Rescuing the labour market in times of COVID-19: Don’t forget new hires!

Christian Merkl, Enzo Weber 07 April 2020



The spread of COVID-19 places economies and labour markets worldwide in a state of emergency. The disruption of supply chains, the slump in world trade and export demand, as well as the loss of working time would be enough for a recession. But the drastic lockdown of domestic economic activity, particularly in the areas of public life, represents a completely new challenge (see Baldwin and Weder di Mauro 2020 for recent policy proposals in the context of COVID-19).

The political reactions are correspondingly far-reaching. Governments around the world are struggling to safeguard jobs and firms. Policymakers are using all possible means to secure what already exists. Short-time work and comprehensive liquidity support for businesses and the self-employed are among the measures. Macroeconomic labour market research has shown that short-time work can be an effective measure to prevent job losses, in particular in severe recessions (Balleer et al. 2016, Gehrke and Hochmuth forthcoming, Weber 2015).

However, even if these measures keep the separation rate in the economy constant (which is unlikely), labour markets will come under substantial stress. To understand this, it is important to think about the labour market in terms of labour market flows. The change in unemployment is approximately the difference between separations into unemployment and hires from unemployment. For instance, in Germany, there were roughly 2 million yearly hires from unemployment in 2019 – a number in the same order of magnitude as the unemployment stock.

The closure of nonessential services in major European economies plus the ensuing uncertainty and demand drop will lead to a substantial reduction of hiring (see e.g. the recent drop of the IAB labour market barometer for Germany). To illustrate the consequences of such a hiring stall, Figure 1 shows how unemployment would evolve when hiring drops by one-third (scenario 1), one-half (scenario 2), or two-thirds (scenario 3) and remains at this lower level for up to 18 months. Note that we keep the separation rate constant at the pre-crisis level of 0.5% per month, i.e. all scenarios isolate the pure hiring effect. 

Figure 1 Months of depressed hiring from unemployment and increase in unemployment rate in Germany

Notes: Scenarios 1, 2, 3 assume a drop by 1/3, 1/2, and 2/3, respectively. We reduce the absolute number of hires, considering job creation as the limiting factor. If we reduced the job-finding rate instead, we would obtain a similar message with a concave instead of a linear adjustment.

For our illustration, we assume that before the crisis, 9% of unemployed workers were hired each month, which is a realistic order of magnitude for Germany. Our illustration shows that the labour market would be under severe stress if the collapse of hiring persisted for a longer time period. In the most unfavourable third scenario, unemployment would roughly double within one and a half years, solely due to the lack of hiring. This clearly shows that it is not sufficient for policymakers to focus their policy measures exclusively on the separation margin.

It is also worth emphasising that the increase of unemployment in Figure 1 shows only part of the problem. In addition to around 2 million hires from unemployment, there were around 2 million yearly hires from non-employment in Germany before 2020 (e.g. when leaving the education system, returning from family care, etc.). Assuming that these hires drop by the same magnitude as hires from unemployment, the decline of employment (in percentage points) would be twice as large. 

Beyond that, there were about 4 million yearly hires as job-to-job transitions before 2020. While many of these workers quit their firm voluntarily, many cases involved were after an involuntary layoff, where individuals directly switch to the next job. A hiring stall would generate unemployment in these cases as well. This means that the unemployment effect in Figure 1 only shows the tip of the iceberg.

It is useful to reconsider the government’s stimulus toolkit in times of COVID-19. An increase of traditional government spending is not particularly suitable to counteract the collapse in hiring in times of COVID-19 because it is untargeted and certain products and services cannot be purchased.

We consider hiring subsidies as the more favourable option. Faia et al. (2013) show in a macroeconomic model simulation that hiring subsidies are a cost-effective stimulus measure for continental European countries. In these economies, collective bargaining and minimum wages are prevalent. Thus, those temporary subsidies for new hires are unlikely to affect the wage structure immediately. Hiring subsidies directly reduce firms’ costs and thereby stimulate hiring. Cahuc et al. (2018, 2019) provide empirical evidence for France that a temporary non-anticipated hiring subsidy has substantial positive employment effects. 

Two additional arguments come into play in times of COVID-19. First, when the hiring rate is depressed, deadweight effects (subsidies for hiring that would have happened in any case) are smaller than in normal times. Second, the hiring subsidy will automatically be targeted towards sectors that remain open, or reopen gradually, and particularly towards firms that have been on an employment growth path. Overall, a hiring subsidy will counteract output drops and uncertainty effects; it will increase hiring and stimulate GDP. It can thereby reduce the risk of unemployment hysteresis effects.

For Germany, Weber (2020a,b) proposes a ‘rescue fund for new hires’. Amongst other measures, the proposal includes direct hiring subsidies. As the most comprehensive and simple way to realise the measures, the proposal suggests that no social security contributions would have to be paid for new employment relationships for a certain time period. The costs for the social security system should be reimbursed with tax money. The subsidy would create strong incentives for new hires and, in addition, would support start-ups, all of whose staff are new hires. Unlike income taxes, social security contributions are usually not progressive such that the rescue fund also has a large impact on low-wage jobs. In other words, the subsidy would be universal and not be confined to specific jobs or skill levels. This is an important feature, since the COVID-19 crisis has a broad effect on various parts of the economy.

In order to avoid abuse, cases should be excluded in which the person concerned has only recently ended an employment relationship with the same employer. In addition, employment relationships would have to be sustained for a certain time period in order to obtain the subsidy. This would also rule out cases in which employees in short-term jobs would be swapped to save social security contributions. 

Certain false incentives to dismiss workers first in order to benefit from subsidised hiring later on may remain, but these would be limited by the timing, as the proposed hiring subsidy would follow the immediate rescue packages for existing jobs and firms already in place. Moreover, labour hoarding is often highly subsidised, e.g. by short-time work, and a firing and hiring strategy would be very costly due to firing costs, hiring, and training costs.

The hiring subsidy would also be granted in many cases in which hiring would have taken place anyway. Although preventing deadweight effects is impossible, in the current crisis, they can be expected to be smaller than in normal times. In an exceptional situation, aiming at the broadest possible impact is key. 

Our proposed measure should be considered as an expansionary fiscal policy that targets the recruitment channel. It would dampen the slump in hiring, improve economic expectations, and speed up labour-market recovery. With immediate rescue packages for existing jobs and firms already in place, a support programme for new hires would be an important next step to prevent a collapse of the labour market.


Baldwin, R, and B Weder di Mauro (2020), Mitigating the COVID economic crisis: Act fast and do whatever it takes, A eBook, CEPR Press.

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.

Cahuc, P, S Carcillo and T Le Barbanchon (2018), “The effectiveness of hiring credits: France evidence from the global crisis”,, 9 January.

Cahuc, P, S Carcillo and T Le Barbanchon (2018), “The effectiveness of hiring credits”, Review of Economic Studies 86: 593–626.

Faia, E, W Lechthaler and C Merkl (2013), “Fiscal stimulus and labor market policies in Europe", Journal of Economic Dynamics and Control 37: 483–99.

Gehrke, B, and B Hochmuth (forthcoming), “Counteracting unemployment in crises: Non-linear effects of short-time policy work”, Scandinavian Journal of Economics.

Weber, E (2015), “The labour market in Germany: Reforms, recession and robustness”, De Economist 163: 461–72.

Weber, E (2020a), “Corona-Schock am Arbeitsmarkt: Wir brauchen auch einen Rettungsschirm für Neueinstellungen”, Der Spiegel, 24 March.

Weber, E (2020b), “Ein Rettungsschirm für Neueinstellungen”, Makronom, 30 March.



Topics:  Covid-19 Labour markets

Tags:  short-time work, Germany, unemployment, COVID-19, labour market policies

Professor of Macroeconomics, University of Erlangen-Nuremberg

Head of Forecasts and Macroeconomic Analyses, Institute for Employment Research; Professor of Economics, University of Regensburg.


CEPR Policy Research