VoxEU Column Labour Markets

Incentive effects of unemployment insurance savings accounts: Evidence from Chile

How can policymakers provide unemployment insurance while minimising adverse incentives? This column presents new evidence from Chile suggesting unemployment insurance savings accounts can increase job-finding rates. This provides a strong endorsement of the savings account component to reform traditional unemployment insurance programmes.

Unemployment insurance offers financial compensation to workers for income loss due to unemployment – providing they satisfy certain requirements. While such programmes usually provide solid protection against the hardship of job loss, the evidence shows that this protection is typically produced at a cost of increased disincentives to work. The unemployed search less intensely for a job than they would have in the absence of benefits and their reservation wage is higher, making them less likely to take a given job offer (Holmlund 1998 and Vodopivec 2004). This generates a so-called moral hazard problem that has been extensively studied. Hence a natural but difficult question is how to minimise adverse incentives created by unemployment insurance programmes while still providing suitable income protection.

There are several mechanisms that help reduce work disincentives in unemployment insurance benefit programmes: monitoring and benefit sanctions, work requirements, and financial incentives (Fredriksson and Holmlund 2006). Within financial incentives, unemployment insurance savings accounts are among the most radical and perhaps promising ones. Under this system, each worker is required to save a fraction of earnings in his or her account, and draw unemployment benefits from it. By internalising the costs of unemployment benefits, the savings accounts system is expected to reinforce worker incentives and thus to avoid or reduce the moral hazard inherent in traditional unemployment insurance programmes. Under some variants, the insurance savings accounts provide the same protection as the traditional insurance system. The system is thus credited with a potential to substantially decrease overall unemployment and, by lowering payroll taxes, increase wages.

But in contrast to the other mechanisms used to address work disincentives, studies on insurance savings accounts have been rare and nearly exclusively limited to theoretical contributions. Taking advantage of the recently introduced, innovative Chilean unemployment benefit system, our study (Reyes, Van Ours and Vodopivec 2010) is the first one to empirically investigate whether unemployment insurance savings accounts improve work incentives. Our results provide strong support to this idea.

How does the Chilean programme work?

In 2002, Chile introduced a new unemployment insurance programme that combines social insurance with self-insurance. Unemployment contributions, paid by both workers and employers, are split between individual-level insurance savings accounts and a common, solidarity fund, the latter being cofinanced by the government. To stimulate reemployment, benefit recipients first draw resources from their savings accounts and, upon depletion, from the solidarity fund. The potential benefit duration is five months.

Withdrawals from individual accounts are triggered by separation from the employer, regardless of the reason. Insufficient resources on individual accounts trigger withdrawals from the common fund, if the claimant satisfies the usual conditions of continuing eligibility under typical unemployment insurance. Only those who, prior to unemployment, have worked under permanent contracts and were laid off for reasons attributable to the employer, can access solidarity funding. But, even if they do qualify, workers may opt not to choose the option of using solitary funding – presumably to avoid additional conditions on the benefit receipt.

By the end of 2008, the programme had about three million active contributors, representing almost 80% of private sector wage and salary workers, and distributed benefits to more than 100,000 members, approximately one quarter of the unemployed.

New evidence on the incentive effects of the Chilean programme

What kind of work incentives can one expect from the Chilean system? Theoretical modelling of Orszag and Snower (2002) shows that workers who rely on unemployment insurance savings accounts internalise the costs of their unemployment and thus they have the incentive to search harder for jobs than workers not relying on the savings accounts. Applying this logic to the Chilean programme and realising that the Chilean programme is of a “hybrid” nature, we derive the following predictions:

  • For persons eligible to savings accounts only, the amount accumulated on their accounts will not affect their job-finding rate.
  • In contrast, for workers using solidarity funding, we can expect that their job-finding rate will increase in proportion to the share of potential benefits that can be financed by their own savings accounts.
  • Moreover, as the benefit expiration date approaches, we can expect that workers who use the solidarity fund will increase the intensity of their job search or reduce their reservation wage, thereby increasing the rate of job-finding (see Mortensen 1977), and no such effect will be detected for workers not using the solidarity funding.

We investigated the validity of these predictions using administrative records of the Chilean unemployment benefit programme, both the contribution histories and paid benefits, for a sample of prime age men and women who lost a permanent job by 2007 (about 50,000 men and 25,000 women). In Reyes et al. (2010) we analyse the labour market position of men and women separately. Here, for brevity, we only report results for prime age men.

The empirical results provide strong support to the above predictions. Figure 1 shows that here are strong, systematic differences in job-finding and unemployment survival rates between prime age male workers who use the solidarity funding, on one hand, and those who do not use it, either because they are not entitled or because they chose not to use it, on the other. The main discrepancy occurs at the start of the unemployment spell, where the job-finding rate for male workers not using the solidarity funding strongly exceeds the job-finding rate of workers who do not use the fund (Figure 1a). The job-finding rate of the latter group is increasing over the first 5 months of unemployment, and thereafter the differences between the three groups become small. Consistent with above, the survival function shows that male workers who use the solidarity funding stay in unemployment longer than workers who do not use the solidarity funding (Figure 1b).

Figure 1. Job finding rates and survival rates; prime age males

a. Job finding rates

b. Survival rates

The above results are fully corroborated by a rigorous econometric analysis. The job-finding rate of unemployment benefit recipients was estimated by a mixed proportional hazard rate model, assuming that the job-finding rate depended on observed and unobserved characteristics and accounting for the unobserved heterogeneity and selectivity (individuals who decided to use solidarity fund might systematically differed in their unobserved characteristics from those who did not).

For beneficiaries who are not entitled to use the solidarity fund, our results show that the number of potential withdrawals has no effect on the job-finding rate, implying that their savings account accumulations do not affect exit rates. Moreover, the time pattern of the job-finding rate shows no correspondence to payment of unemployment benefits and exerts strong and monotonic negative duration dependence i.e. the longer the duration of the unemployment spell, the less likely it is to exit unemployment.

For the beneficiaries who used the solidarity fund, in contrast, we find that the higher their savings accounts accumulation, the higher the job-finding rate. Moreover, confirming the graphical analysis in Figure 1, our results indicate a monotonous increase in the job-finding rate in the first five months of unemployment, followed by a steady reduction of the exit rate thereafter.

Lastly, for the group of beneficiaries not using the solidarity funding because they chose so, the parameter estimates are very similar to those not using the funding because they did not qualify. For men, job-finding rates are unaffected by the number of potential withdrawals, and the exit rates for both men and women show negative duration dependence throughout the unemployment spell.

Policy implications

Our results render a strong support to the idea that insurance savings accounts can improve work incentives. By providing empirical support to the theoretically grounded claims, they provide a strong endorsement of the introduction of the savings account component to reform traditional unemployment insurance programmes (for the US, see recent proposals by Kletzer and Rosen 2006, and in particular, Kling 2006).

The introduction of savings accounts seems particularly attractive for developing countries, because they face a large informal sector and they lack the administrative capacity needed for an effective implementation of the standard unemployment insurance system, particularly of checking continuing eligibility conditions that requires monitoring of job-search behaviour and of labour market status. Under such circumstances, Vodopivec (2009) proposes savings accounts as the main source of financing, complementing it by solidarity funding to boost consumption smoothing (or allowing the recipients to borrow from their savings accounts), thus relying on financial incentives as the main vehicle of imposing work incentives.

The study also prompts questions about the appropriate design of the savings accounts programme. Our results show that, through its solidarity funding component, the Chilean programme is still prone to work disincentives. Hence there remain questions over what is the appropriate scope for the solidarity component and, moreover, what is the scope for monitoring and sanctions in such a system – as our results show that the likelihood of using solidarity funding depends on the perceived costs.

Another open question relates to the relative size of the savings accounts and solidarity funding components – in fact, our results suggest that a larger savings accounts component (perhaps even by partial government matching of contributions) will encourage job-finding and thus reduce resources needed for solidarity funding. Of course, the ultimate design of any unemployment insurance scheme must also respond to the main objective of the programme; providing adequate consumption smoothing.


Fredriksson, Peter and Bertil Holmlund (2006), “Improving incentives in unemployment insurance: a review of recent research”, Journal of Economic Surveys, 20: 357-386.

Holmlund, Bertil (1998), “Unemployment insurance in theory and practice”, Scandinavian Journal of Economics, 100: 113-141.

Kletzer Lori G and Howard F Rosen (2006), “Reforming unemployment insurance for the twenty-first century workforce”, Brookings Institution, Discussion paper 2006-06.

Kling Jeffrey R. (2006), “Fundamental restructuring of unemployment insurance: wage-loss insurance and temporary earnings replacement accounts”, Brookings Institution, Discussion paper 2006-05.

Orszag, Michael and Dennis Snower (2002), “From unemployment benefits to unemployment accounts”, IZA Discussion Paper 532, Bonn.

Reyes, Gonzalo, Jan C van Ours and Milan Vodopivec (2010), “Incentive Effects of Unemployment Insurance Savings Accounts: Evidence from Chile”, CEPR, Discussion Paper 5971.

Vodopivec, Milan (2004), “Income support for the unemployed: issues and options”, mimeo, World Bank, Regional and Sectoral Studies Series.

Vodopivec, Milan (2009), “Introducing unemployment insurance to developing countries”, IZA, Policy Paper 6, Bonn.

1,155 Reads