The Effect of Taxation on Informal Employment

Transcription

1 The Effect of Taxation on Informal Employment Evidence from the Russian Flat Tax Reform Fabián Slonimczyk ICEF-Higher School of Economics, Moscow Joint CLMS-LIRT Seminar. November 2011.

2 Informality and Taxes In developing countries, 50% or more of employment is informal: self-employed, unregistered workers Critical issue: relationship between informality and the level of taxation 1 Challenge 1: Identification 2 Challenge 2: Measurement Related Literatures Shadow economy: government regulation, taxation, and informal work. Segmented markets vs. excessive regulation

3 Informality and Taxes In developing countries, 50% or more of employment is informal: self-employed, unregistered workers Critical issue: relationship between informality and the level of taxation 1 Challenge 1: Identification 2 Challenge 2: Measurement Related Literatures Shadow economy: government regulation, taxation, and informal work. Segmented markets vs. excessive regulation

4 Informality and Taxes In developing countries, 50% or more of employment is informal: self-employed, unregistered workers Critical issue: relationship between informality and the level of taxation 1 Challenge 1: Identification 2 Challenge 2: Measurement Related Literatures Shadow economy: government regulation, taxation, and informal work. Segmented markets vs. excessive regulation

5 Informality and Taxes In developing countries, 50% or more of employment is informal: self-employed, unregistered workers Critical issue: relationship between informality and the level of taxation 1 Challenge 1: Identification 2 Challenge 2: Measurement Related Literatures Shadow economy: government regulation, taxation, and informal work. Segmented markets vs. excessive regulation

6 Informality and Taxes In developing countries, 50% or more of employment is informal: self-employed, unregistered workers Critical issue: relationship between informality and the level of taxation 1 Challenge 1: Identification 2 Challenge 2: Measurement Related Literatures Shadow economy: government regulation, taxation, and informal work. Segmented markets vs. excessive regulation

7 Informality and Taxes In developing countries, 50% or more of employment is informal: self-employed, unregistered workers Critical issue: relationship between informality and the level of taxation 1 Challenge 1: Identification 2 Challenge 2: Measurement Related Literatures Shadow economy: government regulation, taxation, and informal work. Segmented markets vs. excessive regulation

8 Informality and Taxes In developing countries, 50% or more of employment is informal: self-employed, unregistered workers Critical issue: relationship between informality and the level of taxation 1 Challenge 1: Identification 2 Challenge 2: Measurement Related Literatures Shadow economy: government regulation, taxation, and informal work. Segmented markets vs. excessive regulation

9 The Flat Tax Reform In 2001, Russia introduced a tax reform that drastically reduced taxation levels But lower income brackets were mostly unaffected: control group The effect of the reform estimated using a differences-in-differences strategy Other aspects of the reform make it a good quasi-experiment Little or no room for anticipation effects No specific incentives to misreport income around the threshold But not perfect, since treatment is defined based on income bracket

10 The Flat Tax Reform In 2001, Russia introduced a tax reform that drastically reduced taxation levels But lower income brackets were mostly unaffected: control group The effect of the reform estimated using a differences-in-differences strategy Other aspects of the reform make it a good quasi-experiment Little or no room for anticipation effects No specific incentives to misreport income around the threshold But not perfect, since treatment is defined based on income bracket

11 The Flat Tax Reform In 2001, Russia introduced a tax reform that drastically reduced taxation levels But lower income brackets were mostly unaffected: control group The effect of the reform estimated using a differences-in-differences strategy Other aspects of the reform make it a good quasi-experiment Little or no room for anticipation effects No specific incentives to misreport income around the threshold But not perfect, since treatment is defined based on income bracket

12 The Flat Tax Reform In 2001, Russia introduced a tax reform that drastically reduced taxation levels But lower income brackets were mostly unaffected: control group The effect of the reform estimated using a differences-in-differences strategy Other aspects of the reform make it a good quasi-experiment Little or no room for anticipation effects No specific incentives to misreport income around the threshold But not perfect, since treatment is defined based on income bracket

13 The Flat Tax Reform In 2001, Russia introduced a tax reform that drastically reduced taxation levels But lower income brackets were mostly unaffected: control group The effect of the reform estimated using a differences-in-differences strategy Other aspects of the reform make it a good quasi-experiment Little or no room for anticipation effects No specific incentives to misreport income around the threshold But not perfect, since treatment is defined based on income bracket

14 The Flat Tax Reform In 2001, Russia introduced a tax reform that drastically reduced taxation levels But lower income brackets were mostly unaffected: control group The effect of the reform estimated using a differences-in-differences strategy Other aspects of the reform make it a good quasi-experiment Little or no room for anticipation effects No specific incentives to misreport income around the threshold But not perfect, since treatment is defined based on income bracket

15 The Flat Tax Reform In 2001, Russia introduced a tax reform that drastically reduced taxation levels But lower income brackets were mostly unaffected: control group The effect of the reform estimated using a differences-in-differences strategy Other aspects of the reform make it a good quasi-experiment Little or no room for anticipation effects No specific incentives to misreport income around the threshold But not perfect, since treatment is defined based on income bracket

16 Outline 1 Introduction 2 Informality Definition and Measurement Data Definition of Informal Employment 3 Description of Informal Employment in Russia 4 The Tax Reform Identification of the Tax Reform Effect 5 Results Regression Analysis Matching DID Other Experiments Extensive Margin 6 Conclusion

17 The Data The Russian Longitudinal Monitoring Survey Rounds VIII XVIII ( ) In typical round, 10,000 individuals in 4,000 household The adult questionnaire contains information on up to three jobs : main job, second job, irregular remunerated activities Special supplement on informal work (round XVIII)

18 The Data The Russian Longitudinal Monitoring Survey Rounds VIII XVIII ( ) In typical round, 10,000 individuals in 4,000 household The adult questionnaire contains information on up to three jobs : main job, second job, irregular remunerated activities Special supplement on informal work (round XVIII)

19 The Data The Russian Longitudinal Monitoring Survey Rounds VIII XVIII ( ) In typical round, 10,000 individuals in 4,000 household The adult questionnaire contains information on up to three jobs : main job, second job, irregular remunerated activities Special supplement on informal work (round XVIII)

20 The Data The Russian Longitudinal Monitoring Survey Rounds VIII XVIII ( ) In typical round, 10,000 individuals in 4,000 household The adult questionnaire contains information on up to three jobs : main job, second job, irregular remunerated activities Special supplement on informal work (round XVIII)

21 The Data The Russian Longitudinal Monitoring Survey Rounds VIII XVIII ( ) In typical round, 10,000 individuals in 4,000 household The adult questionnaire contains information on up to three jobs : main job, second job, irregular remunerated activities Special supplement on informal work (round XVIII)

22 Informality Productive definition Private unincorporated production units Production for exchange Small scale (<5 employees) Legalistic or social protection definition

23 Informality Productive definition Private unincorporated production units Production for exchange Small scale (<5 employees) Legalistic or social protection definition

24 Informality Productive definition Private unincorporated production units Production for exchange Small scale (<5 employees) Legalistic or social protection definition

25 Informality Productive definition Private unincorporated production units Production for exchange Small scale (<5 employees) Legalistic or social protection definition

26 Informality Productive definition Private unincorporated production units Production for exchange Small scale (<5 employees) Legalistic or social protection definition

27 Working Definition Employed Main Job Entrepreneur Employee Firm Owners Individual Entrepreneur For Firm For Individual Entrepreneur Formal Informal Informal Formal Informal Informal Second Job Formal Informal Irregular Activities Formal Informal

28 Informality at Main Job Entrep Inf Entrep Inf Employee Notes: RLMS, rounds VIII XVIII ( ). The series are defined as a percentage of those with a main job.

29 Informality at 2nd Job and Irreg Activ Sec Job Irreg Act Inf Sec Inf Irreg Notes: RLMS, rounds VIII XVIII ( ). The series for the second job are defined as a percentage of those with a main job. For the irregular activities, the base are all those employed.

30 Real Hourly Earnings Inf Irreg Act Exc Inf Irreg Act Formal Employee Inf Employee Inf Entrep Notes: RLMS, rounds VIII XVIII ( ). Real hourly earnings are monthly receipts divided by usual hours and deflated by the CPI. Informal irregular activities exclusive means that the individual held no other job.

31 Sup for employees Compliance with the Law All Employed Informal Employee Informal Sec. Job Inf. Irreg. Activ Under oral agreement % Labor Law Compliace NA 53.2 % Contract Compliance NA 65.5 % of Inc Declared for SS NA 10.5 Obs Sup for entrepreneurs All Employed Formal Entrep Informal Entrepr. Inf. Irreg. Activ Unregistered % Labor Law Compliance % Contract Compliance % Formal Employees Contributes to SS fund Obs Notes: The data sources are RLMS round XVIII and the supplementary questionnaire on informality by the Center of Labor Market Studies, Higher School of Economics (2009). Based on job-b answers by individuals who do not perform irregular activities. Based on job-a answers by individuals who do not have a main job.

32 The Russian Flat Tax Reform Before (2000) After (2001) Gross Yearly PIT ST PIT ST Income (r.) Employee Employer Employee Employer <3, ,168 4, ,800-50, , , , , , , , , >600, Notes: The data source is Russian Tax Code, part 2 (2001-2). The tax allowance in 2001 was only available to those with income below 20,000 rubles. Rate initially set to 5% and lowered to 2% in 2002.

33 Combined Tax Burden Gross Yearly Income (thousands) Before After

34 Differences in Differences The design of the reform created a natural experiment Individuals earning less than 50,000 rubles a year constitute the control group People with higher incomes faced lower tax rates and are therefore considered treated Problems in practice: 1 Income misreporting Use post-reform income Identify lower bound 2 Income volatility Considered treated if ever in upper brackets 3 The control group comprises the un-treated individuals who were employed

35 Differences in Differences The design of the reform created a natural experiment Individuals earning less than 50,000 rubles a year constitute the control group People with higher incomes faced lower tax rates and are therefore considered treated Problems in practice: 1 Income misreporting Use post-reform income Identify lower bound 2 Income volatility Considered treated if ever in upper brackets 3 The control group comprises the un-treated individuals who were employed

36 Differences in Differences The design of the reform created a natural experiment Individuals earning less than 50,000 rubles a year constitute the control group People with higher incomes faced lower tax rates and are therefore considered treated Problems in practice: 1 Income misreporting Use post-reform income Identify lower bound 2 Income volatility Considered treated if ever in upper brackets 3 The control group comprises the un-treated individuals who were employed

37 Differences in Differences The design of the reform created a natural experiment Individuals earning less than 50,000 rubles a year constitute the control group People with higher incomes faced lower tax rates and are therefore considered treated Problems in practice: 1 Income misreporting Use post-reform income Identify lower bound 2 Income volatility Considered treated if ever in upper brackets 3 The control group comprises the un-treated individuals who were employed

38 Differences in Differences The design of the reform created a natural experiment Individuals earning less than 50,000 rubles a year constitute the control group People with higher incomes faced lower tax rates and are therefore considered treated Problems in practice: 1 Income misreporting Use post-reform income Identify lower bound 2 Income volatility Considered treated if ever in upper brackets 3 The control group comprises the un-treated individuals who were employed

39 Differences in Differences The design of the reform created a natural experiment Individuals earning less than 50,000 rubles a year constitute the control group People with higher incomes faced lower tax rates and are therefore considered treated Problems in practice: 1 Income misreporting Use post-reform income Identify lower bound 2 Income volatility Considered treated if ever in upper brackets 3 The control group comprises the un-treated individuals who were employed

40 Differences in Differences The design of the reform created a natural experiment Individuals earning less than 50,000 rubles a year constitute the control group People with higher incomes faced lower tax rates and are therefore considered treated Problems in practice: 1 Income misreporting Use post-reform income Identify lower bound 2 Income volatility Considered treated if ever in upper brackets 3 The control group comprises the un-treated individuals who were employed

41 Summary Statistics by Treatment Control Treated All Employed Female Age Medium Ed Comp College Ed Comp Schooling (Yrs) Work Experience Married Urban Location Russian National Russian Born Size HH # Fem HH # Youth HH # Elderly HH Obs 17,404 68,475 85,879 Indiv 3,545 11,487 15,032

42 Informal Employees by Treatment Control Treated

43 Informal Entrepreneurs by Treatment Control Treated

44 Informal 2nd Job by Treatment Control Treated

45 Informal Irreg Act by Treatment Control Treated

46 DID FE INF it = θ t + X it β + Z i γ + ψp ost t + µt reat i + α(t reat i P ost t ) + c i + ɛ it Informal Employee Informal Irregular Activities Any Informal Employment Household Characteristics YES YES YES Individual Characteristics YES YES YES Year Dummies YES YES YES DID Estimates P ost (0.099) (0.075) (0.119) T reat P ost ** *** *** (0.010) (0.010) (0.014) Constant * (0.306) (0.232) (0.365) Obs 44,452 53,769 47,718 # of Indiv 11,263 12,411 11,969 R 2 Overall Notes: RLMS, rounds VIII XVIII ( ). Nine year dummies were included but not reported. ***p < 0.01, **p < 0.05, *p < 0.1.

47 Some Robustness Checks Informal Employee Informal Irreg Activ Any Informal Employment All Irregular Activ Informal Irreg Activ as Main Job Baseline ** *** *** *** *** (0.010) (0.010) (0.014) (0.010) (0.009) Including interactions ** *** *** *** *** District Y ear (0.011) (0.010) (0.015) (0.011) (0.009) Control group excludes ** *** *** *** *** unreported income (0.010) (0.010) (0.014) (0.011) (0.009) Treatment defined using ** ** *** ** ** income from all sources (0.014) (0.011) (0.019) (0.013) (0.011) Treatment defined using *** *** *** *** 2001 labor income only (0.012) (0.014) (0.019) (0.015) (0.010) Treatment defined using ** *** *** *** *** labor income (0.011) (0.010) (0.015) (0.011) (0.009) T reat T rend ** *** *** *** *** (0.003) (0.003) (0.003) (0.003) (0.003) P lacebo Reform (0.012) (0.015) (0.019) (0.016) (0.010) Notes: RLMS, rounds VIII XVIII ( ). Includes a post-reform time trend (2000 = 1) instead of the post-reform dummy. ***p < 0.01, **p < 0.05, *p < 0.1.

48 ATT Semi-parametric Estimation M DID ˆ = 1 [(INF i,t INF i,2000) W (i, j)(inf j,t INF j,2000)] N T,t i T j C

49 ATT Semi-parametric Estimation Treat Treat Treat 2001 only Figure: Informal Irregular Activities

50 ATT Semi-parametric Estimation Treat Treat Treat 2001 only Figure: Informal Employees

51 Detailed Treatment Groups 4 INF it = θ t + X it β + Z i γ + ψp ost t + µ h T reat h i + 4 α h (T reat h i P ost t) + u it h=1 h=1 Informal Employee Informal Irregular Activities Any Informal Employment P ost (0.099) (0.075) (0.120) T reat 1 P ost * * (0.012) (0.012) (0.017) T reat 2 P ost * *** *** (0.013) (0.013) (0.018) T reat 3 P ost ** *** *** (0.011) (0.012) (0.016) T reat 4 P ost *** * * (0.014) (0.015) (0.020) Obs 44,452 53,769 47,718 # of Indiv 11,263 12,411 11,969 R 2 Overall Notes: RLMS, rounds VIII XVIII ( ). ***p < 0.01, **p < 0.05, *p < 0.1.

52 Weighted DID n M = ω i [INF it θ t X it β ψp ost t α(t reat i P ost t ) u it ] 2 i=1 ω i = ( ) Yit 3625 n ( ) Yit 3625 K / K h i=1 h Informal Employee Informal Irregular Activities Any Informal Employment P ost (0.121) (0.063) (0.141) T reat P ost * ** (0.019) (0.019) (0.027) Obs 41,930 50,914 45,134 R 2 Overall # of Indiv 10,180 11,220 10,856 Notes: RLMS, rounds VIII XVIII ( ). Treatment effect estimated by a weighted fixed effects regression. ***p < 0.01, **p < 0.05, *p < 0.1.

53 Extensive Margin Informal Employee Informal Irreg Activ Any Informal Employment A. Baseline P ost *** *** *** (0.093) (0.058) (0.114) T reat P ost *** *** (0.025) (0.023) (0.027) Obs 21,224 24,924 22,899 # of Indiv 7,339 8,080 7,709 R 2 Overall B. Robustness Tests Including interactions *** *** District Y ear (0.025) (0.023) (0.028) Treatment defined using *** *** income from all sources (0.029) (0.026) (0.031) Control group excludes *** *** unreported income (0.025) (0.023) (0.027) T reat T rend *** *** (0.004) (0.004) (0.005) Notes: RLMS, rounds VIII XVIII ( ). Sample restricted to those unemployed just before the reform and who were employed at least once in the post-reform period. The dependent variable is set to zero in round 9. Round 8 is excluded. Includes a post-reform time trend (2000 = 1) instead of the post-reform dummy. ***p < 0.01, **p < 0.05, *p < 0.1.

54 Summary of Findings There is evidence that the tax reform reduced participation in informal employment The effect was significant economically and statistically for informal employees ( 2.5%) and for irregular activities ( 4.0%). Semi-parametric DID estimates are higher. The reform had permanent effects. No evidence of an effect on informal entrepreneurs or on the second job Robust to different specifications. Stronger effect on higher income brackets

55 Summary of Findings There is evidence that the tax reform reduced participation in informal employment The effect was significant economically and statistically for informal employees ( 2.5%) and for irregular activities ( 4.0%). Semi-parametric DID estimates are higher. The reform had permanent effects. No evidence of an effect on informal entrepreneurs or on the second job Robust to different specifications. Stronger effect on higher income brackets

56 Summary of Findings There is evidence that the tax reform reduced participation in informal employment The effect was significant economically and statistically for informal employees ( 2.5%) and for irregular activities ( 4.0%). Semi-parametric DID estimates are higher. The reform had permanent effects. No evidence of an effect on informal entrepreneurs or on the second job Robust to different specifications. Stronger effect on higher income brackets

57 Summary of Findings There is evidence that the tax reform reduced participation in informal employment The effect was significant economically and statistically for informal employees ( 2.5%) and for irregular activities ( 4.0%). Semi-parametric DID estimates are higher. The reform had permanent effects. No evidence of an effect on informal entrepreneurs or on the second job Robust to different specifications. Stronger effect on higher income brackets

58 Summary of Findings There is evidence that the tax reform reduced participation in informal employment The effect was significant economically and statistically for informal employees ( 2.5%) and for irregular activities ( 4.0%). Semi-parametric DID estimates are higher. The reform had permanent effects. No evidence of an effect on informal entrepreneurs or on the second job Robust to different specifications. Stronger effect on higher income brackets

59 Summary of Findings Very strong effect on the extensive for irregular activities ( 14%) No extensive margin effect for informal employees These findings are consistent with recent literature on the economics of taxation: high behavioral elasticity but low labor supply elasticity In my opinion, the reform experiment also provides strong evidence that informality is a voluntary state

60 Summary of Findings Very strong effect on the extensive for irregular activities ( 14%) No extensive margin effect for informal employees These findings are consistent with recent literature on the economics of taxation: high behavioral elasticity but low labor supply elasticity In my opinion, the reform experiment also provides strong evidence that informality is a voluntary state

61 Summary of Findings Very strong effect on the extensive for irregular activities ( 14%) No extensive margin effect for informal employees These findings are consistent with recent literature on the economics of taxation: high behavioral elasticity but low labor supply elasticity In my opinion, the reform experiment also provides strong evidence that informality is a voluntary state

62 Summary of Findings Very strong effect on the extensive for irregular activities ( 14%) No extensive margin effect for informal employees These findings are consistent with recent literature on the economics of taxation: high behavioral elasticity but low labor supply elasticity In my opinion, the reform experiment also provides strong evidence that informality is a voluntary state

63 Background Characteristics All Employed Informal Employee Informal Entrepr. Informal Sec. Job Informal Irreg. Activ Female Age College Degree Schooling (Yrs) Work Experience Married Urban Location Russian National Russian Born Size HH After Tax Income This Job (rubles) 13,194 11,043 18,661 7,142 7,043 % Reported for Tax NA NA All Jobs (rubles) 13,446 11,132 18,878 17,024 12,470 Obs

64 All Employed Job Characteristics Informal Employee Informal Entrepr. Informal Sec. Job Tenure (Yrs) Changed Jobs NA Changed Occupation NA Has Subordinates Firm Characteristics Ent Size (# of Emp) State Owns Share Russian Indiv Owns Share Firm from Soviet times Firm owes money Firm pays in kind Job Benefits Paid Vacation Paid Sick Leave NA Paid Maternity Leave Paid Health Care Paid Trips to Sanatoria Paid Child Care Obs

65 Informality in the last 12 months All Employed Informal Employee Informal Entrepr. Informal Sec. Job Inf. Irreg. Activ Worked extra job Raised cattle for sale Agric. on own plot for sale Performed services for pay Obs

66 Distribution by Occupation All Employed Inf Emp Inf Entrep Inf Sec Job Inf Irreg Act 1-digit ISCO Occup Main Job Main Job Main Job Legislators, Sen Manag, Officials Professionals Technicians, Assoc Prof Clerks Service and Market Workers Skilled Agric-Fishery Craft and Related Trades Plant-Machine Oper-Assemblers Unskilled Occupations Obs

67 Distribution by Industry All Employed Inf Emp Inf Entrep Inf Sec Job 1-digit Industry Main Job Sec Job Main Job Main Job Food and Other Light Industry Civil Machine Construction Military Industrial Complex Oil and Gas Industry Other Heavy Industry Construction Transportation, Communication Agriculture Government and Public Adm Education Science, Culture Public Health Army, Security Services Trade, Consumer Services Finances Energy (Power) Industry Housing and Communal Services Obs

68 The Effect of Taxation on Informal Employment Evidence from the Russian Flat Tax Reform Fabián Slonimczyk ICEF-Higher School of Economics, Moscow Joint CLMS-LIRT Seminar. November 2011.