When analysing the impact of such furlough schemes, we must differentiate between the short and medium term. Their positive effects in the short term are widely documented, as they allow firms to retain their staff – thus saving redundancy costs and the cost of searching for replacements when the economy is reactivated – as well as offering workers some certainty. In addition, they allow the aid to be tailored to reductions in working hours and have a lower public cost than alternatives such as wage subsidies or unemployment benefits.

In this regard, the Kurzarbeit in Germany was particularly effective during the Great Recession, saving an estimated 400,000 jobs. According to some estimates, the Kurzarbeit reduced Germany’s unemployment rate in 2009 by 1.3 pps, and 4 out of 5 workers covered by this scheme were able to return to their usual job.1 Furthermore, the positive effects on firms can be long-lasting: it has been documented that French companies that used this workforce reduction mechanism after the financial crisis fared better on average2 than similar companies that did not use this type of programme.3

Various empirical studies have used sophisticated statistical techniques to identify the causal effects of these diverging paths and have corroborated the effectiveness of such programmes implemented during the Great Recession (2008-2009). It is estimated that in Germany, 0.35 jobs were saved for each worker enrolled in the programme.4 This may seem a low figure, but in reality its level well below 1 is logical, since some jobs would have been preserved in the absence of these programmes as well. As an example, in France it is estimated that only 0.17 jobs were saved per worker covered by such schemes.5

One interesting element is the fact that, in the 2008 financial crisis, the positive effects were only observed among workers on permanent contracts, since firms are most interested in retaining them. In contrast, no beneficial effects were observed among temporary workers. This situation is very likely to be repeated in the current crisis, and it will have to be given particular consideration in economies with a high degree of labour market duality such as Spain or Italy.

However, experience with these programmes also shows that, while they are particularly effective for cushioning temporary shocks, they lose their effectiveness if the shocks persist in time. This explains why these programmes were especially effective for Germany, where the shock of the Great Recession had a temporary effect limited to 2009. In contrast, the effects on employment ended up being lost in Italy, where the shock was much more persistent as it continued through to 2012 due the sovereign-debt crisis. In fact, when the shock is temporary, firms that use temporary workforce reduction programmes manage to retain between 10% and 15% of the staff that they would have laid off if the shock were permanent.6 Furthermore, as the last chart shows for the case of Italy, during the last financial crisis the positive effect of maintaining employment was only short-lived (during the year in which the programme was administered). Two years later, in contrast, the probability of being employed was exactly the same for a worker who had been included in a furlough scheme and another with similar characteristics who had been fired at the time of the shock.

In addition, extending these programmes for longer than necessary slows down the reallocation of resources towards the most productive firms and can thus reduce aggregate productivity. Once again, Italy is an illustrative case: the empirical evidence shows that a 1-pp increase in the fraction of workers included in furlough schemes reduced employment in all other firms by 0.94% and cut productivity by 2%.

In short, subsidised furlough schemes are proving to be a key tool for cushioning the shock of the coronavirus outbreak on the labour market. However, looking ahead, we will have to tread very carefully in order to adapt such schemes to the new normal. In particular, their duration will have to be carefully fine-tuned. Withdrawing them too early could stifle some companies with a temporary drop in demand which therefore need them to survive until demand recovers. Keeping them in place for too long, however, may do nothing more than prolong the agony for some firms that will no longer be able to stay float, either because the decline in demand is more permanent or because the firms in question were starting from a position of relative weakness. In addition, furlough schemes covering workforce reductions triggered by force majeure will need to be quickly transformed into schemes covering those triggered by economic factors. On top of this, firms and workers will need to be properly incentivised to successfully return to work and to recommence productive activity. To this end, such schemes will need to be made more flexible and differentiated by sector – for example, being maintained in sectors affected by social distancing measures while a vaccine is still being sought –, while reducing working hours, rather than simply cutting staff numbers, will need to be encouraged.