Unemployment Insurance Eligibility and the Duration ofEmployment
Laura Khoury, Clément Brébion, Simon Briole
31 mai 2019
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 1 / 32
Motivation
Figure – Evolution of the share and length of fixed-term contracts
Source : DARES.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 2 / 32
Introduction
Potential effect of UI on employment relationships :I Affects the labour supply decision of workers by changing their outside option
I Affects the value of short-term contracts for workers
In this paper :
→ Impact of the minimum employment history condition to be eligible to UIbenefits→ Use of the 2009 French reform which changed this condition.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 3 / 32
Introduction
1 Do employers and employees respond to the increase in UI value byseparating more at the eligibility threshold ?
n Through increasing flows to unemploymentn Through shorter contracts after the reform
2 Does receiving UI benefits affect future employment prospects ?
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 4 / 32
Introduction
I Contribution 1.a : We estimate the impact of UI on on-the-job behaviours(Ortega and Rioux, 2010; Hopenhayn and Nicolini, 2009; Andersen et al.,2015)(Rebollo-Sanz, 2012; Van Doornik et al., 2018; Albanese et al., 2019;Martins, 2016)
I Contribution 1.b : Both at the micro and macro levels
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 5 / 32
Introduction
I Contribution 2 : extensive margin impact of UI benefits both on short- andlong-run outcomes(Landais, 2015; Lalive et al., 2006; Chetty, 2006; Schmieder andVon Wachter, 2016)(Davezies and Le Barbanchon, 2017)
I Contribution 3 : document the effect on job quality while still an on-goingdebate(Nekoei and Weber, 2017; Card et al., 2007; Schmieder et al., 2013; Lalive,2007; Van Ours and Vodopivec, 2008)
→ Comprehensive analysis of flows both in and out of employment
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 6 / 32
Institutional background and data
Table of Contents
1 Institutional background and data
2 Empirical evidence on employment outflowsSeparation responseContract duration
3 Empirical evidence on unemployment outflows
4 Conclusion
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 7 / 32
Institutional background and data
Institutional background
Contributory UI schemes characterised by :
I UI benefits and contributions proportional to past earnings
I Potential benefit duration (PBD) proportional to work history
I Eligibility depends on a minimum employment record → focus of the paper
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 8 / 32
Institutional background and data
Institutional background
Contributory UI schemes characterised by :
I UI benefits and contributions proportional to past earnings
I Potential benefit duration (PBD) proportional to work history
I Eligibility depends on a minimum employment record → focus of the paper
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 8 / 32
Institutional background and data
Institutional background
Reform on April, 1st , 2009 :
Table – Pre and Post reform rules
Minimum employment record Potential benefit durationPre-reform 6 months over the last 22 months Different categories according to work historyPost-reform 4 months over the last 28 months 1 to 1 relationship up to 2 years
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 9 / 32
Institutional background and data
Data
I FH-DADS : linked employer-employee dataset matched with UI data for apanel of individuals (1/12th) between 2003-2012
I DADS built at the individual × establishment × year level (’position’)
I FH only available for claimants
I Work history variable built from DADS→ 75% match between FH and DADS
I Sample restrictions : remove workers ever observed as a civil servant orhome-employed for a private employer
I 2,690,114 individuals accounting for 18,114,742 positions ending between2004 and 2012 and 3,071,283 corresponding UI spells
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 10 / 32
Empirical evidence on employment outflows
Table of Contents
1 Institutional background and data
2 Empirical evidence on employment outflowsSeparation responseContract duration
3 Empirical evidence on unemployment outflows
4 Conclusion
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 11 / 32
Empirical evidence on employment outflows Separation response
Separation response - Unemployment inflows
I Analysis of the separation response at the micro level
I Definition of Panel 1 :n Random subsample of 100,000 individualsn Positions converted into biweekly spells starting from the 1st job after thereform over 2 years
n Measure of the transitions from employment to registered UI and tonon-employment
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 12 / 32
Empirical evidence on employment outflows Separation response
Separation response - Unemployment inflows
Figure – Probability to go from employment to registeredunemployment Reg
Source : FH-DADS.Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 13 / 32
Empirical evidence on employment outflows Separation response
Separation response - Unemployment inflows
Figure – Probability to go from employment to non-employment
Source : FH-DADS.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 14 / 32
Empirical evidence on employment outflows Separation response
Separation response - Unemployment inflows
I Both graphs could be rationalised by a change in registration behaviour
I Yet :n Transitions E → UI small share of all transitions E → NEn Cannot distinguish layoffs from quitsn Evidence from other countries (Albanese et al., 2019; Rebollo-Sanz, 2012) andfrom France (Khoury, 2019) point to strategic separation behaviours
n Complementary findings on MMO
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 15 / 32
Empirical evidence on employment outflows Separation response
Separation response - Unemployment inflows
P(layoff )it = α+ β114monthsit + β2Postt + β314monthsit × Postt + uit
Probability of being laid-off Probability to end a fixed-term contract Probability to quit
Eligible=1 0.00012 0.00012 -0.00582*** -0.00598*** 0.00036* 0.00033*(0.000116) (0.000117) (0.000451) (0.000452) (0.000195) (0.000196)
Post=1 -0.00010 -0.00015 0.00179*** 0.00011 0.00054*** 0.00047(0.000116) (0.000215) (0.000451) (0.000829) (0.000195) (0.000360)
Eligible=1 × Post=1 0.00041** 0.00041** 0.00622*** 0.00798*** -0.00007 -0.00001(0.000171) (0.000173) (0.000664) (0.000670) (0.000288) (0.000291)
Month FE - yes - yes - yesConstant 0.00090*** 0.00075*** 0.01594*** 0.02096*** 0.00240*** 0.00195***
(0.000081) (0.000146) (0.000314) (0.000565) (0.000136) (0.000245)Observations 546535 546535 554500 554500 547528 547528* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The treatment (respectively control) group is made of individuals ending their contract between 4 and 5 months (resp. 3 and 4 months).The time window corresponds to 6 months before and after the reform.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 16 / 32
Empirical evidence on employment outflows Contract duration
Contract duration
I Employers may internalise the jump in the value of unemployment at thetime of hiring
I Triple advantage of contracts scheduled to qualify workers for benefits :n They ease adjustments of the workforce to the variation in business activitywith a limited decrease in attractiveness of contracts
n Implicit contract theory : employers and employees may agree to share therent from UI through lower wages (Feldstein, 1976; Baily, 1977)
n If systematic enough, employers’ recalling behaviours can ensure investment infirm-specific human capital without bearing the cost of high employmentprotection
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 17 / 32
Empirical evidence on employment outflows Contract duration
Contract duration
I We analyse the evolution of the duration of contracts at the aggregate level
I Data : the set of last contracts used to open UI entitlements.n From which we remove workers registering with a work history between 4 and6 months
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 18 / 32
Empirical evidence on employment outflows Contract duration
Contract duration
Figure – Share of contracts among contracts of 1 year or less (allseparations) Reg
.02
.04
.06
.08
.1.1
2
Pro
port
ion a
mong c
ontr
acts
equal to
or
less than 1
year
2004 2006 2008 2010 2012 2014
1 to 2−monthcontracts
4 to 5−monthcontracts
6 to 7−monthcontracts
10 to 11−monthcontracts
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 19 / 32
Empirical evidence on employment outflows Contract duration
Contract duration
Figure – Share of contracts among contracts of 1 year or less (end offixed-term contract)
0.0
5.1
.15
Pro
port
ion a
mong c
ontr
acts
equal to
or
less than 1
year
2004 2006 2008 2010 2012 2014
1 to 2−monthcontracts
4 to 5−monthcontracts
6 to 7−monthcontracts
10 to 11−monthcontracts
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 20 / 32
Empirical evidence on employment outflows Contract duration
Contract duration
Figure – Share of contracts among contracts of 1 year or less (otherdismissals)
0.0
5.1
.15
Pro
port
ion a
mong c
ontr
acts
equal to
or
less than 1
year
2004 2006 2008 2010 2012 2014
1 to 2−monthcontracts
4 to 5−monthcontracts
6 to 7−monthcontracts
10 to 11−monthcontracts
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 21 / 32
Empirical evidence on unemployment outflows
Table of Contents
1 Institutional background and data
2 Empirical evidence on employment outflowsSeparation responseContract duration
3 Empirical evidence on unemployment outflows
4 Conclusion
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 22 / 32
Empirical evidence on unemployment outflows
Impact on unemployment outflows - Methodology
I Qualifying for benefits 2 months sooner may affect decisions regardingemployment
I Separation response makes it hard to study the extensive margin impact of UIbenefits on U outcomes
I Further sample restrictions and use of the reformn Workers with 4months ≤ work history ≤ 6months before (controls) and after(treated) the reform
n Under a fixed-term contract that started before the reformn Fuzzy RDD around the time threshold of the reform on 23,559 observations
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 23 / 32
Empirical evidence on unemployment outflows
Impact on unemployment outflows - Methodology
Assumption to be checked :
I No disproportionately high number of contracts ≤ 4 months starting beforethe reform and ending right after Graphs
I McCrary test (McCrary, 2008) Graph
I Continuity of the covariates Graph
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 24 / 32
Empirical evidence on unemployment outflows
Impact on unemployment outflows - 1st stage
Figure – Probability of opening a UI right Reg
0.00
0.10
0.20
0.30
0.40
Bein
g r
egis
tere
d to U
I
−1.00 0.00 1.00
Ending date of the contract
Upper and Lower 95% interval Predicted Within−bin mean
Binsize: .03, Number of observation: 23559
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 25 / 32
Empirical evidence on unemployment outflows
Impact on employment probability
Figure – Impact of UI eligibility on employment probability 6 monthsafter Reg
0.30
0.40
0.50
0.60
0.70
Pro
babili
ty o
f bein
g e
mplo
yed 6
month
s a
fter
−1.00 0.00 1.00
Ending date of the contract
Upper and Lower 95% interval Predicted Within−bin mean
Binsize: .03, Number of observation: 23559
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 26 / 32
Empirical evidence on unemployment outflows
Impact on employment path
Table – Impact of UI benefit receipt on the number ofemployment spells over the following two years
Number of employment spells over the following 2 yearsLinear Quadratic Cubic
RD_Estimate -9.795** -9.298* -9.336(4.076) (5.059) (6.296)
Observations 23559 23559 23559
* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The bandwidth has been computed using the MSE optimal bandwidth selector. Linear, quadratic andcubic specifications.The dependant variable corresponds to the number of employment spells over the two years followingthe end of the contract that defines the treatment status.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 27 / 32
Empirical evidence on unemployment outflows
Impact on employment path
Table – Impact of UI benefit receipt on the total numberof days employed over the following two years
Total number of days employed over the following 2 yearsLinear Quadratic Cubic
RD_Estimate -1027.132* -737.503 -1084.775(566.452) (739.947) (837.905)
Observations 23559 23559 23559
* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The bandwidth has been computed using the MSE optimal bandwidth selector. Linear, quadratic andcubic specifications.The dependant variable corresponds to the total number of days employed over the two years followingthe end of the contract that defines the treatment status.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 28 / 32
Empirical evidence on unemployment outflows
Impact on job quality
Table – Summary of the results on job quality
Unconditionally Conditionally Lee boundsShort-term Medium-term Long-term Next contract Short-term Medium-term Long-term
Permanent contract (-) - (-) (+) (+) n.s. n.s.Full-time - (-) n.s. (-) n.s. n.s. n.s.Wage n.s. (-) - - n.s. n.s. n.s.Duration - - - (-) n.s. n.s. n.s.Matching - n.s. n.s. n.s. n.s. n.s. n.s.
A single + or - means that the result is significant and of the indicated sign. A (+) or (-) in parentheses means that there is a strong pattern of the effectgoing in the indicated direction but coefficients are not significant.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 29 / 32
Conclusion
Table of Contents
1 Institutional background and data
2 Empirical evidence on employment outflowsSeparation responseContract duration
3 Empirical evidence on unemployment outflows
4 Conclusion
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 30 / 32
Conclusion
Conclusion
I Comprehensive evaluation of the 2009 reform affecting the eligibility criterionto UI benefits
I Suggestive evidence of :n A strategic separation response at the eligibility thresholdn May contribute to the increase in short-term contracts at the aggregate level
I One of the first empirical evidence on the effect of receiving any UI benefits :n Strong and long-lasting neg. impact on employment probabilityn Evidence of neg. impact on earningsn Less clear-cut results on job quality
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 31 / 32
Conclusion
Appendix
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 32 / 32
Table – Discontinuity in the transition rate fromemployment to registered unemployment Graph
Transition E → UI probabilityLocal Quadratic Cubic
RD_Estimate 0.008*** 0.007*** 0.010***(0.001) (0.001) (0.001)
Observations 1270880 1270880 1270880
* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The regression shows in a regression discontinuity design spirit the discontinuity in thebiweekly transition rate from employment to UI. The running variable is the work historyover the last 28 months and the cutoff value is 4 months. Bandwidth has been computedusing the mean squared error (MSE) optimal bandwidth selector with a linear specification.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 1 / 19
Table – Post-reform evolution of 4 to 5-month contracts(6-month and 1 to 3-month control group) Graph
Proportion among contracts ≤1 year Proportion among contracts ≤1 yearPost=1 -0.00515*** -0.00500***
(0.0000395) (0.0000549)4 to 5-month contracts=1 -0.06441*** -0.06436***
(0.0000649) (0.0001418)Post=1 × 4 to 5-month contracts=1 0.01470*** 0.01451***
(0.0000811) (0.0001751)Constant 0.08778*** 0.08758***
(0.0000298) (0.0000414)Observations 122867 299939
* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The first specification shows the difference in the evolution of the shares of 4-month and 6-month contracts. Thesecond specification compares 4-month contracts to contracts between 1 and 3 months.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 2 / 19
Figure – Share of 1 to 30-day contracts among contracts of 1 year orless Back
.1.2
.3.4
.5.6
Pro
port
ion a
mong
contr
acts
of 1 y
ear
or
less
01jan2009 01mar2009 01may2009 01jul2009 01sep2009
Ending date of the contract
2009 2010
2011
Source : FNA.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 3 / 19
Figure – Share of 31 to 60-day contracts among contracts of 1 year orless
.05
.1.1
5.2
Pro
port
ion a
mong
contr
acts
of 1 y
ear
or
less
01jan2009 01mar2009 01may2009 01jul2009 01sep2009
Ending date of the contract
2009 2010
2011
Source : FNA.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 4 / 19
Figure – Share of 61 to 90-day contracts among contracts of 1 year orless
.05
.1.1
5.2
Pro
port
ion a
mong
contr
acts
of 1 y
ear
or
less
01jan2009 01mar2009 01may2009 01jul2009 01sep2009
Ending date of the contract
2009 2010
2011
Source : FNA.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 5 / 19
Figure – Share of 91 to 120-day contracts among contracts of 1 year orless
.05
.1.1
5.2
Pro
port
ion a
mong
contr
acts
of 1 y
ear
or
less
01jan2009 01mar2009 01may2009 01jul2009 01sep2009
Ending date of the contract
2009 2010
2011
Source : FNA.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 6 / 19
Figure – Share of 121 to 150-days contracts among contracts of 1 yearor less
0.0
5.1
.15
.2.2
5
Pro
port
ion a
mong
contr
acts
of 1 y
ear
or
less
01jan2009 01mar2009 01may2009 01jul2009 01sep2009
Ending date of the contract
2009 2010
2011
Source : FNA.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 7 / 19
Figure – Mc Crary test contract ending date distribution Back
Source : FH-DADS.The running variable, ending date of the contract, has been normalised around the time threshold to beequal to be equal to 0 at the threshold, -1 six months before, and +1 six months after.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 8 / 19
Figure – Mc Crary test contract ending date distribution
Source : FH-DADS.The running variable, ending date of the contract, has been normalised around the time threshold to beequal to be equal to 0 at the threshold, -1 six months before, and +1 six months after.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 9 / 19
Figure – Histogram of the contract ending date frequencies
Source : FH-DADS.The running variable, ending date of the contract, has been normalised around the time threshold to beequal to be equal to 0 at the threshold, -1 six months before, and +1 six months after. The bin size isequal to 0.02.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 10 / 19
Figure – Magnitude of the difference in covariates at the cutoff Back
−50 0 50 100
Gender Education
Wage Fulltime
Establishment size
The three levels of significance of the confidence intervals depicted are 5%, 1% and 0.1%.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 11 / 19
Table – Impact of separating after the reform on UItakeup Back
Being registered to UILinear Quadratic Cubic
RD_Estimate 0.069*** 0.061** 0.058**(0.021) (0.024) (0.024)
F-stat 72.64 46.38 21.14Observations 23559 23559 23559
Bandwidth has been computed using the mean squared error (MSE) optimal bandwidthselector.Standard errors in parentheses.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 12 / 19
Table – Impact of UI benefits receipt on employmentprobability (1-4 months) Back
Probability of beingemployed 1 months
after
Probability of beingemployed 2 months
after
Probability of beingemployed 3 months
after
Probability of beingemployed 4 months
afterRD_Estimate 0.023 0.019 -0.536 -0.902*
(0.556) (0.495) (0.426) (0.488)Observations 23559 23559 23559 23559
* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The bandwidth has been computed using the MSE optimal bandwidth selector. Linear specification.The dependant variable corresponds to the probability of being employed at different time horizons after the end ofthe contract that defines the treatment status.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 13 / 19
Table – Impact of UI benefits receipt on employmentprobability (5-8 months)
Probability of beingemployed 5 months
after
Probability of beingemployed 6 months
after
Probability of beingemployed 7 months
after
Probability of beingemployed 8 months
afterRD_Estimate -0.895* -1.068** -1.020** -1.141***
(0.492) (0.494) (0.469) (0.342)Observations 23559 23559 23559 23559
* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The bandwidth has been computed using the MSE optimal bandwidth selector. Linear specification.The dependant variable corresponds to the probability of being employed at different time horizons after the end ofthe contract that defines the treatment status.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 14 / 19
Table – Impact of UI benefits receipt on employmentprobability (9-12 months)
Probability of beingemployed 9 months
after
Probability of beingemployed 10months after
Probability of beingemployed 11months after
Probability of beingemployed 12months after
RD_Estimate -0.376 -0.949** -0.978** -0.695(0.381) (0.453) (0.455) (0.438)
Observations 23559 23559 23559 23559
* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The bandwidth has been computed using the MSE optimal bandwidth selector. Linear specification.The dependant variable corresponds to the probability of being employed at different time horizons after the end ofthe contract that defines the treatment status.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 15 / 19
Table – Impact of UI benefits receipt on employmentprobability (13-16 months)
Probability of beingemployed 13months after
Probability of beingemployed 14months after
Probability of beingemployed 15months after
Probability of beingemployed 16months after
RD_Estimate -0.641 -0.719 -0.684 -0.772*(0.453) (0.446) (0.435) (0.438)
Observations 23559 23559 23559 23559
* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The bandwidth has been computed using the MSE optimal bandwidth selector. Linear specification.The dependant variable corresponds to the probability of being employed at different time horizons after the end ofthe contract that defines the treatment status.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 16 / 19
Table – Impact of UI benefits receipt on employmentprobability (17-20 months)
Probability of beingemployed 17months after
Probability of beingemployed 18months after
Probability of beingemployed 19months after
Probability of beingemployed 20months after
RD_Estimate -0.779* -0.979** -0.940** -1.064**(0.461) (0.480) (0.453) (0.486)
Observations 23559 23559 23559 23559
* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The bandwidth has been computed using the MSE optimal bandwidth selector. Linear specification.The dependant variable corresponds to the probability of being employed at different time horizons after the end ofthe contract that defines the treatment status.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 17 / 19
Table – Impact of UI benefits receipt on employmentprobability (21-24 months)
Probability of beingemployed 21months after
Probability of beingemployed 22months after
Probability of beingemployed 23months after
Probability of beingemployed 24months after
RD_Estimate -1.108** 0.069 0.079 0.048(0.444) (0.437) (0.428) (0.424)
Observations 23559 23559 23559 23559
* p<0.05, ** p<0.01, *** p<0.001. Standard errors in parentheses.The bandwidth has been computed using the MSE optimal bandwidth selector. Linear specification.The dependant variable corresponds to the probability of being employed at different time horizons after the end ofthe contract that defines the treatment status.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 18 / 19
Albanese, Andrea, Corinna Ghirelli, and Matteo Picchio, “Timed to saygoodbye : Does unemployment benefit eligibility affect worker layoffs ?,” 2019.
Andersen, Torben M, Mark S Kristoffersen, and Michael Svarer, “BenefitReentitlement Conditions in Unemployment Insurance Schemes,” 2015.
Baily, Martin Neil, “On the theory of layoffs and unemployment,”Econometrica : Journal of the Econometric Society, 1977, pp. 1043–1063.
Card, David, Raj Chetty, and Andrea Weber, “The spike at benefitexhaustion : Leaving the unemployment system or starting a new job ?,”American Economic Review, 2007, 97 (2), 113–118.
Chetty, Raj, “A general formula for the optimal level of social insurance,” Journalof Public Economics, 2006, 90 (10), 1879–1901.
Davezies, Laurent and Thomas Le Barbanchon, “Regression discontinuitydesign with continuous measurement error in the running variable,” Journal ofEconometrics, 2017, 200 (2), 260–281.
Doornik, Bernardus Nazar Van, David Schoenherr, and Janis Skrastins,“Unemployment Insurance, Strategic Unemployment, and Firm-WorkerCollusion,” 2018.
Feldstein, Martin, “Temporary layoffs in the theory of unemployment,” Journalof political economy, 1976, 84 (5), 937–957.
Hopenhayn, Hugo A and Juan Pablo Nicolini, “Optimal unemploymentinsurance and employment history,” The Review of Economic Studies, 2009, 76(3), 1049–1070.
Khoury, Laura, “Unemployment Benefits and the Timing of Redundancies :Evidence from Bunching,” 2019.
Lalive, Rafael, “Unemployment benefits, unemployment duration, andpost-unemployment jobs : A regression discontinuity approach,” AmericanEconomic Review, 2007, 97 (2), 108–112., Jan Van Ours, and Josef Zweimüller, “How changes in financial incentivesaffect the duration of unemployment,” The Review of Economic Studies, 2006,73 (4), 1009–1038.
Landais, Camille, “Assessing the welfare effects of unemployment benefits usingthe regression kink design,” American Economic Journal : Economic Policy,2015, 7 (4), 243–278.
Martins, Pedro S, “Working to get fired ? Regression discontinuity effects ofunemployment benefit eligibility on prior employment duration,” 2016.
McCrary, Justin, “Manipulation of the running variable in the regressiondiscontinuity design : A density test,” Journal of Econometrics, 2008, 142 (2),698–714.
Nekoei, Arash and Andrea Weber, “Does Extending Unemployment BenefitsImprove Job Quality ?,” American Economic Review, 2017, 107 (2), 527–61.
Ortega, Javier and Laurence Rioux, “On the extent of re-entitlement effects inunemployment compensation,” Labour Economics, 2010, 17 (2), 368–382.
Ours, Jan C Van and Milan Vodopivec, “Does reducing unemploymentinsurance generosity reduce job match quality ?,” Journal of Public Economics,2008, 92 (3), 684–695.
Rebollo-Sanz, Yolanda, “Unemployment insurance and job turnover in Spain,”Labour Economics, 2012, 19 (3), 403–426.
Schmieder, Johannes F and Till Von Wachter, “The effects of unemploymentinsurance benefits : New evidence and interpretation,” Annual Review ofEconomics, 2016, 8, 547–581., , and Stefan Bender, “The causal effect of unemployment duration onwages : Evidence from unemployment insurance extensions,” Technical Report,National Bureau of Economic Research 2013.
Khoury, Brébion Unemployment Insurance Eligibility and the Duration of Employment 31 mai 2019 19 / 19
Top Related