Germany’s economic performance during the Global Crisis in 2009 stands in stark contrast to that of other countries in the OECD. During the last quarter of the year, Germany experienced the largest drop of economic activity when GDP collapsed by 6.9% (Figure 1). At the same time, the unemployment rate increased by less than 1% (Figure 2).
Figure 1 Quarterly one-year GDP growth, OECD and selected countries
Notes: Quarterly GDP growth compared to one year ago for selected OECD-countries. The vertical axis shows the growth rate in percent. The red dashed line shows Germany. Source: OECD.
Figure 2 Unemployment rates, OECD and selected countries
Notes: 2005q1 = 100. The red dashed line shows Germany.
Policymakers and the economic press have largely attributed this ‘German miracle’ to the use of ‘short-time work’.1 This policy instrument added flexibility to the German labour market by allowing firms to temporarily reduce labour input through a reduction in hours worked (the intensive margin), rather than through layoffs (the extensive margin). Absent the short-time work policy, unilateral reductions in hours worked are extremely difficult to achieve, although the recent rise in working time accounts has given firms in certain sectors more flexibility (Burda and Hunt 2011). Workers which are placed on short-time work receive a partial compensation for the earnings loss from the social security system.
The policy, which goes back to at least the Weimar Republic, is aimed at mitigating cyclical shocks on the labour market and should be placed in the context of a labour law that is very protective of employees, especially when compared to that of the US. After a firm successfully applies for short-time work, it is free to temporarily reduce hours worked by up to 100%, either for an entire production site or an identified part of a site. The employment relationship with the affected workers is maintained and actual hours worked are paid as usual. However, during short-time work, a firm continues to pay workers’ social security contributions in full. These ‘remanence costs’ lead to an increase in average wages.
Workers affected by short-time work were eligible for partial compensation of lost earnings. Depending on family status, the German government compensated up to 67% of the net earnings difference due to working time reductions.
Firm behaviour during the recession
The short-time work policy described above was significantly extended during the recent recession. At the heart of this extension stood a drastic relaxation of eligibility criteria for applying firms. At its height, the programme included around 60,000 establishments and 1.5 million workers (about 3.5% of the labour force). The vast majority of firms and workers affected by short-time work were operating in the manufacturing sector.
The recession reached Germany in 2009 and led to a sharp decrease in total hours worked. The aggregate time series for changes in total hours worked in the manufacturing sector is depicted in Figure 3. It shows that the reduction (and subsequent increase) in total hours worked in 2009 was largely due to a decrease in hours per employee. This is the effect of the additional flexibility in hours reductions at the firm level.
Figure 3 Changes in total hours worked (manufacturing)
Source: Cooper et al. (2017)
Several papers have studied the employment effects of the most recent iteration of short-time work in Germany between 2009 and 2011. While Balleer et al. (2016) argue that short-time work has significantly reduced unemployment, Burda and Hunt (2011) place more emphasis on a reticence to hire in the previous expansion. In a related contribution, Dustmann et al. (2014) argue that long-term trends in the decentralisation of the wage-setting process have led to increases in competitiveness that made Germany’s labour market more flexible.
In a new paper, we use data from the universe of German manufacturing firms to assess the effectiveness of short-time work in saving jobs, while at the same time studying the potentially negative effects on output and productivity (Cooper et al. 2017). These negative effects could be coming from a worsening in the allocation of factors to production sites, as expanding firms find it more costly to hire workers, which are prevented from reallocating across firms due to the short-time work policy.
The effect of the policy is clearly visible in the micro data, as Figure 4 shows. It depicts the distribution of changes in annual hours per worker. Compared to the years between 1995-2008, the expansion of the short-time work policy in 2009 led to a large and significant increase in reductions of average hours by more than 10% and 20%. With the phasing out of the short-time work policy, a similar pattern of increases in average hours can be observed in 2010.
Figure 4 Changes in annual hours per worker
Source: Cooper et al. (2017)
Output effects of short-time work
Our paper answers the questions of whether short-time work can save jobs and what the effects on output and productivity will be by simulating a counterfactual scenario to determine the response to the recession without short-time work. In our analysis, the choice of the labour input is placed in a search context with heterogeneous multi-worker firms, allowing for bargaining between a firm and its workers (Cooper et al. 2007, Elsby and Michaels 2013).2 Firms are subject to persistent aggregate and idiosyncratic productivity shocks and choose both employment and hours worked. Variations in hours worked are costly as workers must be compensated. Further, existing policy limits adjustments on the intensive margin.
Adjustments on the extensive margin, through hires and fires, is also costly due to search frictions and hiring costs. The optimal policy of the firm balances costs and benefits given employment policies set by the government. Variations in these policies, say through an expansion of short-time work, lead to a response by firms and thus in the fluidity of labour markets.
We use confidential German plant-level micro data to estimate the parameters of a search model of the German labour market. Our quantitative exercise uses a simulated method of moments (SMM) approach to match the distributions of hours and changes in German firms prior to the policy.
Using these estimates, we simulate the impact of short-time work. Our results confirm the intuition that short-time work mitigates the negative impact of the economic downturn on the labour market and reduces job losses. Firms are induced to respond to adverse demand conditions by adjustments in hours rather than in the number of workers. From our estimated model, in the absence of short-time work, the output loss during the recession would have been 5.3% and unemployment would have risen by about four percentage points.
A second effect is on reallocation. This is more subtle as it operates through the general equilibrium of the model and describes (negative) externalities of firm behaviour. Under the policy, less productive firms are not firing as many workers, and there are fewer workers searching for jobs. Consequently, more productive firms find it more difficult to hire labour. Thus, the rate at which vacancies are filled is reduced by short-time work. In market economies, the efficiency of the allocation of factors across production sites has been shown to play an important role for aggregate productivity (Hsieh and Klenow 2009, Restuccia and Rogerson 2008). By intervening into the reallocation of factors across production sites (i.e. preventing labour from flowing towards the most productive firms), short-time work can generate adverse effects on GDP through this ‘reallocation channel’.
Findings from the analysis show that short-time work had a profound impact on labour market outcomes and shaped the short- and medium-term impact of the crisis. First, short-time work saves jobs in the short term, as firms adjust at the intensive, instead of extensive, margin. Second, firms keep their workers under contract and labour hoarding has a negative impact on the reallocation of labour between more and less productive firms. This reduces productivity in the medium term.
Balleer, A, B Gehrke, W Lechthaler, and C Merkl (2016), “Does short-time work save jobs? A business cycle analysis”, European Economic Review 84: 99–122.
Brenke, K, U Rinne and K F Zimmermann (2011), “Short-time work: The German answer to the Great Recession”, IZA Discussion Paper No 5780.
Burda, M C and J Hunt (2011), “What explains the German labor market miracle in the Great Recession”, Brookings Papers on Economic Activity 42: 273–335.
Burdett, K and R Wright (1989), “Unemployment insurance and short-time compensation: The effects on layoffs, hours per worker, and wages”, Journal of Political Economy 97: 1479–1496.
Cahuc, P and S Carcillo (2011), “Is short-time work a good method to keep unemployment down?” Nordic Economic Policy Review, 133–164.
Cooper, R, J Haltiwanger and J L Willis (2007), “Search frictions: Matching aggregate and establishment observations”, Journal of Monetary Economics 54: 56–78.
Cooper, R, M Meyer and I Schott (2017), “The employment and output effects of short-time work in Germany”, NBER Working Paper No 23688.
Dustmann, C, B Fitzenberger, U Schönberg and A Spitz-Oener (2014), “From sick man of Europe to economic superstar: Germany’s resurgent economy”, Journal of Economic Perspectives 28: 167–188.
Elsby, M W L and R Michaels (2013), “Marginal jobs, heterogeneous firms, and unemployment flows”, American Economic Journal: Macroeconomics 5: 1–48.
Hsieh, C-T and P Klenow (2009), “Misallocation and manufacturing TFP in China and India”, The Quarterly Journal of Economics 124(4): 1403-1448.
Restuccia, D and R Rogerson (2008), “Policy distortions and aggregate productivity with heterogeneous plants”, Review of Economic Dynamics 4: 707-720.
 An overview of the institutional framework in Germany is provided by Burda and Hunt (2011) and Brenke et al. (2011).
 In contrast, Burdett and Wright (1989) model the choice of hours and employees in a static contracting environment. Our analysis is different from Balleer et al. (2016), who study STW in a search setting where firms are not heterogeneous and thus reallocation effects are absent.