Post on 01-Jan-2020
HOW VIABLE ARE UISAs: SIMULATION RESULTS FOR
SLOVENIA
Milan Vodopivec (The World Bank)Bonn, May 4, 2009
Even in developed countries, unemployment insurance (UI) creates a moral hazard
problem; still worse in developing countries
Survival of “identical” groups of UI recipients, Slovenia, before and after change of law (1998)
Probability of Being Unemployed [%]
M o n t hs Un emp loy ed 0 6 12 18 24 30
0
25
50
75
100
Source: van Ours and Vodopivec (2006)
Can other programs do better?
YES Unemployment insurance savings
accounts (UISAs): workers internalize the costs of
benefits
But can we trust UISAs?
1. Only intertemporal risk pooling -- no pooling across individuals
the problem of viability: can workers generate enough savings?
2. Do we have evidence that theoretical predictions of improving job-search incentives really work?
Main messages
• UISAs are a viable option under a modest, less so under the generous UI system
• Yes, UISAs can improve work incentives (we have the first, brand new evidence from Chile)
Route map
1. How the UISA system works2. Methodology and data3. Simulation results4. Evidence from the Chilean UI program5. Concluding remarks
1. How does the UISA system work?
Remember - rules under UI:• Workers and employers pay contributions to
a public fund• Qualifying workers collect benefits, if they
satisfy:– Initial eligibility conditions– Continuing eligibility conditions
• The program prescribes a replacement rate and potential benefit duration (PBD).
How does the UISA system work?
• Contributions are saved on the UISA of an individual worker
• Benefits paid from UISAs (and solidarity fund)• Options
– Pure UISAs– UISAs-cum-borrowing: at retirement, the
account merged with pension account– Combination of UISAs and solidarity fund (Chile)
UISAs with solidarity fund –The Chilean case
2. Methodology and dataMethodology: simulation of the UISA system,
on a representative sample of labor force participants, over their entire working life (Feldstein and Altman, 1998).
==> analysis of the history of contributions to and withdrawals from the UISAs for each worker
• No behavioral changes
Methodology and data==>examining the viability of the system via:
– the fraction of workers who:• end their active lives with negative cumulative
balances on their UISAs; • have had ever negative balances; and
– the taxation rate (to cover terminal negative balances)
==>examining the changes in the redistribution of the system by the replacement of UI with UISA
Distinction of UI regimes
Taking advantage of the 1998 UI law change that strongly reduced potential benefit duration:
• Modest: 2000-01 period• Generous: 1995-98 period
Data sources
• Work history data base, including earnings• Registered unemployment data base,
including the receipt of UB• Entire labor force included (formal sector0
Obtaining complete work history
• Challenge: obtaining a complete working career, based on data spanning a few years
• I generated a synthetic labor force participants (“persons”) by connecting working histories of workers with similar characteristics (“parents”)
• Persons merged by gender, age, education, and type of spell
3. Simulation results
• Definition of the baseline scenario• Viability• Redistribution
Simulation results
Baseline scenario:– contribution rate: 3 percent of person's current wage
– maximum accumulation amount: 3.9 average wages (cover a 6-month unemployment spell)
– duration of benefits: as occurred historically
– replacement rate: 70%, dropping to 60%– no grace period– interest rate: 2 percent
Simulation resultsShare of workers
with negative UISA balance at retirement
Ever negative UISA during working life
Generous Modest Generous Modest
49.2 ± 0.9 24.6 ± 1.3 66.3 ± 1.6 42.5 ± 1.4
Simulation results Average negative
balance at retirement
(in average MW
Fraction financed from UISAs (in percent)
Generous Modest Generous Modest Total 13.5 ±
0.6 7.7 ± 0.4
41.6 ± 1.2
59.1 ± 1.4
Males 14.0 ± 0.8
8.3 ± 0.6
41.8 ± 1.6
55.3 ± 1.5
Females 13.0 ± 0.9
6.8 ± 1.0
41.5 ± 2.1
64.8 ± 4.4
Viability and UISA: using pension funds as the collateral
Estonia: viability if using pension funds as the collateral
Sensitivity analysis: contribution rate
Negative balance Taxation rate
Sensitivity analysis: replacement rate
Negative balance Taxation rate
Redistribution: subsidization/ taxation rate under UI and UISA
4. Evidence from the Chilean UI program
• Theoretical predictions • Evidence from Chile: UISAs component
improves incentives, but moral hazard still present due to SF Source: van Ours, J.C., G. Reyes Hartley, and M. Vodopivec (2009). “Incentive Effects of Unemployment Insurance Savings Accounts: Evidence from Chile” (preliminary results)
Theoretical predictions
By internalizing costs of benefits, UISAs –in comparison to UI – increase: – job-search effort of unemployed workers and– on-the-job effort of employed workers
(while offering the same level of protection).
Source: Orszag and Snower, 1997; Orszag et al, 1999).
Evidence for the job-finding rate from the Chilean UI program
• The larger UISA resources at the start of the spell, the higher the job-finding probability
• For persons not using SF, the accumulation on UISAs does not affect the job-finding probability
• Using SF creates disincentives - exit rate lower than for other groups, increases as the benefit exhaustion approaches
Estimates of for job-finding rate in Chilean system, used and did not use the Solidarity Fund among those entitled to it
(proportionate hazard model accounting for selectivity)
Source: van Ours, Reyes Hartley and Vodopivec (2009)
Job-finding rate for males, having and not having right to the solidarity fund
Source: van Ours, Reyes Hartley and Vodopivec (2009)
5. Concluding remarks
• UISA system is a viable alternative to traditional UI under non-generous regime
• We now also have evidence that UISAs“work”-- they can affect incentives!UISAs provide an attractive alternative to UI, particularly for developing countries (large informal sector, weak monitoring capacity: “Introducing UI to Developing Countries,” Vodopivec 2009)
Concluding remarks
• Under an innovative design: UISAs with borrowing, backed by pension wealth, the program is also suitable as a rapid response to the current crisis
• Many design questions still remain: setting UISA interest rates if allowing for borrowing; split of UISAs and SF for UISA with SF
• Danger of policy time inconsistencies