VoxEU Column COVID-19 Labour Markets

Cutting labour taxes brings back the jobs lost to COVID-19

The COVID-19 crisis has disproportionately affected different occupations in the labour market. Workers in contact-intensive and personal-service oriented sectors bear the brunt of the COVID-19 recession, but blue-collar workers suffer heavy job losses as well. This column uses a multi-sector, multi-occupation macroeconomic model to study how different fiscal stimulus measures can boost aggregate demand and help the economy recover faster. It finds that a cut in taxes on labour income outperforms other stimulus plans in promoting job creation for those who lost their jobs in the COVID-19 downturn.

The Covid-19 recession stands out in two ways. First, it is enormous and unfolding at an unprecedented speed. Second, the mix of workers who have been struck is unusual. While a typical recession hits blue-collar workers in construction and manufacturing hardest, the brunt of the Covid-19 recession is borne by workers in sectors with a high intensity of worker-client interaction and workers in personal-service oriented occupations (Adams-Prassl et al. 2020). In the US, the crisis has destroyed ten million jobs in ‘retail trade’ and ‘leisure and hospitality’ industries alone, and one out of three workers in service occupations have lost their job. This is why some call the Covid-19 recession a ‘pink-collar recession’ (e.g. Ribeiro 2020, Wang 2020). However, one should not forget that this recession has, like previous ones, also struck blue-collar workers hard.

The role of fiscal policy

Unlike previous recessions, there is no immediate role for aggregate demand management in the Covid-19 crisis as long as public-health measures depress the economy’s potential output. During the initial phase of the pandemic, the essential role of governments was to support infection fighting and to provide transfers to households and firms to avoid excessive hardship and bankruptcies. Bayer et al. (2020) and Faria-e-Castro (2020) study the effectiveness of the CARES Act in providing disaster relief during the Covid-19 pandemic. When infections are under control and restrictions are slowly relaxed, governments should give more focus to supporting demand and helping the economy recover faster (Bénassy-Quéré et al. 2020, Odendahl and Springford 2020). Governments around the world are indeed discussing how best to tie together large stimulus packages. In the US, for example, House Democrats passed a $3 trillion tax cut and spending bill, which also includes transfers to taxpayers and state and local governments. Republicans resist calls for more spending and instead consider a wide range of tax cuts, mostly for businesses. Proposals include reducing the tax rate on capital gains, allowing companies to deduct investment costs, and suspending payroll taxes. In Germany, a temporary sales tax cut is a centrepiece of the country’s €130 billion stimulus package that parliament is expected to pass at the end of June.

How to design a fiscal recovery plan

When designing a stimulus package, policymakers should not only concentrate on pushing up the total number of jobs. They should also be concerned with the industry mix and – in particular – the occupation mix of employment to help those hit hardest by the crisis. After all, an unemployed salesclerk will benefit little if fiscal policy induces firms to hire software engineers. From an aggregate perspective, stabilising the employment composition can reduce the economic costs of the Covid-19 pandemic by avoiding excessive losses of industry-specific and occupation-specific human capital. For example, Kambourov and Manovskii (2009) show that displaced workers’ future earnings losses are three times larger when they are unable to find a job in their initial occupation.

In a new paper, we investigate how a fiscal recovery plan should be designed to promote job creation for those hit hardest by the Covid-19 crisis (Bredemeier et al. 2020a). To address this question, we use a multi-sector, multi-occupation New Keynesian business-cycle model that can explain heterogeneous employment dynamics by industry and occupation found in US data, as shown in earlier work (Bredemeier et al. 2020b). We distinguish between two large sectors of the economy and three broad occupation groups. Following Kaplan et al. (2020), we differentiate between a ‘social’ sector that comprises industries with high physical proximity between clients and workers, such as retail trade and hospitality, and a ‘distant’ sector where less face-to-face contact is required. The broad occupation groups we consider are white-collar occupations, blue-collar occupations, and service and sales (‘pink-collar’) occupations. Our model generates heterogeneity in employment dynamics as a consequence of differences in sectoral economic activity and changes in the occupation mix within sectors due to differences in the substitutability with capital services across occupations (similar to Autor and Dorn 2013). In particular, labour provided by blue-collar occupations is, on average, more easily substitutable with capital than labour provided by white-collar and pink-collar occupations.

We calibrate the model to the US economy and expose it to a ‘Covid-19 shock’ that generates employment losses by industry and occupation as seen in spring 2020. We use stochastic wedges that combine aspects of both supply and demand disturbances to construct a Covid-19 shock, in line with the evidence by Brinca et al. (2020). We then perform policy experiments where expansionary fiscal policy supports the recovery from early 2021 on. We consider a variety of stimulus measures, both spending-based and tax-based. We quantify the size of the expansionary impulse to achieve a full recovery of aggregate employment by the third quarter of 2021. While this constitutes an ambitious goal, normalising the aggregate effects of the stimulus plans enables us to compare how the different plans would affect employment prospects by industry and occupation. Our two main findings are: 

1. A fiscal stimulus, independent of its specific design, promotes job growth in pink-collar occupations and the social sector. In this sense, fiscal policy is successful in helping create jobs where they were lost during the Covid-19 crisis.

2. Stimulating blue-collar job creation is more challenging. Only a cut in labour income taxes would quicken the recovery for blue-collar workers considerably. That is why a reduction in taxes on labour income outperforms other stimulus programs in stimulating job creation for all those who lost their job in the Covid-19 downturn.

Figures 1 and 2 illustrate our results. Figure 1 shows the model-predicted recovery supported by a spending expansion, and Figure 2 shows the economy’s recovery when the stimulus takes the form of a cut in the tax rate on labour income. For comparison, the dashed lines display how the economy’s recovery is predicted to evolve without policy interventions.

Figure 1 Covid-19 recovery with an equal spending expansion across sectors

Notes: Deviations from steady state. Budget deficit and government spending by sector in percent of steady-state GDP. All other variables in percent of their steady-state values.

Figure 2 Covid-19 recovery with a reduction in the labour income tax rate

Notes: Deviations from steady state. Budget deficit in percent of steady-state GDP. Tax rate in percentage points. All other variables in percent of their steady-state values.

Both measures significantly accelerate the economy’s recovery. The employment gains are most substantial in the hard-hit social sector, thereby stabilising the composition of employment by sector. Both a spending expansion and a cut in taxes push up the number of jobs for white-collar and pink-collar workers substantially. However, blue-collar workers do not benefit much from the increase in government spending: it takes almost four years for blue-collar employment to recover to its pre-Covid level (see Figure 1). A different picture emerges when the government reduces the tax on labour income in Figure 2. Blue-collar employment recovers far more quickly than in the other scenario, achieving a full recovery to pre-crisis levels by mid-2022. Not all tax cuts have this appealing property, though. We also consider reductions in tax rates on capital income and find that they hurt the employment prospects of blue-collar workers.

Differences in capital-labour substitutability across occupation

The reason why blue-collar employment does not benefit much from most stimulus packages lies in its close substitutability with capital services. In the model, the spending stimulus leads to an increase in factor demands, which also boosts the recovery in factor prices. This boost is more pronounced for wages because labour is in less elastic supply than capital services. Thus, firms increase their use of capital services by more than they enlarge their payroll. The more intensive use of capital lowers the marginal productivity of its close substitute, blue-collar labour, weakening the recovery of blue-collar work.

On the contrary, the more intensive use of capital raises the marginal productivity of its close complement, pink-collar labour, reinforcing the recovery of pink-collar employment. The different occupational employment responses also have repercussions for sectoral employment changes. A spending expansion fosters job growth relatively strongly in sectors that employ many pink-collar workers (the social sector) and more weakly in industries employing relatively many blue-collar workers (the distant sector).

To stimulate job creation for blue-collar workers, the government needs to incentivise firms to enlarge their payroll by more than they raise their use of machines. This will happen, for example, if the government reduces the tax on labour income. With this stimulus, the government would not only boost the recovery, but it would also create jobs for all those workers whom the Covid-19 crisis hurt the most.


Adams-Prassl, A, T Boneva, M Golin and C Rauh (2020), “Inequality in the Impact of the Coronavirus Shock: Evidence from Real Time Surveys”, CEPR Discussion Paper 14665, Centre for Economic Policy Research.

Autor, D and D Dorn (2013), “The Growth of Low Skill Service Jobs and the Polarization of the U.S. Labor Market”, American Economic Review 103: 1553-1597.

Bayer, C, B Born, R Luetticke and G Müller (2020), “The coronavirus stimulus package: Quantifying the transfer multiplier”,, 24 April.

Bénassy-Quéré, A, R Marimon, J Pisani-Ferry, L Reichlin, D Schoenmaker and B Weder di Mauro (2020), “COVID-19: Europe needs a catastrophe relief plan”, In R Baldwin and B Weder di Mauro (eds.), Mitigating the COVID Crises: Act Fast and Do Whatever It Takes, a eBook, CEPR Press.

Bredemeier, C, F Juessen and R Winkler (2020a), “Bringing back the jobs lost to Covid-19: The role of fiscal policy”, Covid Economics: Vetted and Real-Time Papers 29: 99-140.

Bredemeier, C, F Juessen and R Winkler (2020b), “Fiscal Policy and Occupational Employment Dynamics”, Journal of Money, Credit and Banking.

Brinca, P, J B Duarte and M Faria-e-Castro (2020), “Decomposing demand and supply shocks during COVID-19”,, 17 June.

Faria-e-Castro, M (2020), “Fiscal policy during a pandemic”, Covid Economics: Vetted and Real-Time Papers 2: 67-101.

Kaplan, G, B Moll and G Violante (2020), “Pandemics According to HANK”, Presentation at LSE, 4 May.

Kambourov, G and I Manovskii (2009), “Occupational Specificity of Human Capital”, International Economic Review 50: 63-115.

Odendahl, C and J Springford (2020), “Bold policies needed to counter the coronavirus recession”, In R Baldwin and B Weder di Mauro (eds.), Mitigating the COVID Crises: Act Fast and Do Whatever It Takes, a eBook, CEPR Press.

Ribeiro, C (2020), “’Pink-collar recession’: how the Covid-19 crisis could set back a generation of women”, The Guardian, 23 May.

Wang, N (2020), “COVID Leads To A Pink Collar Recession”, Forbes, 24 May.

2,520 Reads