In 2008-2009, like most economies, the US was immersed in the worst crisis since the Great Depression. As a result unemployment went from 5% in 2007 to 10% in 2010. Given this situation, the US Congress decided to extend the maximum entitlement period for unemployment insurance from the 26 weeks in force at that time to 99 weeks. Extending this benefit was a success in coverage terms (benefitting 18 million people in the first two years) but was nevertheless controversial with public debate regarding its possible impact on slowing down the employment's subsequent recovery.1 What lies behind this controversy?

This debate is actually a new version of the academic discussion regarding the repercussions of unemployment insurance. Before reviewing these, it is useful to describe the features that constitute the typical unemployment insurance in advanced countries. First of all, it is important to note that, as the name suggests, unemployment insurance is a means of insuring against the possibility of losing a job. It is therefore, and we repeat, insurance against a risk, insurance which, moreover, is redistributive in nature because, to some extent, there is a transfer between those who are less likely to lose their job and those who are more likely to experience a loss.

How is this insurance structured? The most usual system is as follows: workers who lose their job involuntarily and who meet certain criteria (particularly a minimum prior employment period), receive an unemployment payout whose amount, in most countries, is related to their most recent salary. This relationship between a worker's prior wage (or income) is known, in economics, as the replacement rate. This payout is generally received for a maximum period of time and tends to decrease as this period runs out. Entitlement to unemployment insurance is subject to meeting certain conditions which vary slightly from country to country: in France only those who are seeking employment in a «proved and continued» way receive unemployment insurance while those rejecting a job offer in the United Kingdom or Austria may have their benefit suspended temporarily or permanently.

In spite of tending to share a common system, advanced countries vary considerably in two key variables of unemployment insurance: how much is paid and for how long. The amount is usually calculated using the replacement rate. Limiting our analysis to EU countries, this rate varies considerably, as can be seen in the graph: while this is above 70% in Latvia, Luxembourg, Bulgaria, Portugal, the Netherlands and Croatia, in the case of Malta, Greece and the United Kingdom it falls to below 30%. The maximum period for receiving unemployment insurance also varies greatly. In the EU, in 2010, this period ranged from 21 weeks in Lithuania to an unlimited period in practice in Belgium. In general terms, the maximum period in many EU countries is either around six months or one year. Having looked at the unemployment insurance system, let us now return to our discussion of its various effects. As has already been mentioned, it is designed to fulfil a primary function: namely ensuring income when employment is lost involuntarily. The reason for this guarantee is to help individuals to gradually adjust their consumption to their unemployed status. As a rule, this function is duly fulfilled; according to estimates by Gruber (1997) for the US, unemployed people experience a 6% drop in consumption compared with when they were employed while, without insurance, this drop would have been 22%.2 In addition to this vital function, unemployment insurance also improves the process known as labour market matching in economics. This consists of finding a job that is more in line with the characteristics and preferences of the person seeking employment. Unemployment insurance allows job seekers to be more selective in the kind of employment they are looking for and therefore matches their skills to their job as closely as possible.

Beyond the individual benefits of unemployment insurance, the satisfactory fulfilment of these two functions ensures that, overall, an economy functions more efficiently, with a reduction in aggregate consumption that is more in line with the optimal reduction in a recession and with gains in productivity thanks to better labour market matching. However, unemployment insurance also has negative effects on economic efficiency. First of all it affects the effort made to find a job and lowers the minimum wage the unemployed person is willing to accept to return to work (the so-called reservation wage). As a result people are less likely to accept a new job as the amount of the unemployment insurance increases.3 Moreover, empirical literature has repeatedly shown that the likelihood of finding a job increases as the end of the coverage period approaches.4 Lastly, at a more general level, workers' negotiating powers improve when there is unemployment insurance, leading to better wages and a higher equilibrium level of unemployment.5

In summary, after reviewing the effects of unemployment insurance it seems clear that the most optimal system is one which reduces the incentives not to seek employment while also maintaining all or most of the function of guaranteeing income. In practice two different approaches tend to be used to achieve this aim. Firstly, the system can be reformed by increasing state involvement to reinforce the active search for a job while the insurance is being received. According to available evidence, this method tends to produce positive results when applied correctly6 and it also helps to link unemployment insurance with active policies.

A second way to bring unemployment insurance closer to its optimal design is by using financial mechanisms that reduce the disincentives to actively seek employment. Along these lines we should mention Chile's reform of its unemployment insurance in 2002. Essentially this system includes both individual insurance savings accounts and a common solidarity fund, ensuring the system is redistributive in nature. The solidarity fund is co-financed by employers, employees and the state while workers and companies are responsible for the individual savings accounts. The system basically works as follows: when workers become unemployed they are paid from their own individual account and can only access the solidarity fund under certain conditions, such as when their individual fund has run out or if, when they became unemployed, they had a permanent contract and were dismissed for reasons exclusively attributable to the company. In this last case the unemployed person can opt to remain outside the solidarity fund (the conditions to be entitled to a payout from the solidarity fund are stricter). In short, the system is designed to encourage workers to actively seek employment while also guaranteeing an income, since the sooner they return to work the less of a drain on their individual savings account.7

In conclusion, the Chilean experience, and other good practices that appropriately combine active and passive policies via appropriate public and administrative action, remind us that, although the optimal design of unemployment insurance is surely not achievable, we are capable of designing alternatives that at least come close to achieving this aim. Given the social and economic drama of unemployment, any step in the right direction is worthy of consideration.

Àlex Ruiz

Macroeconomics Unit, Strategic Planning and Research Department, CaixaBank

1. This debate led to the Congressional Budget Office admitting that the programme had not had much effect on creating jobs.

2. See Gruber, J. (1997), «The consumption smoothing benefits of unemployment insurance», The American Economic Review, 87, 1.

3. Specifically, Bover, Arellano and Bentolila (2002) estimate that an individual without benefits who has remained unemployed at least three months has a 25% chance of finding work during the third month, while this figure is 11% for a comparable individual who has been receiving unemployment benefit. See Bover, O. , Arellano, M. and Bentolila, S. (2002), «Unemployment Duration, Benefit Duration and the Business Cycle», The Economic Journal, Vol. 112, No. 479.

4. See David, C., Chetty, R. and Weber, A. (2007). The Spike at Benefit Exhaustion: Leaving the Unemployment System or Starting a New Job? IZA Discussion Paper No. 2590.

5. This effect has been underlined, at a theoretical level, by Pissarides (2000). See Pissarides, C. (2000), Equilibrium Unemployment Theory, MIT Press, Cambridge.

6. For a review of this evidence, see Fredriksson, P. and Holmlund, B. (2003), Improving incentives in unemployment insurance: a review of recent research, IFAU Working Paper 2003:5.

7. Reyes, van Ours and Vodopivec (2010) obtain convincing empirical evidence that there are greater incentives to seek employment among unemployed people using an individual savings account. See Reyes, G., van Ours, J. and Vodopivec, M. (2010). «Incentive Effects of Unemployment Insurance Savings Accounts: Evidence from Chile» IZA Discussion Paper No. 4681.