Impact assessment of targeted wage subsidies using administrative data

Transcription

1 Zsombor Cseres-Gergely IE-HAS, Budapest Institute Árpád Földessy Budapest Institute, UCL Ágota Scharle Budapest Institute Impact assessment of targeted wage subsidies using administrative data Eastern-Central Regional Workshop on Counterfactual Impact Evaluation Budapest,

2 Outline Relevance of wage subsidies Hungarian context and details of START scheme Identification strategy Data Comparison of treatment and control group Regression analysis Results Issues for further analysis 24 Sept. 2015

3 Relevance Recent global financial crisis aggrevates low-skilled unemployment/inactivity in post-socialist economies Wage subsidies and services/sanctions proved to be most effective in increasing reemployment rates (Kluve 2006) But evidence on wage subsidies is mixed, especially for CEE (Betcherman et al 2004, Kluve 2006, Card et al 2009, Schünemann et al, 2011). Problems expensive: cost efficiency may be poor may stigmatise recipients deadweight and substitution costs Few studies in transition countries and mixed results (Betcherman et al 2004, O Leary 1998, Csoba and Nagy 2011) transition environment or bad policy design?

4 Targeted wage subsidies Typical features of a targeted wage subsidy scheme Paid to employer Conditional on hiring additional worker belonging to target group Target group faces some disadvantage Tied to wage costs (lump sum/worker or % of gross wage with ceiling) May require continued employment Aims to reduce temporary (e.g. discrimination, lack of experience) or permanent (e.g. disability) disadvantage

5 CEE/Hungarian context Long term non-employment by education, 2012 low emprate for uneducated and older workers persists in Hu: high share of low educated

6 Policy development Job creation subsidies since early 1990s Low admin wage subsidy for youth in Oct 2005 (START) For long term unemployed in July 2007 (START plusz and extra ) well targeted/designed scheme in ~stable environment evaluation may help clarify the effectiveness of wage subsidies in transition countries

7 Table 1. Rules of the various START schemes at the time of introduction Name Eligibility Amount of subsidy (% of employers contribution)* Ceiling on subsidy *** Date of introduction 1st year 2nd year START START Plusz School leavers: below 25 (30 for graduates), no prior paid job On parental leave or care allowance, or registered unemployed for 12 months within preceding 16 months, not eligible for old age pension 53 % HUF / month 22 % HUF /month 1.5x min wage (2x for gradutes) Oct x min wage July 2007 START Extra Over 50 or primary education only, and registered unemployed for 12 months within preceding 16 months, not eligible for old age pension 100 % HUF /month 53 % HUF /month 2x min wage July 2007 Receiving unemployment assisstance, not eligible for old age pension Jan * In 2007, the employer s contribution was 32% of the gross wage, and this was waived in full during the first year of employing a person with a START extra voucher. The flat rate health contribution was waived in both years in all schemes, which was 1950 HUF a month (about 8 EUR), or around 3% of the minimum wage. Source: Cseres-Gergely and Scharle (2009)

8 START scheme conditions registered unemployed for 12 months in preceding 16 months and aged aged primary education extra extra secondary or higher plusz extra subsidy on employers contributions: plusz 53 % in 1st year, 22 % in 2nd extra 100 % in 1st year, 53 % in 2nd

9 START scheme details Quasi-voucher: offers temporary reduction on payroll tax (social security contribution) job seekers claim a plastic card issued by the tax authority before they start the job almost automatic: <2 % of claims declined by tax authority eligibility starts on day when card is issued NOT when hired Claims closely linked to employment 6109 START extra cards issued bween July 2007 and Dec employed in November 2008 Take-up started to grow rapidly after February 2008 No obligation to continue employment after the subsidised period Total cost of the scheme is lower than re-training or public work programmes: annual 1.8 billion HUF (7.2 million EUR)

10 Identification strategy Focus on the employment effect of the additional subsidy for educated older workers Compare similar individuals slightly below and above 50 Use the age condition in the START extra scheme Regression-discontinuity analysis Note that the probability of re-employment decreases with age

11 Identification strategy Extra offers a subsidy almost twice as high as plusz for long term unemployed older workers aged aged primary education extra extra secondary or higher plusz: control extra: treat

12 Control and treatment groups 3 year window for the control group (age ) 2.5 years for the treatment group (age ) 18 month gap in between (will be eligible) Wider age range: more observations but estimates more prone to age-related effects Jobs seekers with primary education are excluded (no age condition for them) Treatment = eligible for voucher NOT claim cf: incentive to claim only after job offer Control = not eligible only because below age 50

13 Dataset Institute of Economics, Hungarian Academy of Sciences Matched records of 50 % of adult population, admin data from 1) health and 2) pension insurance 3) the treasury 4) unemployment register From January 2002 to December 2008 Information on labour market status in each month 2043 workers in the treated, 2195 in the control group (long-term unemployed in June 2007) Dependent variable is timing of the first exit to job

14 Raw outcomes Table 2. Exits to job / survival in unemployment and employed spells after July 2007 Men Stays (N) Exits (N) Exits (%) Total (N) employed spells (month) Control Treatment Difference Women Control Treatment Difference average exit rate for total observed period is 0.03 for both T and C

15 Group differences Modest or no treatment effect at first sight: difference in exit rates is zero for men and negative for women Treated and control group ~same minor differences in labour market history But reemployment probability decreases with age and the 7-year window is fairly large

16 Figure 5. Cumulative exits for age cohorts apart and near MEN time time aged 52 (t) aged 45 (c) aged 50 (t) aged 48 (c)

17 WOMEN time time aged 52 (t) aged 45 (c) aged 50 (t) aged 48 (c) Note: (c) control group (t) treatment group

18 Regression analysis T = E(I P=1) E(I P=0) T' = E(I P=1, X=x1) E(I P=0, X=x0) Important to control for age and other observed (small) differences between the treatment and the control group Discrete time duration model following Jenkins (1995) with logistic hazard for exit after T66 (intro of subsidy) Unit of analysis is the spell month Dummy for each observation (person*month): (0) continued non-employment (1) or exit to a job

19 Main results 15 months after Employment effects of older job-seekers 18 months after Calendar time duration CT duration for lower sec vocational Men ** *** ** Women *** p<0.01, ** p<0.05, * p<0.1 See full regression results in Tables B1 and B2 in the Appendix. Significant positive effect for men after controlling for age, education and participation in training (in the 2 years preceding the treatment) Larger effect for lower secondary vocational vs higher level of edu Positive but not significant for women robust to definition of employment and unemployment in the data There is no treatment effect when age is omitted No significant substitution effect

20 More work to do Why no effect for women? Less likely to actively look for work? Examine wage outcomes What is the effect of the subsidy on reemployment probabilities for the long term unemployed below 50? (effect of plusz ) Add further job history variables Reestimate on longer time span to check long term impact

21 Thank you for your attention more detail in Cseres-Scharle-Földessy BWP 2015/3 Impact assessment of targeted wage subsidies

22 Inactivity and long-term unemployment by age - LFS before July 2007