform 1) vs no tax-benefit linkage (reforms 2 and 3) ˆ Other explanations? different time period changes in labor market conditions in France ˆ Meta-analysis of payroll tax shifting estimates papers using well identified payroll tax reforms estimating degree of linkage from authors quote 52 / 68

75 Figure 27: Meta-Analysis of Payroll Tax Incidence no shifting full shifting Gruber & Krueger (1991) Gruber (1994) Gruber (1997) Anderson & Meyer (1997) Anderson & Meyer (2000) Komamura & Yamada (2004) Baicker & Chandra (2006) Murphy (2007) Kugler & Kugler (2009) Korkeamäki & Uusitalo (2009) Bennmarker et al. (2009) Cruces et al. (2010) Saez et al. (2012) Tax-benefit linkage: Strong Weak Uncertain Saez et al. (2017) Bozio et al. (2018) Estimated pass-through to workers 53 / 68

76 Discussion (2): Incidence on Employers? ˆ Why not full shifting when no tax-benefit linkage? inelastic labor demand / small elasticity of substitution σ fairness norms within the firm (Saez, 2012, 2017) 54 / 68

77 Discussion (2): Incidence on Employers? ˆ Why not full shifting when no tax-benefit linkage? inelastic labor demand / small elasticity of substitution σ fairness norms within the firm (Saez, 2012, 2017) ˆ Can results be rationalized with standard framework? shifting estimates between 0.06 and 0.21 for reforms 2 and 3 would imply elasticity of substitution between 0.03 and 0.13 implausibly low given available estimates 54 / 68

78 Discussion (2): Incidence on Employers? ˆ Why not full shifting when no tax-benefit linkage? inelastic labor demand / small elasticity of substitution σ fairness norms within the firm (Saez, 2012, 2017) ˆ Can results be rationalized with standard framework? shifting estimates between 0.06 and 0.21 for reforms 2 and 3 would imply elasticity of substitution between 0.03 and 0.13 implausibly low given available estimates ˆ Rejection of full shifting at the individual level but not necessarily at firm level or at market level 54 / 68

79 Conclusion ˆ What have we found? Full and quick shifting of employer payroll tax to workers in the case of strong and salient tax-benefit linkage No or limited shifting, up to 7-8 years, of payroll taxes with no linkage 55 / 68

80 Conclusion ˆ What have we found? Full and quick shifting of employer payroll tax to workers in the case of strong and salient tax-benefit linkage No or limited shifting, up to 7-8 years, of payroll taxes with no linkage ˆ Contributions Salience of tax-benefit linkage can rationalize conflicting findings in the literature When no linkage, limited shifting at the individual level cannot be easily explained with the standard framework 55 / 68

81 Conclusion ˆ What have we found? Full and quick shifting of employer payroll tax to workers in the case of strong and salient tax-benefit linkage No or limited shifting, up to 7-8 years, of payroll taxes with no linkage ˆ Contributions Salience of tax-benefit linkage can rationalize conflicting findings in the literature When no linkage, limited shifting at the individual level cannot be easily explained with the standard framework ˆ Policy implications Higher efficiency of SSC can come only with salience of the tax-benefit linkage 55 / 68

82 Tax-benefit Linkage and Incidence of Social Security Contributions: Evidence from France Antoine Bozio, Thomas Breda and Julien Grenet Paris School of Economics (PSE) NBER Summer Institute Boston, July 25, 2018

83 Varying degree of tax-benefit linkage the agreement also entails that wage earners whose wage is above the Social Security threshold would be able to constitute themselves a better pension: the contribution rate will be raised to 16 percent by 2005 for workers of existing firms, and as soon as 2000 for firms created after January 1st 1997 Jean-Michel Bezat, La baisse des retraites complémentaires est programmée, Le Monde, 27 April back 57 / 68

84 Figure 28: Reform 1 (Increase in Pensions SSCs): Gross Earnings Gross Earnings (100 in 1999) Reform 1 Increase in Pensions SSCs Year Treatment: 1 to 1.4 SST Control:.9 to 1 SST Sources: DADS Panel 2010; TAXIPP / 68

85 Figure 29: Reform 1 (Increase in Pensions SSCs): Labor Cost Labor Cost (100 in 1999) Reform 1 Increase in Pensions SSCs Year Treatment: 1 to 1.4 SST Control:.9 to 1 SST Sources: DADS Panel 2010; TAXIPP 0.4. back 59 / 68

86 Figure 30: Reform 1: log(zh) vs log(wh) Years since reference year Labour cost Gross earnings Sources: DADS Panel 2010; TAXIPP / 68

87 Figure 31: Reform 1: Pass-Through Rate on Workers wh Years since reference year Estimate 95% CI Sources: DADS Panel 2010; TAXIPP 0.4. back 61 / 68

88 Figure 32: Reform 1: Sensitivity Tests (t 0 + 8) Lower earnings (%SST) of the control group Sources: DADS Panel 2010; TAXIPP / 68

89 Figure 33: Reform 2: Sensitivity Tests (t 0 + 8) Lower earnings threshold of the control group (% of SST) Sources: DADS Panel 2010; TAXIPP / 68

90 Figure 34: Reform 3: Sensitivity Tests (t 0 + 8) Lower earnings threshold of the control group (% of SST) Sources: DADS Panel 2010; TAXIPP 0.4. back 64 / 68

91 Figure 35: Reform 3: hours Responses no trends Years since reference year Estimate 95% CI Sources: DADS Panel 2010; TAXIPP 0.4. back 65 / 68

92 Behavioral responses Table 3: Impact of SSC Reforms on Probability of Entering Full-time Employment with Earnings above the SST Reform: Reform 1: Reform 2: Reform 3: t ** (0.003) (0.003) (0.002) t (0.003) (0.004) (0.002) t *** (0.003) (0.004) (0.002) t *** (0.003) (0.004) (0.002) t 0 +9 n/a n/a (0.003) (0.002) 66 / 68

93 Behavioral responses Table 4: Impact of SSC Reforms on Probability of Exiting Full-time Employment with Earnings above the SST Reform: Reform 1: Reform 2: Reform 3: t *** (0.003) (0.003) (0.002) t *** 0.024*** (0.003) (0.004) (0.002) t ** 0.005** (0.003) (0.004) (0.002) t * 0.006*** (0.003) (0.003) (0.002) t 0 +9 n/a 0.005* 0.004** n/a (0.003) (0.002) back 67 / 68

94 Discussion: incidence vs. earnings responses ˆ Incidence is a change in wage rate Hours not observed in the data before 1993 Not possible to distinguish incidence from behavioral response Need to assume no behavioral response 68 / 68

95 Discussion: incidence vs. earnings responses ˆ Incidence is a change in wage rate Hours not observed in the data before 1993 Not possible to distinguish incidence from behavioral response Need to assume no behavioral response ˆ Incidence or behavioral responses? We use only full-time employees Substitution effects would lead to a reduction in hours, hence lower earnings (opposite for income effects) We interpret our earnings responses as being a close approximation of incidence 68 / 68