Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Germany and the United States in coronavirus distress: internal versus external labour market flexibility

Germany and the United States in coronavirus distress: internal versus external labour market... Germany and the United States pursued different economic strategies to minimise the impact of the Coronavirus Crisis on the labour market. Germany focused on safeguarding existing jobs through the use of internal flexibility measures, especially short‑time work (STW ). The United States relied on a mix of external flexibility and income pro ‑ tection. On this basis, we use macroeconomic time series to examine the German strategy of securing employment through internal flexibility by contrasting it with the chosen strategy in the United States. In Germany, temporary cyclical reductions in working hours are mainly driven via STW. US unemployment rose at an unprecedented rate, but unlike in previous recessions, it was mostly driven by temporary layoffs. However, a closer look at the blind spots of the chosen strategies in both countries showed that despite the different approaches, people in weaker labour market positions were less well protected by the chosen strategies. Keywords: Working‑time reduction, Safeguarding employment, Unemployment, Internal flexibility, External flexibility, Short ‑time work, Temporary layoffs, Great recession, Coronavirus recession, Covid‑19 pandemic JEL: E24, E32, J08, J20 1 Introduction hoarding through STW programmes, the United States During both the Great Recession and the Coronavirus decided to insure worker’s incomes with instruments such Recession, unemployment in Germany increased only mod- as cash transfers and temporary increases in unemployment erately, especially given the severity of these recessions. benefits instead of protecting employment. However, in both recessions there were instead significant Giupponi et  al. (2022) discuss extensively upsides temporary reductions in working time by means of internal and downsides of both strategies. We contribute to this flexibility, i.e., the internal adjustment of the labour input debate by providing a detailed descriptive analysis of the used in the production process along the intensive margin German strategy of safeguarding employment via inter- especially through the use of short-time work (STW). While nal flexibility by contrasting it with the strategy chosen most European countries have also relied on STW to tackle in the United States and the labour market experience in the crisis, in the United States external flexibility was domi - the two countries in 2020. nant, i.e., the adjustment of labour input via the external For Germany, we show that despite the dramatic labour market, and there was a sharp temporary increase in decrease in real gross domestic product (GDP), unem- unemployment of historic proportions. Instead of labour ployment only increased moderately. While during the Great Recession all working-time instruments con- tributed to the reduction in working time, STW now *Correspondence: alexander‑herzog‑stein@boeckler.de accounts for almost all of the working-time reduction Macroeconomic Policy Institute (IMK), Georg‑Glock‑Str. 18, 40474 Düsseldorf, Germany For more details on the difference between internal and external (numerical) Full list of author information is available at the end of the article flexibility see for example Herzog-Stein and Zapf (2014). © The Author(s) 2022. Open Access. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. 11 Page 2 of 22 A. Herzog‑Stein et al. Fig. 1 Real gross domestic product in Germany and the United States (2005–2021). Quarterly change in real GDP; seasonally adjusted. Sources: Federal Statistical Office (Destatis); Bureau of Economic Analysis; authors’ calculations as the government focused on this instrument to main- of the chosen policies in the two countries. Section  4.1 tain employment. In contrast, the United States expe- concludes. rienced a comparable decrease in real  GDP, but also a temporary increase in unemployment on an unprec- 2 The German and US labour market edented scale. However, the nature of unemployment during the Covid‑19 pandemic changed as well. In contrast to the Great Recession, The outbreak of the global Covid-19 pandemic and its temporary layoffs during the Coronavirus Recession economic impact on the world economy caused a major played a dominant role in the United States and allowed economic crisis. Both the German and the US economy for a fast recovery of the unemployment rate. While were severely affected by the pandemic and experi - both countries approached the crisis differently, enced economic contractions of similar magnitude. The remarkably their weakness is the same. The methods to German economy started to be severely affected by the secure employment and income were less pronounced pandemic at the end of the first quarter 2020 and went for individuals in weaker positions on the labour mar- into a partial lockdown from mid-March to May 2020. ket (atypical employment and low wage earners). The result was an economic slump of historic propor - The remainder is structured as follows: In Section  2 tions. In the second quarter  2020, real GDP contracted we provide a concise summary of how the German by 9.9%, after it fell already by 0.7% in the first quarter labour market was affected by the Coronavirus Cri - 2020 (Fig.  1). In total in the first two quarters 2020 real sis relative to the experience in the United States and GDP fell by 10.5% seasonally adjusted. The United States describe the differing labour-market related policy declared a public health emergency at the End of Janu- responses in the two countries. Section  3 presents a ary 2020 and in mid-March 2020 a national emergency comparative business-cycle analysis of the Coronavi- in response to the Covid-19 pandemic. From Spring rus Recession and the Great Recession in Germany as 2020 onwards there were also widespread business clo- well as in the United States and has a closer look at the sures and social distancing practices (Houseman 2022). country specific margins of labour market adjustments. Real  GDP fell seasonally adjusted by 8.9% in the second Then, in Section  4  we investigate potential blind spots quarter 2020 after a 1.3% decrease in the first quarter. Germany and the United States in coronavirus distress: internal versus external labour market… Page 3 of 22 11 Fig. 2 Unemployment rates in Germany and the United States (2005 to 2021). The definition used in the Labour Force Survey follows the definition of the Internal Labour Organization. Number of all unemployed people aged 15–74 as percentage of the labour force. Sources: Eurostat; own presentation The Coronavirus Crisis had a marked impact on labour In comparison to the Great Recession, the increase market performance in Germany and the United States. in the US unemployment rate from its minimum was Unemployment increased markedly in both countries around twice as large in the Coronavirus Recession, but (Fig.  2). However, the magnitude of the rise in unem- the following decline was much quicker this time. In Ger- ployment in the two countries was different. many, unemployment rose faster this time than in the While in the United States the unemployment rate Great Recession, when unemployment started to rise in increased dramatically from 3.5 to 14.7% from February November 2008 and peaked in July 2009 (see Fig. 2). With to April 2020, in Germany the rise in unemployment was a total increase of 0.6 percentage points the rise in unem- much less pronounced but went on for longer. From Feb- ployment was now slightly less than at that time (+ 0.9 ruary to August 2020 the German unemployment rate percentage points). Compared to the United States and increased from 3.5 to 4.1% and started only to decline against the backdrop of the massive decline in economic again in December 2020. In contrast, in the Unites States activity the rise in unemployment was in both downturns unemployment declined relatively fast and continuously relatively moderate in Germany. after its peak in April 2020. In addition, the US civilian The differences in the development of unemployment labour force declined by 5.0% from February to April in the two countries in the Coronavirus Recession are 2020, while in Germany the labour  force decrease from quite remarkable. But the policy responses in Germany February to May 2020 was much smaller at 1.5%. and the United States regarding the labour market as a result of the economic impact of the pandemic were also quite different. Table  1 contrasts the discretionary pol- icy changes in Germany with those in the United States in 2020 sorted into four categories: legislative changes To enable a consistent comparison between the two countries, we use the regarding the respective STW scheme, the provision of unemployment rate as defined by the International Labour Office (ILO). For Germany, the monthly development of this unemployment rate during the business support, the expansion of unemployment ben- Covid-19 pandemic differs with respect to the timing and the magnitude of efits, and the provision of income support for house - the increase from the unemployment rate of the Federal Employment Agency, holds. STW and business support reflect measures that which is defined differently. 11 Page 4 of 22 A. Herzog‑Stein et al. Table 1 Discretionary policy changes during the Coronavirus Crisis in the United States and Germany (dating corresponds to date of passage) Germany United States 2020Q1 Short‑time work Short‑time work • Simplified eligibility criteria • Federal financial support for states’ STC schemes • 100% reimbursement of social insurance contributions for hours affected by STW • Expansion of additional income opportunities during STW Business support Business support • Soforthilfen: grants for small businesses & self‑ employed, administrated • Paycheck Protection Program (PPP): forgivable loans to SME by Länder • Employee Retention Tax Credit: payroll tax credit for employers • Loans and credit guarantees • Tax credits for obliged paid leave by employees • Disaster Loans Program: Federal Funding of loans from Small Businesses Administration Unemployment benefits Unemployment benefits • Simplified eligibility criteria for basic income support (ALG II) (e.g., fur ‑ • Federal Pandemic Unemployment Compensation (FPUC): additional benefit ther inclusion of self‑ employed, suspension of means testing) of $600 per week until end of July • Pandemic Emergency Unemployment Compensation (PEUC): extension of max. eligibility period • Pandemic Unemployment Assistance (PUA): simplified eligibility criteria (e.g., inclusion of self‑ employed) Income support (households) Income support (households) • Simplified eligibility criteria for child supplements • Child Tax Credits • Compensation for earnings losses due to child care • Economic Impact Payments: Stimulus checks for households per eligible adult ($1200) and child ($500) • Obligation to SME to provide paid sick leave, paid family leave, paid medi‑ cal leave 2020Q2 Short‑time work Business support • Extension of maximum eligibility period • Improved conditions regarding loans and forgiveness of loans (June) • Temporary increase in replacement rates until end of 2020 • Amendment of PPP: Provision of additional $320 bn. (April) Unemployment benefits • Expansion of max. eligibility period for unemployment insurance (ALG I) by three months 2020Q3 Business support Unemployment benefits • Grants for SME & self‑ employed (Überbrückungshilfe) •Extension and modification of expiring FPUC: $300 per week and condi‑ • Hiring credit for apprentices for SME of €2000 tioning eligibility on receiving at least $100 from other state unemploy‑ ment benefits Unemployment benefits Income support (households) •Extension of simplified eligibility criteria for ALG II •Provision of assistance to renters and homeowners Income support (households) • Stimulus Checks for families: €300 per child (Kinderbonus) • Tax Credits for single parents • Expansion of max. eligibility period for elderly/child care compensation • Temporary VAT reduction 2020Q4 Short‑time work Business support • Extension of the key measures until end of 2021 •Modification & extension of different programs (i.a., the Employee Retention Tax Credit, PPP, etc.) Business support Unemployment benefits • Grants for foregone revenues (November-/Dezemberhilfen) • FPUC: $300 per week until March 2021 • Grants for SME (Überbrückungshilfe II) • PEUC applies until March 2021, expansion of max. eligibility period to 24 weeks • PUA applies until March 2021; expansion of max. eligibility period to 50 weeks Income support (households) •Economic Impact Payments for adults ($600) and children ($600) Sources: Federal Ministry of Labour and Social Affairs, Federal Ministry of Finance, Steffen (2021) for Germany; U.S. Government, Chetty et al. (2020), Houseman (2022) for United States Germany and the United States in coronavirus distress: internal versus external labour market… Page 5 of 22 11 incentivise the use of internal flexibility to safeguarding In Germany all programs providing business support employment. Expansions of unemployment benefits and in 2020 account for €54.73 billion  or 1.6% of nominal income support for households aim at insuring incomes GDP in 2020 (Federal Ministry of Finance 2021). The given that establishments adjust labour input along the main programs at the federal level are the Soforthilfen in extensive margin. March, the Überbrückungshilfen  I–III from the Stabiliz- In short, while Germany put a focus on subsidising the ing Package in July, and the November-/Dezemberhilfen. use of internal flexibility, especially by providing more These programs contained grants and forgivable loans. generous STW allowance, the United States put a strong Apart from the differences in incentivising the use of focus on insuring incomes by providing generous unem- measures of internal flexibility between Germany and ployment benefits and allow for adjustments along the the United States there are also differences in the extent extensive margin. of insuring workers’ incomes in case establishments Policy measures to expand the use of STW had only adjust their labour input along the extensive margin. The a relevant impact in Germany. In line with the empiri- CARES Act introduced additional federal unemployment cal evidence on the effectiveness of the rule-based and benefits (FPUC) of $600 per week until July 2020. In fur - the discretionary component of STW in safeguard- ther programs the maximum eligibility period and eligi- ing employment in the Great Recession (Balleer et  al. bility of further worker groups has been introduced. In 2019; Gehrke and Hochmuth 2021), discretionary policy August the FPUC was replaced by $300 per week, and in changes made the use of STW in the pandemic more the Consolidated Appropriations Act the maximum eligi- attractive. The similarity of the changes in both crises is bility period was further expanded. striking (Herzog-Stein et  al. 2021), albeit the extensions In contrast to the US, the discretionary changes in were much faster and more generous this time. Crucial unemployment benefits in Germany had only minor rele - for the increased take up of STW were the extension of vance in insuring incomes of the unemployed. In the first the eligibility period of STW and the simplified eligibility Social Security Package in March  2020 eligibility criteria criteria with respect to the scope of STW in March 2020. for basic income support (ALG II) were simplified. In the Moreover, immediately a full reimbursement of social subsequent second Social Security Package from May security contributions for hours affected by STW was 2020 the maximum eligibility period for unemployment introduced to reduce residual costs of companies when benefits was expanded by three months. using STW. Thus, strong incentives for companies to use Besides insuring incomes with unemployment ben- STW were created. efits, stimulus payments are of particular importance to In contrast, in the United States, STW measures stabilize incomes. In Germany, the Stabilizing package mainly consist of financial support for state-level STW contains a stimulus check for families by providing €300 schemes by the federal level. Moreover, only 26 states had per child and tax credits for single parents. Still, they are STW schemes at the beginning of the crisis implemented by far not as expansive as the Economic Impact Payments (Houseman 2022). in the US. The CARES Act entailed so-called Economic Instead of subsidizing working-time reductions by Impact Payments of up to $1200 per adult for eligible expanding the STW schemes, the US government aimed individuals (earning less than $75,000) and $500 per at subsidised labour hoarding by providing business sup- qualifying child under age 17. Additionally, a child tax port especially with the CARES Act in March 2020. These credit was implemented. measures aimed at incentivising businesses to retain their employees, too. Major elements of the CARES Act were 3 Germany and the United States: internal the Paycheck Protection Program (PPP), which mainly versus external flexibility in the Covid‑19 entailed forgivable loans to SME, and the Employee pandemic Retention Tax Credits. However, as Autor et  al. (2022) To gain a better understanding of the impact of the differ - show, 66 to 77% of the issued loans in 2020 do not seem ent policy strategies in Germany and the United States on to actually have been used to retain their employees but the labour market, we conduct a business-cycle analysis were kept by business owners and shareholders. Both comparing, on the one hand, the Coronavirus Recession measures, the PPP and the Employee Retention Tax with the Great Recession in both countries and contrast- Credits were amended in subsequent laws later in 2020. ing the German experience with that of the United States Overall, these two measures of business support from the on the other hand. Then, we look at the case of Germany CARES Act account for $861 billion  or 4.1% of nominal in the Coronavirus Recession with particular interest in GDP in 2020 (CRFB 2021). the relative importance of working-time instruments (overtime, regular working time, working time accounts (WTA), and STW) in safeguarding employment, before This view is also documented in Giupponi et al. (2022). 11 Page 6 of 22 A. Herzog‑Stein et al. Table 2 Dating the Great Recession and the Coronavirus Recession in Germany and the United States Contractions Germany United States Peak quarter Trough quarter Peak quarter Trough quarter Great Recession 2008 Q1 2009 Q2 2007 Q4 2009 Q2 Coronavirus Recession 2019 Q4 Not yet determined 2019 Q4 2020 Q2 Sources: German Council of Economic Experts (GCEE); NBER’s Business Cycle Dating Committee examining the role of external flexibility in the United on cyclical variations in economic activity in the follow- States via temporary and permanent lay-offs during the ing business-cycle analysis. Therefore, we extract the Covid-19 pandemic. cyclical and trend component using the Hodrick-Prescott Filter (HP-Filter). As is common practice for quarterly 3.1 G ermany and the United States: a business cycle data, we use a HP-Filter with a smoothing parameter analysis equal to 1600 to detrend the quarterly time series from For a business-cycle analysis of the cyclical variations in 1991 to 2021. It is well known that the HP-Filter, like economic activity, employment, productivity, and work- other filter methods, suffers from an end-point problem. ing hours in these two countries, we must first determine Since our main focus is on the slump until 2020q2 and the peak and trough of the Great Recession and the latest we use six additional data points, the impact of this end- recession in Germany and the United States. We follow point problem is still there but of a smaller importance the business cycle dating of the NBER’s Business Cycle for the analysis of the recession periods. However, we are Dating Committee for the United States and of the Ger- careful in interpreting results after 2020q2 and closer to man Council of Economic Experts (GCEE) for Germany. the end of the data set. For completeness and clarity, we In Germany the economic downturn of the Great present results up to 2021q4, the end of our dataset. Recession started after the first quarter of 2008 (peak) For Germany and the United States, Fig.  3 examines and ended in the second quarter 2009 (trough). Accord- the economic dynamics of the cyclical components of ing to the NBER, in the United States it started already real  GDP, employment, productivity and working time, after the fourth quarter 2007 (peak) and also ended in the i.e., average hours worked per employee, during the second quarter 2009 (Table 2). As for the latest downturn, Great Recession (Germany: Panel A, and United States: the Coronavirus Recession, in both countries economic Panel C) and the Coronavirus Recession (Germany: Panel activity peaked in the fourth quarter 2019. The NBER B, and United States: Panel D). All figures are normalised dated the trough of economic activity to the second quar- to the respective beginning of the two economic down- ter 2020. For Germany, the GCEE has not yet determined turns, i.e., for the Great Recession 2008q1 for Germany the trough of economic activity. However, the develop- and 2007q4 for the United States, and for the Coronavi- ment of economic activity in Germany in 2020 and 2021, rus Recession 2019q4 for both economies. especially GDP growth, also points towards the second Due to the economic shock caused by the Covid-19 quarter 2020 as time of the economic trough. Therefore, pandemic, the Coronavirus Recession was much more in the further analysis we assume that 2020q2 is also the severe. From 2019q4 to 2020q2, cyclical real  GDP con- trough of economic activity in Germany, but present data tracted by 12.3% in Germany and by 11.4% in the United for both countries until the end of 2021. States as a direct consequence of the Covid-19 pan- Given the determination of the Great Recession and demic. In the Great Recession, the corresponding cyclical the Coronavirus Recession in Germany and the United decline in output from peak to trough was 8.1 and 5.1%, States by the GCEE and the NBER, respectively, we focus respectively. As in the Great Recession, most of the economic shock Throughout this business -cycle analysis whenever we use the logarithmic in Germany was absorbed by internal flexibility in the transformation of a time series x, growth rates are approximated by differ - labour market via a temporary working-time reduction ences in log points, i.e., the growth rate of x in per cent is g ≈ lnx × 100. Growth of 1 log point is equal to a growth rate of approximately 1 per cent. and labour hoarding in the form of a procyclical decline The exact relationship is For simplicity we always g = 100 ∗ (exp(�lnx) − 1). in labour productivity. However, this time the relative speak of percentage changes. contribution of internal flexibility was even larger than Both institutions determine the peaks and troughs of economic activity in the Great Recession. From peak to trough, the cycli- separately on a monthly and a quarterly basis. In general, the peak or trough quarter contain the peak or trough month. However, the determination of cal reduction in the average number of hours worked the peak of economic activity before the Coronavirus Recession is an excep- per employee was twice as high as in the Great Reces- tion. Both, NBER and GCEE determine February 2020 as the latest peak sion (− 8.8 vs. − 3.4%). In Germany, productivity reacted month but the fourth quarter 2019 as the peak quarter. Germany and the United States in coronavirus distress: internal versus external labour market… Page 7 of 22 11 Fig. 3 Great Recession vs. Coronavirus Recession in Germany and the United States. Log deviations from peak quarter (Germany: 2008q1 respectively 2019q4; United States: 2007q4 respectively 2019q4) measured in log points. Output (= real GDP), employment, working time in hours worked per quarter per employee, and productivity (= labour productivity per hours worked) are seasonally and/or calendar adjusted. Sources: Federal Statistical Office (Destatis); Bureau of Economic Analysis; U.S. Bureau of Labor Statistics (BLS); own calculations 11 Page 8 of 22 A. Herzog‑Stein et al. much stronger in the Great Recession than in the Coro- to changes in the economic environment, such as over- navirus Recession (− 5.6 vs. − 2.1%). Even though speed time, working-time accounts, temporary changes in reg- and intensity of job losses were more pronounced in the ular working time and STW. Figure 4 therefore shows the Coronavirus Recession, in both economic recessions development of cyclical working time per employee and cyclical employment continued to decline, even after its components regular working time, paid and unpaid the trough of the business cycle. Overall, employment overtime, STW, as well as WTA, again detrended with declined cyclically by 0.6% from 2008q1 to 2009q2; by the HP-filter (  = 1600) if the component has a trend. 2010q1 it had fallen by a further 0.7%. Thereafter, cycli - Over the period from the beginning of 2005 to the end of cal employment started to recover. In the Coronavirus 2020, working time and all its components follow a clear Recession, cyclical employment declined by 1.4% from cyclical pattern. However, while all these components 2019q4 to 2020q2 and a further 0.4% by 2021q1. It then contributed to the safeguarding of employment during started to recover over the remaining quarters of 2021. the financial crisis (Herzog-Stein et  al. 2018), this is no In contrast to the economic developments observed in longer the case in the Coronavirus Recession. Germany, in the United States external flexibility bore the brunt of adjustment in response to the economic shock 3.2.1 Shor t‑time work (STW) as a consequence of the Covid-19 pandemic (Panel D). In terms of safeguarding jobs, STW has two important From peak to trough, cyclical employment decreased by dimensions: the number of workers in STW and the 13%. However, in contrast to the cyclical development in intensity of STW, i.e., the number of reduced working Germany, employment then started to recover. This also hours per short-time worker due to STW. Comparing contrasts the cyclical behaviour of employment in the the development of STW in both recessions, two aspects Great Recession when it continued to decline beyond the stand out particularly. First, policy makers reacted fast trough quarter. and made the use of STW more attractive for establish- Average hours worked per employee decreased by 1.0% ments immediately after the outbreak of the Covid-19 in the Coronavirus Recession, while labour productivity pandemic at the end of the first quarter 2020. This had cyclically increased by 2.1% from peak to trough. This the effect of introducing STW on a uniquely large scale in also contrasts with developments in the United States both dimensions of STW. In April 2020, the month with during the Great Recession, when internal flexibility the highest incidence of STW in the Coronavirus Reces- from cyclical reductions in working hours per worker sion, almost 6 million, or 17.9% of all employees subject and changes in labour productivity together accounted to social security contributions were in STW. The average for about a quarter (− 1.2%) of the labour market adjust- loss of working time for a short-time worker was nearly ment relative to the cyclical decline in real  GDP (Panel 50%. In employment equivalents this corresponded to C). In the Great Recession, from peak to trough working 9.1% of all employees subject to social security contribu- time per employee decreased cyclically by 1.7%, a larger tions (Fig. 5). decline than during the Coronavirus Recession. But in Although the number of employees in STW declined the latest contraction the speed of the working-time steadily after April 2020, there were still more employees reduction was faster than in the Great Recession. How- in STW in October 2020 than at the peak of the Great ever, the major difference in the cyclical labour-market Recession. As a result of the second wave of the Covid- responses between the two recessions lies in the develop- 19 pandemic, the number of workers in STW rose again ment of cyclical labour productivity in the United States. from November 2020 and peaked in February 2021 Cyclical labour productivity behaved slightly pro- to acy- before declining again. clical in the Great Recession and anticyclical in the Coro- Consequently, in the Coronavirus Recession there navirus Recession. was a rapid cyclical reduction in average working time Overall, this section has shown that the German and per worker of 2.4  h already in 2020q1 alone (relative to US labour market reacted quite differently during the 2019q4). This is comparable in its magnitude to the cycli - last two recessions. In Germany, internal flexibility domi - cal working-time reduction induced by the use of STW nated labour market adjustment, while in the United from peak to trough in the whole Great Recession of States it was external flexibility. This finding fits with the 3.3 h per worker—of which 3.1 h were reduced in the first descriptions of the policy responses outlined above. two quarters of 2009 relative to the last quarter in 2008. Second, while the immediate response in STW was 3.2 G ermany: internal flexibility already comparable to the Great Recession, at the trough In Germany, several instruments of internal flexibility are of the Coronavirus Recession in the second quarter 2020, available at the establishment level to temporarily adjust STW reduced the average working time per worker by the number of hours worked per employee in response 18.4 h compared to the peak quarter 2019q4, more than Germany and the United States in coronavirus distress: internal versus external labour market… Page 9 of 22 11 Fig. 4 Components of cyclical changes in working hours per employee per quarter (2005–2021). The term ‘cyclical’ refers to the difference of actual and trend changes for each working‑time instrument (if the series shows a trend). STW and WTA show no trend. The trend is constructed applying the Hodrick‑Prescott filter with  = 1600 . All components are measured in working hours per employee per quarter. Sources: Institute for Employment Research (IAB) working time calculations; own calculations Fig. 5 Short‑time work and employment equivalents (2008–2020). Proportion of short ‑time workers (realised numbers or employment equivalents) in total employment subject to social security contributions. Sources: Federal Employment Agency; own presentation 11 Page 10 of 22 A. Herzog‑Stein et al. Fig. 6 Share of recipients of short‑time allowance, average working time reduction, and employment change by economic sector. B: Mining and quarrying; C: Manufacturing; D: Electricity, gas, steam and air conditioning supply; E: Water supply; sewerage, waste management and remediation activities; F: Construction; G: Wholesale and retail trade; repair of motor vehicles and motorcycles; H: Transportation and storage; I: Accommodation and food service activities; J: Information and communication; K: Financial and insurance activities; L: Real estate activities; M: Professional, scientific and technical activities; N: Administrative and support service activities; O: Public administration and defence; compulsory social security; P: Education; Q: Human health and social work activities; R: Arts, entertainment and recreation; S: Other service activities. Short‑time work refers to the share of recipients of short‑time allowance by economic sector in May 2009 and April 2020 respectively (columns). The intensity of STW refers to the average reduction in working time of a short‑time worker (in %) due to STW (dots) and is calculated by dividing the employment equivalent by short‑time workers. Change in employment (diamonds) refers to the sum of employment subject to social security contributions (seasonally adjusted) and exclusively marginally paid employees by economic sector from March to April 2020. Data on marginally paid employees by sectors are only available since 2020. Hence no seasonally adjusted data are available. Given that employment is not provided in each economic sector, employment changes of the sectors B, D, E, L, M, O, U, R, S, T are approximated by the corresponding average employment changes by the sums of B + D + E, L + M, O + U, R + S + T. Sources: Federal Employment Agency; own calculations five times the working-time reduction due to the use April 2020 was nearly twice as high as in May 2009. The of STW in the Great Recession. On average, STW is average intensity of STW use was particularly high in the accounting for around 89% of the total reduction in hours services sector, exceeding 70% in sections ‘Accommoda- worked per worker from 2019q4 to 2020q2. tion and food service activities’ (I), ‘Arts, entertainment In the two recessions, employees subject to social and recreation’ (R), and ‘Other service activities’ (S). In security contributions were affected differently by STW the past, STW intensity of 100% was not common. In the in the individual economic sections (Fig.  6). A compari- Coronavirus Recession it was used only modestly, despite son between May 2009 and April 2020, the months with the severity of the crisis. According to Kruppe and Osi- the highest incidence of short-time work in both down- ander (2020) using information on the individual STW turns, shows that this time not only was the number of intensity from a survey in May 2020, 24.1% of all STW- short-time workers significantly higher, but in the econ - workers reported a loss in hours of 100%, but still more omy as a whole STW was used more heavily (columns in than half a loss in hours of only up to 50%. Fig. 6). While more than 80% of short-time workers were employed in manufacturing during the Great Recession, With respect to employment subject to social security contributions the it was only about 31% during the Coronavirus Recession. employment structure of the German economy has changed only moderately since the Great Recession. The employment share in manufacturing (section In the Coronavirus Recession, STW is also used more C) has decreased from 23.3 to 20.8%. In turn, the employment share in the intensively across all economic sections (dots in Fig.  6). services sector (sections G-N) has risen. In the total economy, the intensity of the use of STW in Germany and the United States in coronavirus distress: internal versus external labour market… Page 11 of 22 11 In contrast to the importance of internal flexibility and time from 2019q4 to 2020q2 on a quarterly basis (− 0.3 h especially the use of STW, external flexibility—unlike in vs −  0.2  h per quarter) is similar to that in the Great the United States (see Sect.  3.2.1)—hardly played a role Recession (− 0.2 h and − 0.2 h per quarter), but together in Germany between March and April 2020. The overall accounting only for less than 5% of the total working- change in employment, measured by the sum of employ- time reduction per worker during that time period, in ees subject to social security contributions and workers contrast to nearly 20% in the Great Recession. only marginal employed, was only about − 1% on average (diamonds in Fig.  6). Only in section H (Transportation & Storage) there is a substantial drop in employment of 3.2.4 R egular working time − 7.5%. Unlike in the Great Recession, there is not really a cycli- cal response in regular working time to reduce working hours in the Coronavirus Recession. The cyclical com - 3.2.2 Working‑time accounts (WTA) ponent of regular working time per worker even slightly Together with STW, they were the most important increased average working hours per worker by on aver- instrument of internal flexibility during the Great Reces - age 0.3  h from 2019q4 to 2020q2. Overall, this observa- sion. Like STW, the use of WTA at that time reduced tion might be explained by the dominance of STW, which the average working time per employee by 3.3  h in total made further adjustments to working time unnecessary. or 0.7  h per quarter from peak to trough. In the Coro- navirus Recession, the contribution of WTA to the tem- 3.2.5 Summary porary reduction in average hours worked per worker In conclusion, although external flexibility again was of is this time much smaller than in the Great Recession. minor importance and instruments of internal flexibility From peak to trough, WTA contributed 1.7 h, or on aver- played a crucial role in the safeguarding of employment age 0.8  h per quarter, to the reduction in average hours in both the Great Recession and the Coronavirus Reces- worked per worker in the latest downturn. sion in Germany, a closer look at various working-time At first glance, this is unexpected, as WTA became components shows marked differences between the two more common over time and 56% of all employees had recessions. While in the Great Recession several instru- WTA in 2016 (Ellguth et al. 2018). However, one possible ments contributed markedly to the temporary decline in explanation could be the respective economic dynamics hours worked per worker, in the Coronavirus Recession in the boom periods before the two recessions. STW is the instrument that has contributed by far the In the upswing before the Great Recession, WTA were most to the reduction in working hours (Fig. 7). filled, providing firms with a considerable working-time- In the Great Recession, STW and WTA contributed account buffer for the following downturn. In contrast, in equally to the cyclical reduction in working time from the long boom period before the Coronavirus Recession, peak to trough (− 3.3 h each). Paid and unpaid overtime working time was closer to its long run trend with smaller and a temporary reduction in regular working hours cyclical variations. As a result, opportunities to increase both reduced cyclical working time by an additional two the balances in the WTA were more limited than in the hours. In contrast, while most instruments responded as boom period before the Great Recession. Therefore, the expected in the latest downturn, in absolute and in rela- working-time reductions due to WTA account only for tive terms STW was by far the main driver to safeguard 8% of total working-time reduction in the latest recession employment in the Coronavirus Recession (−  18.4  h). from 2019q4 to 2020q2. WTA was again the second most important instru- ment of internal flexibility used. However, its quantita - 3.2.3 Overtime tive importance was smaller, reducing average working In general, paid and unpaid cyclical overtime vary hours per employee by 1.7  h. The same is true for paid between ± 1  h per quarter over the business cycle. and unpaid overtime which together reduced average Unpaid overtime was most important at the beginning of working hours by another 0.9  h. Reductions in regular the considered period (Fig. 4). After the minor economic working time do not contribute to the cyclical reduc- slowdown in Germany related to the so-called Euro Cri- tion in working time in the Coronavirus Recession. The sis from 2011q3 to 2013q1, it lost its relevance for cycli- observed dominance of STW in the attempt to safeguard cal fluctuations. Interestingly, unlike unpaid overtime, employment in the Coronavirus Recession is in line with the cyclical variation of paid overtime continues after the the made discretionary policy changes governing the use Great Recession and can also be observed in the Covid- of STW. It is conceivable that the extended and simpli- 19 pandemic. fied use of short-time work “crowded out” to some extent In the Coronavirus Recession, the contributions of paid the use of other measures like e.g., WTA since already in and unpaid overtime to the cyclical reduction in working 11 Page 12 of 22 A. Herzog‑Stein et al. Fig. 7 Contributions to the cyclical working‑time reductions in the Great Recession and the Coronavirus Recession. The term ‘cyclical’ refers to the difference of actual and trend changes for each working‑time instrument (if the series shows a trend). STW and WTA show no trend. The trend is constructed applying the Hodrick‑Prescott filter with  = 1600 . All components are measured in working hours per employee. Sources: Institute for Employment Research (IAB) working time calculations; own calculations March 2020 no negative balances on WTA were required employers (Gallant et  al. 2020), was the prominent tool anymore as eligibility criteria for the use of STW. for dealing with the crisis. Finally, the important impact of the use of STW on While temporary unemployment has been between unemployment respectively employment is best seen 0.4 and 1.2% throughout the years and even during the by looking at the seasonally adjusted inflow rate from economic and financial crisis it played no prominent role employment into unemployment and the exit rate from with respect to the overall increase in unemployment, the unemployment into employment on a monthly basis share of workers on temporary layoffs jumped to 11.5% in (Fig. 8). From February to April 2020 the inflow rate from April 2020, accounting for almost 80% of all unemployed employment into unemployment increased from 0.5 persons (Fig.  9). While unemployment declines slowly to 0.7% and declined than quickly back to 0.5% in June, during an economic recovery, the work-finding rate for while the exit rate from unemployment into employment the temporarily laid-off unemployed is usually twice as decreased from 8.3 to 4.7% from February to May 2020 high as for the unemployed. Accordingly, unemployment and did not reach its pre-pandemic level until the end of fell faster this time than in previous recoveries (Hall and 2021. This can be seen as some indication that the mas - Kudlyak 2022). Thus, the rate of temporary unemploy - sive use of STW was able to prevent large and prolonged ment halved from April to July 2020, while the jobless flows from existing employment into unemployment. unemployment rate increased by 1.2 percentage points. External flexibility via temporary lay-offs was the 3.3 Unit ed States: external flexibility main means of the labour-market adjustment in the Although STW programs exist in about half of the U.S. states and STW utilisation was much higher than in the past, the use of STW has overall not played a major role The job-finding rate is often used to describe the labour market tightness. in the United States (Krolikowski and Weixel 2020). Here, Hall and Kudlyak (2022) use the work-finding rate instead in the context of the focus was rather on external flexibility. However, for the Covid-19 pandemic given that unemployed on temporary layoffs are not obliged to look for a job while waiting to be recalled. Some of them might the first time, the use of temporary lay-offs, i.e. laid-off nevertheless take a new job rather than waiting to be recalled by the previous individuals who expect to be recalled by their former employer. Germany and the United States in coronavirus distress: internal versus external labour market… Page 13 of 22 11 Fig. 8 Monthly unemployment flows (2009 to 2021). The monthly inflow rate from employment into unemployment (national definition) is defined as the number of inflows from employment to unemployment in month t relative to the employment level in month t ‑1. The monthly exit rate from unemployment into employment is defined as the number of outflows from unemployment (national definition) to employment in month t relative to the unemployment level in month t‑1. The numbers are seasonally adjusted. Sources: Federal Employment Agency; own calculations Fig. 9 US unemployment rate: temporary layoffs and jobless unemployed. Jobless unemployed comprises job losers not on layoff, job leavers, reentrants to labour force and new entrants to labour force. Sources: U.S. Bureau of Labor Statistics (BLS), own calculations 11 Page 14 of 22 A. Herzog‑Stein et al. Fig. 10 Proportion of persons in the United States unable to work due to lost business in the coronavirus pandemic. Supplemental data measuring the effects of the Coronavirus (COVID ‑19) Pandemic on the labour market. Persons unable to work at some point in the last 4 weeks because their employer closed or lost business due to the Coronavirus pandemic by receipt of pay from their employer for hours not worked and employment status. Sources: U.S. Bureau of Labor Statistics (BLS) Release “Eec ff ts of the Coronavirus (COVID ‑19) Pandemic on the Labor Market”, https:// www. bls. gov/ cps/ effec ts‑ of‑ the‑ coron avirus‑ covid‑ 19‑ pande mic. htm; own calculations Coronavirus Recession, but complementary to this, com- The majority of them reported that they were not com - panies also used some measures of internal flexibility by pensated by their employers. Only less than a quarter reducing the working hours of their employees. While received some compensation. Therefore, the reduction unlike in Germany there are no detailed information on respectively loss in working hours in the United States the average number of working hours lost per worker due took place in a way that is quite different from the short- to the Covid-19 pandemic in the United States, an addi- time allowance in Germany. tional survey conducted by the BLS beginning in May Overall, these information on unpaid as well as 2020 as part of the Current Population Survey provides on compensated temporary working time losses fit a good insight into the extent to which workers in non- together with the macroeconomic evidence for the agricultural industry were affected by the crisis. In this United States presented in Sect.  3.1. As shown above survey, workers were asked whether they had been una- there was some cyclical reduction in the average work- ble to work due to the pandemic in the last four weeks ing time per employee from peak to trough of 1.0% in and whether they had received any kind of payment the Coronavirus Recession. If we take into account that from their employers. Unfortunately, no information was the Coronavirus Recession was much shorter than the asked about the form of compensation paid or the exact Great Recession, the average individual working-time number of hours lost as a consequence of the pandemic. reduction per quarter was stronger this time. Given u Th s, it is also not known whether workers who were that “job-losses have disproportionally hit the low-wage compensated for working time lost due to the pandemic workforce” (Bateman and Ross 2021) it is also likely received any payments via one of the short-time work that the reported working-time losses were concen- programs at the state level. trated among the low-skilled. Since low-skilled workers In May 2020, 20% of workers reported that they were generally have a lower hourly labour productivity, this affected by some kind of loss of working time, and in Given that the survey asks about the employment situation in the previous June and July 2020 still more than 15 respectively 10% of four weeks, the May survey probably covers most of the employment situation workers experienced some loss of working time (Fig. 10). in April 2020, the month with the highest crisis impact. Germany and the United States in coronavirus distress: internal versus external labour market… Page 15 of 22 11 -10 -20 -30 -40 -50 Change in employment Unable to work Unable to work (received pay) (did not receive pay) Fig. 11 Proportion of persons in the United States unable to work (with and without compensation) due to lost business and change in employment by industry in April 2020. NAICS classification. 21 = Mining, quarrying, and oil and gas extraction, 23 = Construction, 31 = Durable goods manufacturing, 32–33 = Nondurable goods manufacturing, 42 = Wholesale trade, 44–45 = Retail trade,48–49 = Transpor tation and warehousing, 22 = Utilities, 51 = Information, 52 = Financial Activities, 54 = Professional & Business Services, 61–62 = Education & Health Services, 71–72 = Leisure & Hospitality, 81 = Other Services, 92 = Public administration, N.I. = Nonagricultural industries. Proportion of persons unable to work are from May 2020 which refers to the previous four weeks. Data for the change in employment refers to the monthly change from March to April 2020. Sources: U.S. Bureau of Labor Statistics (BLS) Release “Eec ff ts of the Coronavirus (COVID ‑19) Pandemic on the Labor Market”, https:// www. bls. gov/ cps/ effec ts‑ of‑ the‑ coron avirus‑ covid‑ 19‑ pande mic. htm, Current Employment Statistics (CES), own calculations would explain the anticyclical increase in labour pro- imply that firms do not use measures of internal flex - ductivity observed in the Coronavirus Recession (see ibility, too. The information from the economic sectors Fig. 4D). indicates that economic sectors that were hit hard by the More details about the American way of dealing with Covid-19 pandemic relied on external as well as on inter- the Coronavirus Crisis are revealed by looking closer nal flexibility in response to the Coronavirus Recession. at the change in employment and the share of work- There is a strong positive correlation between layoffs ers affected by working hours lost due to the inability to and working-time reductions with a correlation coeffi - work in different economic sections of the US economy cient of 0.9: in economic sectors with a larger reduction (Fig.  11). Given that for economic sectors no data for in employment there is also a larger share of workers temporary layoffs are available the change in employ - with a loss of working hours due to the inability to work. ment is used instead to indicate the intensity with which However, this positive relationship is driven by the posi- employers were hit by job losses across economic sectors. tive correlation between employment reductions and As in Germany, the economic sections have been affected working-time losses without compensation; there is no differently by the Covid-19 pandemic, the service sectors correlation between the magnitude of employment losses more than the industry. and the size of the share of workers with renumerated Interestingly, a combination of layoffs and reductions working-time losses. This suggests that, in contrast to the in working hours dominates in all sectors of the econ- experience in Germany, in the United States the burden omy. Hence, the dominance of external flexibility in the of labour market flexibility in the Coronavirus Recession labour market adjustment in the United States does not is borne primarily by workers. 32-33 44-45 48-49 61-62 71-72 N.I. 11 Page 16 of 22 A. Herzog‑Stein et al. Fig. 12 Contributions (in percentage points) of the components of US unemployment in the Great Recession and the Coronavirus Recession. Jobless unemployed comprises job losers not on layoff, job leavers, reentrants to labour force and new entrants to labour force. Sources: U.S. Bureau of Labor Statistics (BLS); own calculations In conclusion, the United States have relied again heav- points were due to temporary layoffs and only 0.7 per - ily on the use of external flexibility. However, there are centage points to workers who lost their jobs. Interest- also major differences in its response compared to the ingly, fewer employees seemed to leave their job of their Great Recession as a breakdown of the change in the own accord, and no change in labour force entry was unemployment rate in the two crises reveals (Fig. 12). observable. Most of the change in the unemployment rate is deter- mined by the number of employees losing their jobs, usu- 4 Blind spots of the chosen strategy ally without being on recall. During the Great Recession, Germany and the United States pursued different goals the unemployment rate rose by 4.5 percentage points. with their Coronavirus Crisis responses. Germany Only a small part of 0.5 percentage points was due to focused on employment protection via mechanisms temporary layoffs. Another 3.1 percentage points of the designed to promote internal flexibility. The United increase in the unemployment rate was due to workers States focused on a mixture of external flexibility and losing their jobs. The proportion of job leavers among the income protection. As strategy objectives were differ - unemployed is hardly influenced by the business cycle ent from one another, the blind spots are likely to differ and lies typically in a range between 0.5 and 0.6%. Hence, between the countries. Therefore, we discuss the chal - its contribution to the change in unemployment over lenges each strategy poses in this last section. time is negligible. Almost one percentage point of the increase was due to re-entrants and new entrants into the 4.1 Germany’s employment protection labour market during the Great Recession. For the success of Germany’s chosen strategy, the main As shown in the analysis above the labour market goal was to secure existing employment relationships to response during the Coronavirus Recession was extraor- save firm specific human capital, prevent unemployment, dinary and very different to the one observed in the Great and to reduce future training costs. Recession. For the first time, temporary layoffs played a The burden of job losses in the Coronavirus Reces - prominent and dominant role in the United States. The sion was unevenly distributed in Germany. Job loss unemployment rate rose by a total of 9.4 percentage rates of employment subject to social security contribu- points from peak to through, of which 8.8 percentage tions, which enjoys the protection of the STW scheme, Germany and the United States in coronavirus distress: internal versus external labour market… Page 17 of 22 11 were less severe than the losses of marginal employment employment as well as self-employment are not pro- (Minijobs). tected by the STW scheme. Employment subject to social security contributions, the backbone of the German welfare state, decreased by 4.2 Differences in income protection in Germany around 450,000 jobs or 1.3% between March and May One advantage of STW besides securing existing employ- 2020 (see Fig. 13A). In line with their growth trends dur- ment relationships and firm specific human capital is that ing the long boom before the outbreak of the Coronavi- the income decrease during STW is less severe than with rus Crisis, the recovery of employment subject to social an immediate fall back to unemployment benefits. How - security contributions was more dynamic than that of ever, since income replacement rates of STW are compa- total employment in the summer and fall 2020. rable with the initial replacement rates of unemployment Workers in marginal employment (Minijobs), who benefits, the impact of STW on income is dependent on overwhelmingly work in the services sector were severely the previous wage and the reduction in working time. hit by the economic crisis due to the Covid-19 pan- This is also why Germany’s focus on employment protec - demic. In the months March to May 2020 the percent- tion via STW also helped with income protection. The age decrease in marginal employment (−  7.5%) was five widespread use of STW not only safeguarded employ- times as large as in employment subject to social security ment, but also secured part of household income for contributions (see Fig. 13B). households whose members were affected by STW. Even after accounting for the different growth trends One challenge of STW during the Coronavirus Reces- of these two employment forms, marginal employment sion was that the average short-time worker was very dif- was more severely hit by the Coronavirus Crisis. Fur- ferent from the average short-time worker in the Great thermore, while both employment subject to social secu- Recession. The massive use of STW in other economic rity contributions and marginal employment started to sections than manufacturing as well as the more intensive recover in the summer months, marginal employment use of STW in general during the Coronavirus Recession declined again with the second wave of the pandemic. have immediate income effects. There are two obvious reasons for these remark - In Fig. 15 the average income losses due to STW in each able employment patterns during the current Corona- economic section are plotted for the Great Recession and virus Recession, which are interlinked. First, this time the Coronavirus Recession. Interestingly, whereas in the the services sector was much more hit by the economic Great Recession there was no clear negative correlation crisis than during the Great Recession. Furthermore, in between the level of earnings and the percentage earn- the Coronavirus Crisis the necessity to temporarily lock ings loss due to STW (blue dots), the situation during down and interrupt parts of economic activity to prevent the Coronavirus Recession is completely different. We the spread of infection cannot be overcome by stimulat- observe a clear negative correlation with a correlation ing aggregate demand. Rather, economic policies must coefficient of −  0.7 (orange dots). The lower the average try to sustain businesses, and hence employment, during earnings are in an economic section the higher is the per- these periods of (partial) lockdown and interruption of centage income loss due to STW. The difference is likely production processes. to be even more pronounced if we were to consider addi- Second, STW, the major pillar of the government’s tional supplements of the short-time work allowance due strategy to safeguard employment is not applicable to to the employer which is more often paid in jobs with marginal employment. This left more jobs unprotected higher earnings (Pusch and Seifert 2020, Table 3). Pusch in the services sector, where marginal employment con- and Seifert (2020, Table  3) present the share of employ- stitutes a larger share of total employment. In services ees who receive a supplement to STW allowance. There sectors like accommodation and food service activities is a positive correlation: The higher the average earnings (section  I), or arts, entertainment and recreation (sec- in an economic sector, the higher the share of employees tion  R), more than 40% of all employees were either who receive a supplementary STW allowance from their working in marginal employment as their only or as their employers. This implies that while the initial income loss second job (Fig. 14). was strongest in sectors with the lowest incomes, addi- Overall, since the services sectors were more severely tional supplements to STW were concentrated among affected by the Coronavirus Recession than by the Great sectors with the highest incomes. Recession, some weaknesses in the approach to safeguard As pointed out before, not only short-time work- employment became visible. In contrast to employ- ers suffered income losses due to the loss of work. In ment subject to social security contributions, marginal general, all types of employment, who lost (temporary) part or all of their work, or even became unemployed, Marginal employment exhibits a marked negative growth trend since the suffered income losses. Groups like the self-employed introduction of the general legal minimum wage at the beginning of 2015. 11 Page 18 of 22 A. Herzog‑Stein et al. (A)EMPLOYMENT SUBJECT TO SOCIAL SECURITY CONTRIBUTIONS (B) MARGINAL EMPLOYMENT Fig. 13 Different types of employment (2008–2021). Level (line, left scale) and change (columns, right scale) measured in 1000 persons (seasonally adjusted). Sources: Federal Employment Agency; Bundesbank; own presentation Germany and the United States in coronavirus distress: internal versus external labour market… Page 19 of 22 11 Fig. 14 Composition of employment according to economic sectors (March 2020). B: Mining and quarrying; C: Manufacturing; D: Electricity, gas, steam and air conditioning supply; E: Water supply; sewerage, waste management and remediation activities; F: Construction; G: Wholesale and retail trade; repair of motor vehicles and motorcycles; H: Transportation and storage; I: Accommodation and food service activities; J: Information and communication; K: Financial and insurance activities; L: Real estate activities; M: Professional, scientific and technical activities; N: Administrative and support service activities; O: Public administration and defence; compulsory social security; P: Education; Q: Human health and social work activities; R: Arts, entertainment and recreation; S: Other service activities. Sources: Federal Employment Agency, own calculations or workers in marginal employment were not entitled and the focus is on re-employment, this approach has to STW or unemployment benefits. Walwei (2021) the potential to target individual’s particular in need of stresses that, in contrast to the previous crisis, employ- income stabilization. ment associated with weak income security (marginal Temporary layoffs were the dominant component employment and solo self-employment) were particu- of the unemployment increase (Fig.  9). However, larly hard hit in the Coronavirus Crisis. they were rather unevenly distributed, and concen- To conclude, STW did not protect employment and trated among high wage employees. While there was income of everyone equally. This is why further discre - a rebound of employment for high wage earners, there tional measures were taken in a later stage of the Cor- were persistent job losses among low wage earners onavirus Crisis to reduce the blind spots of the STW (Cajner et al. 2020; Chetty et al. 2020). instrument. These additional discretionary policies as As discussed in Sect. 2, the CARES Act covered vari- in example the one-off payments for children and an ous instruments to secure income of individuals. Three increase in STW replacement rates for long term recip- policies to stabilize income directly or indirectly are ients helped to cushion the income loss (Christl et  al. of particular interest: The PPP, the Federal Pandemic 2021). Unemployment Compensation and Stimulus Checks. While the initial idea of the PPP was to stabilise the 4.3 D ifferences in income protection in the United States income of individuals otherwise losing their jobs, stabilise There are two objectives for the success of the United companies’ financial flows, and support business owners States’ chosen strategy. First, the job loss should be through direct support for small business, the actual ben- temporary and not permanent. Second, the income of efits had little to do with employment protection. Evalu - individuals should be protected despite the temporary ations overall indicate that the program was untargeted job loss. As long as unemployment is only temporary and inefficient (Chetty et al. 2020; Autor et al. 2022). Even 11 Page 20 of 22 A. Herzog‑Stein et al. Q F Q J 0500 1000 1500 2000 2500 3000 Average gross monthly earnings Fig. 15 Earnings loss due to short‑time work by economic sections. Blue (orange) dots indicate average income losses in the Great Recession (Coronavirus Recession) in %. For the comparison we use the information about the average amount of working time lost due to STW in the two crises which is provided by the Federal Employment Agency (Employment equivalent/number of short‑time workers). Given that STW is paid on the basis of net earnings losses we calculate the impact of STW on average gross monthly earnings (SOEP) in Euro by using the Kurzarbeiterrechner on the assumption that the short‑time worker is single, in tax class 1, without children. Sources: Federal Employment Agency, SOEP, Kurzarbeiterrechner (https:// www. netto lohn. de/ rechn er/ kurza rbeit ergeld. html), own calculations when it safeguarded employment, this was concentrated Germany, in contrast, labour hoarding was encouraged among employees in the top income quintile (Autor et al. through STW programmes, which maintain employment 2022). However, as pointed out by Autor et al. (2022) this relations between workers and firms. As a result, in the mainly results from the trade-off between a timely ver - United States the unemployment rate (which is largely sus a targeted intervention under limited administrative driven by temporary layoffs) and the non-employment capacities. rate rose sharply, as have the take up rates of STW in In contrast, the Stimulus Checks and the Federal Pan- Germany. demic Unemployment Compensation proved to be an Put differently, initial income support was stronger in efficient tool to secure income. While large earning the United States, but despite high levels of temporary declines were more likely for low wage workers, both layoffs, lead to a higher share of permanent job losses instruments together outweighed the otherwise result- (Barrero et  al. 2021), which in turn required more job ing income losses (Larrimore et  al. 2022) and stabilised search and reallocation activities. This was particu - income particularly at the lower end of the income distri- larly true for lower wage earners. It remains to be seen bution (Autor et al. 2022). Ganong et al. (2020) show that whether these reallocation effects will be beneficial in the the majority of workers eligible for unemployment ben- succeeding recovery. Germany, in contrast, secured exist- efits between April and July 2020 had replacement rates ing employment relations at the expense of temporary above 100%. This lead to only a modest increase in pov - income losses without protecting workers in marginal erty rates until the unemployment supplements expired employment. While unemployment also leads to a per- (Parolin et al. 2020). sistent decline in wages, Giupponi et al. (2022) point out that Germany’s previous experience with STW indicates 4.4 Comparison of the strategies that wages of secured workers adjust to its pre-crisis The United States decided to insure workers’ incomes level. The German approach therefore allowed to secure against the costs of job losses by increasing the generos- employment relations and human capital in the short run ity, eligibility criteria and eligibility period of unemploy- and secured income in the long run. ment benefits and other income support measures. In Average income loss due to STW Germany and the United States in coronavirus distress: internal versus external labour market… Page 21 of 22 11 Overall, both countries protect individuals in weaker However, a closer look at the blind spots of the positions on the labour market (atypically employed and chosen strategies in Germany and the United States low wage earners) to a smaller extent through the chosen showed that despite the differences in the respective strategies. This is an indication of the segmentation of the approaches, people in weaker labour market positions labour market by wages and working conditions (Reich were less well protected by the chosen strategies. In et al. 1973), which also reflect their lack of protection by Germany, marginally employed workers who lost their labour market policies during the crisis. jobs were not protected by either STW or unemploy- Additionally, regardless of the labour market measure ment insurance. Moreover, low-income earners for chosen, STW as well as the unemployment insurance whom the short-time allowance was not sufficient were carry the risk of moral hazard for labour market actors additionally dependent on basic income support. In and are associated with social costs. Regarding STW, the United States, the Stimulus Checks and the Federal Cahuc et al. (2021) document moral hazard problems on Pandemic Unemployment Compensation proved to be the firm side by showing that firms with relatively low an effective instrument to secure income, especially for revenue shocks tend to reduce their employees’ working low wage  earners. But they were even less protected hours without actually safeguarding their jobs. In con- from job losses and suffered disproportionately. trast, too generous unemployment benefits may induce Acknowledgements moral hazard issues if they reduce search effort for new We thank Lukas Lehner, Janine Leschke, Fabian Lindner, Camille Logeay, Hart‑ jobs (Schmieder et al. 2016). Despite the discussed blind mut Seifert, Sabine Stephan, two anonymous referees and the editor Britta Gehrke for valuable comments. spots and social costs, Giupponi et al. (2022, 50) conclude “…that short-time work can be an efficient and expedi - Author contributions ent way to attenuate the social costs created by “excess” All four authors contributed in equal shares to the paper. All authors read and approved the final manuscript. layoffs in recessions”. Furthermore, since they argue for instance  that both instruments cover different types of Funding workers, STW and unemployment insurance can be val- Not applicable. uable complements. Availability of data and materials The data that support the findings of this study are available from the 5 Conclusion Hans-Böckler-Stiftung and the Research Data Center of the Socio-Economic Panel but restrictions apply to the availability of these data, which were used The global Covid-19 pandemic hit the economies of Ger - under license for the current study, and so are not publicly available. Data many and the United States  hard. Both countries expe- are however available from the authors upon request and with permission of rienced an economic downturn of similar magnitude. the Hans-Böckler-Stiftung and the Research Data Center of the Socio-Economic Panel. Interestingly, Germany and the United States pursued very different economic strategies to minimise the impact Declarations of the Coronavirus Crisis on the labour market. While Germany focused on safeguarding existing jobs through Ethics approval and consent to participate the use of internal flexibility measures, especially STW, Not applicable. the United States relied on a mix of external flexibility Consent for publication and income protection. This fact allowed us to examine Not applicable. more closely the German strategy of securing employ- Competing interests ment through internal flexibility and the German labour The authors declare that they have no competing interests. market development during the Coronavirus Recession by contrasting it with the chosen strategy and the labour Author details Macroeconomic Policy Institute (IMK), Georg‑Glock‑Str. 18, 40474 Düsseldorf, market development in the United States. 2 3 Germany. University of Koblenz‑Landau, 76829 Landau, Germany. K iel Uni‑ Our analysis has shown that Germany responded to 4 versity, Wilhelm‑Seelig‑Platz 1, 24118 Kiel, Germany. University of Muenster, the economic shock with a massive temporary cyclical Schlossplatz 2, 48149 Münster, Germany. reduction in working hours, mainly through STW, on Received: 24 November 2021 Accepted: 11 July 2022 a historic scale and unemployment rose only moder- ately. In the United States, on the other hand, unem- ployment rose at an unprecedented rate, but unlike previous recessions, the nature of unemployment was References quite different, being driven mostly by temporary lay- Autor, D., Cho, D., Crane, L.D., Goldar, M., Lutz, B., Montes, J., Peterman, W.B., Ratner, D., Villar, D., Yildirmaz, A.: The $800 billion paycheck protection offs. This allowed for a very fast recovery of unem- program: where did the money go and why did it go there? J. Econ. ployment in the United States—much faster than in Perspect. 36(2), 55–80 (2022). https:// doi. org/ 10. 1257/ jep. 36.2. 55 previous recessions. 11 Page 22 of 22 A. Herzog‑Stein et al. Balleer, A., Gehrke, B., Hochmuth, B., Merkl, C.: Autonomes Fahren statt stop Schmieder, J.F., von Wachter, T., Bender, S.: The effect of unemployment ben‑ and go: Vorschläge zur effektiven Gestaltung der deutschen Kurzarbeit. Z. efits and nonemployment durations on wages. Am. Econ. Rev. 106(3), Wirtsch. 68(3), 252–260 (2019) 739–777 (2016) Bateman, N., Martha, R.: The pandemic hurt low‑ wage workers the most—and Steffen, J.: Sozialpolitische Chronik. Die wesentlichen Änderungen in der so far, the recovery has helped them the least. https:// www. brook ings. Arbeitslosen‑, Renten‑, Kranken‑ und Pflegeversicherung sowie bei der edu/ resea rch/ the‑ pande mic‑ hurt‑ low wage‑ w‑orke rs‑ the‑ most‑ and‑ so‑ Sozialhilfe (HLU) und der Grundsicherung für Arbeitsuchende—von den far‑ the‑ recov ery‑ has‑ helped‑ them‑ the‑ least/ (2021). Accessed 3 May siebziger Jahren bis heute.” Portal Sozialpolitik. www. portal‑ sozia lpoli tik. 2022 de (2021) Barrero, J.M., Bloom, N., Davis, S.J., Meyer, B.H.: COVID‑19 is a persistent realloca‑ Walwei, U.: Erwerbsformen in Krisenzeiten: was folgt aus Corona? WSI‑Mit ‑ tion shock. AEA Papers Proc. 111, 287–291 (2021) teilungen 74(2), 151–159 (2021) Cajner, T., Crane, L., Decker, R., Grigsby, J., Hamins‑Puertolas, A., Hurst, E., Kurz, C., Yildirmaz, A.: The U.S. labor market during the beginning of the pan‑ Publisher’s Note demic recession. Brookings Papers on Economic Activity, Summer 2020, Springer Nature remains neutral with regard to jurisdictional claims in pub‑ 3‑33 (2020) lished maps and institutional affiliations. Cahuc, P., Kramarz, F., Nevoux, S.: The heterogeneous impact of short‑time work from saved jobs to windfall effects. IZA Discussion Paper No. 14381 (2021) Chetty, R., Friedman, J. N., Hendren, N., Stepner, M., The Opportunity Insights Team: How did COVID‑19 and stabilization policies affect spending and employment? A new real‑time economic tracker based on private sector data. NBER Working PaperNo. 27431 (2020) Christl, M., de Poli, S., Hufkens, T., Peichl, A., Ricci, M.: The role of short‑time work and discretionary policy measures in mitigating the effects of the COVID ‑ 19 crisis in germany. CESifo Working Paper No. 9072 (2021) Committee for a Responsible Federal Budget (CRFB): COVID Money Tracker. Committee for a Responsible Federal Budget. https:// www. covid money track er. org/ (2021). Accessed 19 July 2022 Ellguth, P., Gerner, H‑D., Zapf, I.: Flexible Arbeitszeitgestaltung wird immer wichtiger. IAB‑Kurzbericht 15/2018 (2018) Federal Ministry of Finance (BMF): Corona‑Unternehmenshilfen—eine vor ‑ läufige Bilanz. Monatsbericht des BMF (November 2021). Analysen und Berichte (2021) Ganong, P., Noel, P., Vavra, J.: US unemployment insurance replacement rates during the pandemic. J. Public Econ. 191, Article 104273 (2020) Gallant, J., Kroft, K., Lange, F., Notowidigdo, M.: Temporary unemployment and labor market dynamics during the COVID‑19 recession. Brookings Papers on Economic Activity, Fall 2020, 167‑216 (2020) Gehrke, B., Hochmuth, B.: Counteracting unemployment in crises: non‑linear effects of short ‑time work policy*. Scand. J. Econ. 123(1), 144–183 (2021). https:// doi. org/ 10. 1111/ sjoe. 12395 Giupponi, G., Landais, C., Lapeyre, A.: Should we insure workers or jobs during recessions? J. Econ. Perspect. 36(2), 29–54 (2022) Hall, R., Kudlyak, M.: The unemployed with jobs and without jobs. NBER Work‑ ing Paper No. 27886 (2022) Herzog‑Stein, A., Zapf, I.: Navigating the great recession. ILR Rev. 67(3), 891–925 (2014) Herzog‑Stein, A., Lindner, F., Sturn, S.: The German employment miracle in the great recession: the significance and institutional foundations of tempo ‑ rary working‑time reductions. Oxf. Econ. Pap. 70(1), 206–224 (2018) Herzog‑Stein, A., Nüß, P., Peede, L., Stein, U.: Germany’s labour market in Coronavirus distress. New challenges to safeguarding employment. IMK Working Paper No. 209 (2021) Houseman, S.: IZA COVID‑19 crisis response monitoring: United States (Janu‑ ary 2022). Country Report (2022) Krolikowski, P.M., Weixel, A.: Short‑time compensation: an alternative to layoffs during COVID‑19. Economic Commentary No. 2020‑26. Federal Reserve Bank of Cleveland (2020). https:// doi. org/ 10. 26509/ frbc‑ ec‑ 202026 Kruppe, T., Osiander, C.: Kurzarbeit in der Corona‑Krise: Wer ist wie stark betrof‑ fen?—IAB‑Forum. https:// www. iab‑ forum. de/ kurza rbeit‑ in‑ der‑ corona‑ krise w‑er‑ ist wie‑‑ stark‑ betroen/ ff (2020). Accessed 11 May 2022 Larrimore, J., Mortenson, J., Splinter, D.: Earnings shocks and stabilization dur‑ ing COVID‑19. J. Public Econ. 206, Article 104597 (2022) Parolin, Z., Curran, M., Matsudaira, J., Waldfogel, J., Wimer, C.: Monthly poverty rates in the United States during the COVID‑19 pandemic. Poverty and Social Policy Working Paper (2020) Pusch, T., Seifert, H.: Kurzarbeit in der Corona‑Krise mit neuen Schwerpunkten. Wirtschafts‑ und Sozialwissenschaftliches Institut ( WSI). WSI Policy Brief 47 (09/2020) (2020) Reich, M., Gordon, D.M., Edwards, R.C.: A theory of labor market segmentation. Am. Econ. Rev. 63(2), 359–365 (1973) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal for Labour Market Research Springer Journals

Germany and the United States in coronavirus distress: internal versus external labour market flexibility

Loading next page...
 
/lp/springer-journals/germany-and-the-united-states-in-coronavirus-distress-internal-versus-FFGXIAsjzS
Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2022
ISSN
1614-3485
eISSN
2510-5027
DOI
10.1186/s12651-022-00316-5
Publisher site
See Article on Publisher Site

Abstract

Germany and the United States pursued different economic strategies to minimise the impact of the Coronavirus Crisis on the labour market. Germany focused on safeguarding existing jobs through the use of internal flexibility measures, especially short‑time work (STW ). The United States relied on a mix of external flexibility and income pro ‑ tection. On this basis, we use macroeconomic time series to examine the German strategy of securing employment through internal flexibility by contrasting it with the chosen strategy in the United States. In Germany, temporary cyclical reductions in working hours are mainly driven via STW. US unemployment rose at an unprecedented rate, but unlike in previous recessions, it was mostly driven by temporary layoffs. However, a closer look at the blind spots of the chosen strategies in both countries showed that despite the different approaches, people in weaker labour market positions were less well protected by the chosen strategies. Keywords: Working‑time reduction, Safeguarding employment, Unemployment, Internal flexibility, External flexibility, Short ‑time work, Temporary layoffs, Great recession, Coronavirus recession, Covid‑19 pandemic JEL: E24, E32, J08, J20 1 Introduction hoarding through STW programmes, the United States During both the Great Recession and the Coronavirus decided to insure worker’s incomes with instruments such Recession, unemployment in Germany increased only mod- as cash transfers and temporary increases in unemployment erately, especially given the severity of these recessions. benefits instead of protecting employment. However, in both recessions there were instead significant Giupponi et  al. (2022) discuss extensively upsides temporary reductions in working time by means of internal and downsides of both strategies. We contribute to this flexibility, i.e., the internal adjustment of the labour input debate by providing a detailed descriptive analysis of the used in the production process along the intensive margin German strategy of safeguarding employment via inter- especially through the use of short-time work (STW). While nal flexibility by contrasting it with the strategy chosen most European countries have also relied on STW to tackle in the United States and the labour market experience in the crisis, in the United States external flexibility was domi - the two countries in 2020. nant, i.e., the adjustment of labour input via the external For Germany, we show that despite the dramatic labour market, and there was a sharp temporary increase in decrease in real gross domestic product (GDP), unem- unemployment of historic proportions. Instead of labour ployment only increased moderately. While during the Great Recession all working-time instruments con- tributed to the reduction in working time, STW now *Correspondence: alexander‑herzog‑stein@boeckler.de accounts for almost all of the working-time reduction Macroeconomic Policy Institute (IMK), Georg‑Glock‑Str. 18, 40474 Düsseldorf, Germany For more details on the difference between internal and external (numerical) Full list of author information is available at the end of the article flexibility see for example Herzog-Stein and Zapf (2014). © The Author(s) 2022. Open Access. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. 11 Page 2 of 22 A. Herzog‑Stein et al. Fig. 1 Real gross domestic product in Germany and the United States (2005–2021). Quarterly change in real GDP; seasonally adjusted. Sources: Federal Statistical Office (Destatis); Bureau of Economic Analysis; authors’ calculations as the government focused on this instrument to main- of the chosen policies in the two countries. Section  4.1 tain employment. In contrast, the United States expe- concludes. rienced a comparable decrease in real  GDP, but also a temporary increase in unemployment on an unprec- 2 The German and US labour market edented scale. However, the nature of unemployment during the Covid‑19 pandemic changed as well. In contrast to the Great Recession, The outbreak of the global Covid-19 pandemic and its temporary layoffs during the Coronavirus Recession economic impact on the world economy caused a major played a dominant role in the United States and allowed economic crisis. Both the German and the US economy for a fast recovery of the unemployment rate. While were severely affected by the pandemic and experi - both countries approached the crisis differently, enced economic contractions of similar magnitude. The remarkably their weakness is the same. The methods to German economy started to be severely affected by the secure employment and income were less pronounced pandemic at the end of the first quarter 2020 and went for individuals in weaker positions on the labour mar- into a partial lockdown from mid-March to May 2020. ket (atypical employment and low wage earners). The result was an economic slump of historic propor - The remainder is structured as follows: In Section  2 tions. In the second quarter  2020, real GDP contracted we provide a concise summary of how the German by 9.9%, after it fell already by 0.7% in the first quarter labour market was affected by the Coronavirus Cri - 2020 (Fig.  1). In total in the first two quarters 2020 real sis relative to the experience in the United States and GDP fell by 10.5% seasonally adjusted. The United States describe the differing labour-market related policy declared a public health emergency at the End of Janu- responses in the two countries. Section  3 presents a ary 2020 and in mid-March 2020 a national emergency comparative business-cycle analysis of the Coronavi- in response to the Covid-19 pandemic. From Spring rus Recession and the Great Recession in Germany as 2020 onwards there were also widespread business clo- well as in the United States and has a closer look at the sures and social distancing practices (Houseman 2022). country specific margins of labour market adjustments. Real  GDP fell seasonally adjusted by 8.9% in the second Then, in Section  4  we investigate potential blind spots quarter 2020 after a 1.3% decrease in the first quarter. Germany and the United States in coronavirus distress: internal versus external labour market… Page 3 of 22 11 Fig. 2 Unemployment rates in Germany and the United States (2005 to 2021). The definition used in the Labour Force Survey follows the definition of the Internal Labour Organization. Number of all unemployed people aged 15–74 as percentage of the labour force. Sources: Eurostat; own presentation The Coronavirus Crisis had a marked impact on labour In comparison to the Great Recession, the increase market performance in Germany and the United States. in the US unemployment rate from its minimum was Unemployment increased markedly in both countries around twice as large in the Coronavirus Recession, but (Fig.  2). However, the magnitude of the rise in unem- the following decline was much quicker this time. In Ger- ployment in the two countries was different. many, unemployment rose faster this time than in the While in the United States the unemployment rate Great Recession, when unemployment started to rise in increased dramatically from 3.5 to 14.7% from February November 2008 and peaked in July 2009 (see Fig. 2). With to April 2020, in Germany the rise in unemployment was a total increase of 0.6 percentage points the rise in unem- much less pronounced but went on for longer. From Feb- ployment was now slightly less than at that time (+ 0.9 ruary to August 2020 the German unemployment rate percentage points). Compared to the United States and increased from 3.5 to 4.1% and started only to decline against the backdrop of the massive decline in economic again in December 2020. In contrast, in the Unites States activity the rise in unemployment was in both downturns unemployment declined relatively fast and continuously relatively moderate in Germany. after its peak in April 2020. In addition, the US civilian The differences in the development of unemployment labour force declined by 5.0% from February to April in the two countries in the Coronavirus Recession are 2020, while in Germany the labour  force decrease from quite remarkable. But the policy responses in Germany February to May 2020 was much smaller at 1.5%. and the United States regarding the labour market as a result of the economic impact of the pandemic were also quite different. Table  1 contrasts the discretionary pol- icy changes in Germany with those in the United States in 2020 sorted into four categories: legislative changes To enable a consistent comparison between the two countries, we use the regarding the respective STW scheme, the provision of unemployment rate as defined by the International Labour Office (ILO). For Germany, the monthly development of this unemployment rate during the business support, the expansion of unemployment ben- Covid-19 pandemic differs with respect to the timing and the magnitude of efits, and the provision of income support for house - the increase from the unemployment rate of the Federal Employment Agency, holds. STW and business support reflect measures that which is defined differently. 11 Page 4 of 22 A. Herzog‑Stein et al. Table 1 Discretionary policy changes during the Coronavirus Crisis in the United States and Germany (dating corresponds to date of passage) Germany United States 2020Q1 Short‑time work Short‑time work • Simplified eligibility criteria • Federal financial support for states’ STC schemes • 100% reimbursement of social insurance contributions for hours affected by STW • Expansion of additional income opportunities during STW Business support Business support • Soforthilfen: grants for small businesses & self‑ employed, administrated • Paycheck Protection Program (PPP): forgivable loans to SME by Länder • Employee Retention Tax Credit: payroll tax credit for employers • Loans and credit guarantees • Tax credits for obliged paid leave by employees • Disaster Loans Program: Federal Funding of loans from Small Businesses Administration Unemployment benefits Unemployment benefits • Simplified eligibility criteria for basic income support (ALG II) (e.g., fur ‑ • Federal Pandemic Unemployment Compensation (FPUC): additional benefit ther inclusion of self‑ employed, suspension of means testing) of $600 per week until end of July • Pandemic Emergency Unemployment Compensation (PEUC): extension of max. eligibility period • Pandemic Unemployment Assistance (PUA): simplified eligibility criteria (e.g., inclusion of self‑ employed) Income support (households) Income support (households) • Simplified eligibility criteria for child supplements • Child Tax Credits • Compensation for earnings losses due to child care • Economic Impact Payments: Stimulus checks for households per eligible adult ($1200) and child ($500) • Obligation to SME to provide paid sick leave, paid family leave, paid medi‑ cal leave 2020Q2 Short‑time work Business support • Extension of maximum eligibility period • Improved conditions regarding loans and forgiveness of loans (June) • Temporary increase in replacement rates until end of 2020 • Amendment of PPP: Provision of additional $320 bn. (April) Unemployment benefits • Expansion of max. eligibility period for unemployment insurance (ALG I) by three months 2020Q3 Business support Unemployment benefits • Grants for SME & self‑ employed (Überbrückungshilfe) •Extension and modification of expiring FPUC: $300 per week and condi‑ • Hiring credit for apprentices for SME of €2000 tioning eligibility on receiving at least $100 from other state unemploy‑ ment benefits Unemployment benefits Income support (households) •Extension of simplified eligibility criteria for ALG II •Provision of assistance to renters and homeowners Income support (households) • Stimulus Checks for families: €300 per child (Kinderbonus) • Tax Credits for single parents • Expansion of max. eligibility period for elderly/child care compensation • Temporary VAT reduction 2020Q4 Short‑time work Business support • Extension of the key measures until end of 2021 •Modification & extension of different programs (i.a., the Employee Retention Tax Credit, PPP, etc.) Business support Unemployment benefits • Grants for foregone revenues (November-/Dezemberhilfen) • FPUC: $300 per week until March 2021 • Grants for SME (Überbrückungshilfe II) • PEUC applies until March 2021, expansion of max. eligibility period to 24 weeks • PUA applies until March 2021; expansion of max. eligibility period to 50 weeks Income support (households) •Economic Impact Payments for adults ($600) and children ($600) Sources: Federal Ministry of Labour and Social Affairs, Federal Ministry of Finance, Steffen (2021) for Germany; U.S. Government, Chetty et al. (2020), Houseman (2022) for United States Germany and the United States in coronavirus distress: internal versus external labour market… Page 5 of 22 11 incentivise the use of internal flexibility to safeguarding In Germany all programs providing business support employment. Expansions of unemployment benefits and in 2020 account for €54.73 billion  or 1.6% of nominal income support for households aim at insuring incomes GDP in 2020 (Federal Ministry of Finance 2021). The given that establishments adjust labour input along the main programs at the federal level are the Soforthilfen in extensive margin. March, the Überbrückungshilfen  I–III from the Stabiliz- In short, while Germany put a focus on subsidising the ing Package in July, and the November-/Dezemberhilfen. use of internal flexibility, especially by providing more These programs contained grants and forgivable loans. generous STW allowance, the United States put a strong Apart from the differences in incentivising the use of focus on insuring incomes by providing generous unem- measures of internal flexibility between Germany and ployment benefits and allow for adjustments along the the United States there are also differences in the extent extensive margin. of insuring workers’ incomes in case establishments Policy measures to expand the use of STW had only adjust their labour input along the extensive margin. The a relevant impact in Germany. In line with the empiri- CARES Act introduced additional federal unemployment cal evidence on the effectiveness of the rule-based and benefits (FPUC) of $600 per week until July 2020. In fur - the discretionary component of STW in safeguard- ther programs the maximum eligibility period and eligi- ing employment in the Great Recession (Balleer et  al. bility of further worker groups has been introduced. In 2019; Gehrke and Hochmuth 2021), discretionary policy August the FPUC was replaced by $300 per week, and in changes made the use of STW in the pandemic more the Consolidated Appropriations Act the maximum eligi- attractive. The similarity of the changes in both crises is bility period was further expanded. striking (Herzog-Stein et  al. 2021), albeit the extensions In contrast to the US, the discretionary changes in were much faster and more generous this time. Crucial unemployment benefits in Germany had only minor rele - for the increased take up of STW were the extension of vance in insuring incomes of the unemployed. In the first the eligibility period of STW and the simplified eligibility Social Security Package in March  2020 eligibility criteria criteria with respect to the scope of STW in March 2020. for basic income support (ALG II) were simplified. In the Moreover, immediately a full reimbursement of social subsequent second Social Security Package from May security contributions for hours affected by STW was 2020 the maximum eligibility period for unemployment introduced to reduce residual costs of companies when benefits was expanded by three months. using STW. Thus, strong incentives for companies to use Besides insuring incomes with unemployment ben- STW were created. efits, stimulus payments are of particular importance to In contrast, in the United States, STW measures stabilize incomes. In Germany, the Stabilizing package mainly consist of financial support for state-level STW contains a stimulus check for families by providing €300 schemes by the federal level. Moreover, only 26 states had per child and tax credits for single parents. Still, they are STW schemes at the beginning of the crisis implemented by far not as expansive as the Economic Impact Payments (Houseman 2022). in the US. The CARES Act entailed so-called Economic Instead of subsidizing working-time reductions by Impact Payments of up to $1200 per adult for eligible expanding the STW schemes, the US government aimed individuals (earning less than $75,000) and $500 per at subsidised labour hoarding by providing business sup- qualifying child under age 17. Additionally, a child tax port especially with the CARES Act in March 2020. These credit was implemented. measures aimed at incentivising businesses to retain their employees, too. Major elements of the CARES Act were 3 Germany and the United States: internal the Paycheck Protection Program (PPP), which mainly versus external flexibility in the Covid‑19 entailed forgivable loans to SME, and the Employee pandemic Retention Tax Credits. However, as Autor et  al. (2022) To gain a better understanding of the impact of the differ - show, 66 to 77% of the issued loans in 2020 do not seem ent policy strategies in Germany and the United States on to actually have been used to retain their employees but the labour market, we conduct a business-cycle analysis were kept by business owners and shareholders. Both comparing, on the one hand, the Coronavirus Recession measures, the PPP and the Employee Retention Tax with the Great Recession in both countries and contrast- Credits were amended in subsequent laws later in 2020. ing the German experience with that of the United States Overall, these two measures of business support from the on the other hand. Then, we look at the case of Germany CARES Act account for $861 billion  or 4.1% of nominal in the Coronavirus Recession with particular interest in GDP in 2020 (CRFB 2021). the relative importance of working-time instruments (overtime, regular working time, working time accounts (WTA), and STW) in safeguarding employment, before This view is also documented in Giupponi et al. (2022). 11 Page 6 of 22 A. Herzog‑Stein et al. Table 2 Dating the Great Recession and the Coronavirus Recession in Germany and the United States Contractions Germany United States Peak quarter Trough quarter Peak quarter Trough quarter Great Recession 2008 Q1 2009 Q2 2007 Q4 2009 Q2 Coronavirus Recession 2019 Q4 Not yet determined 2019 Q4 2020 Q2 Sources: German Council of Economic Experts (GCEE); NBER’s Business Cycle Dating Committee examining the role of external flexibility in the United on cyclical variations in economic activity in the follow- States via temporary and permanent lay-offs during the ing business-cycle analysis. Therefore, we extract the Covid-19 pandemic. cyclical and trend component using the Hodrick-Prescott Filter (HP-Filter). As is common practice for quarterly 3.1 G ermany and the United States: a business cycle data, we use a HP-Filter with a smoothing parameter analysis equal to 1600 to detrend the quarterly time series from For a business-cycle analysis of the cyclical variations in 1991 to 2021. It is well known that the HP-Filter, like economic activity, employment, productivity, and work- other filter methods, suffers from an end-point problem. ing hours in these two countries, we must first determine Since our main focus is on the slump until 2020q2 and the peak and trough of the Great Recession and the latest we use six additional data points, the impact of this end- recession in Germany and the United States. We follow point problem is still there but of a smaller importance the business cycle dating of the NBER’s Business Cycle for the analysis of the recession periods. However, we are Dating Committee for the United States and of the Ger- careful in interpreting results after 2020q2 and closer to man Council of Economic Experts (GCEE) for Germany. the end of the data set. For completeness and clarity, we In Germany the economic downturn of the Great present results up to 2021q4, the end of our dataset. Recession started after the first quarter of 2008 (peak) For Germany and the United States, Fig.  3 examines and ended in the second quarter 2009 (trough). Accord- the economic dynamics of the cyclical components of ing to the NBER, in the United States it started already real  GDP, employment, productivity and working time, after the fourth quarter 2007 (peak) and also ended in the i.e., average hours worked per employee, during the second quarter 2009 (Table 2). As for the latest downturn, Great Recession (Germany: Panel A, and United States: the Coronavirus Recession, in both countries economic Panel C) and the Coronavirus Recession (Germany: Panel activity peaked in the fourth quarter 2019. The NBER B, and United States: Panel D). All figures are normalised dated the trough of economic activity to the second quar- to the respective beginning of the two economic down- ter 2020. For Germany, the GCEE has not yet determined turns, i.e., for the Great Recession 2008q1 for Germany the trough of economic activity. However, the develop- and 2007q4 for the United States, and for the Coronavi- ment of economic activity in Germany in 2020 and 2021, rus Recession 2019q4 for both economies. especially GDP growth, also points towards the second Due to the economic shock caused by the Covid-19 quarter 2020 as time of the economic trough. Therefore, pandemic, the Coronavirus Recession was much more in the further analysis we assume that 2020q2 is also the severe. From 2019q4 to 2020q2, cyclical real  GDP con- trough of economic activity in Germany, but present data tracted by 12.3% in Germany and by 11.4% in the United for both countries until the end of 2021. States as a direct consequence of the Covid-19 pan- Given the determination of the Great Recession and demic. In the Great Recession, the corresponding cyclical the Coronavirus Recession in Germany and the United decline in output from peak to trough was 8.1 and 5.1%, States by the GCEE and the NBER, respectively, we focus respectively. As in the Great Recession, most of the economic shock Throughout this business -cycle analysis whenever we use the logarithmic in Germany was absorbed by internal flexibility in the transformation of a time series x, growth rates are approximated by differ - labour market via a temporary working-time reduction ences in log points, i.e., the growth rate of x in per cent is g ≈ lnx × 100. Growth of 1 log point is equal to a growth rate of approximately 1 per cent. and labour hoarding in the form of a procyclical decline The exact relationship is For simplicity we always g = 100 ∗ (exp(�lnx) − 1). in labour productivity. However, this time the relative speak of percentage changes. contribution of internal flexibility was even larger than Both institutions determine the peaks and troughs of economic activity in the Great Recession. From peak to trough, the cycli- separately on a monthly and a quarterly basis. In general, the peak or trough quarter contain the peak or trough month. However, the determination of cal reduction in the average number of hours worked the peak of economic activity before the Coronavirus Recession is an excep- per employee was twice as high as in the Great Reces- tion. Both, NBER and GCEE determine February 2020 as the latest peak sion (− 8.8 vs. − 3.4%). In Germany, productivity reacted month but the fourth quarter 2019 as the peak quarter. Germany and the United States in coronavirus distress: internal versus external labour market… Page 7 of 22 11 Fig. 3 Great Recession vs. Coronavirus Recession in Germany and the United States. Log deviations from peak quarter (Germany: 2008q1 respectively 2019q4; United States: 2007q4 respectively 2019q4) measured in log points. Output (= real GDP), employment, working time in hours worked per quarter per employee, and productivity (= labour productivity per hours worked) are seasonally and/or calendar adjusted. Sources: Federal Statistical Office (Destatis); Bureau of Economic Analysis; U.S. Bureau of Labor Statistics (BLS); own calculations 11 Page 8 of 22 A. Herzog‑Stein et al. much stronger in the Great Recession than in the Coro- to changes in the economic environment, such as over- navirus Recession (− 5.6 vs. − 2.1%). Even though speed time, working-time accounts, temporary changes in reg- and intensity of job losses were more pronounced in the ular working time and STW. Figure 4 therefore shows the Coronavirus Recession, in both economic recessions development of cyclical working time per employee and cyclical employment continued to decline, even after its components regular working time, paid and unpaid the trough of the business cycle. Overall, employment overtime, STW, as well as WTA, again detrended with declined cyclically by 0.6% from 2008q1 to 2009q2; by the HP-filter (  = 1600) if the component has a trend. 2010q1 it had fallen by a further 0.7%. Thereafter, cycli - Over the period from the beginning of 2005 to the end of cal employment started to recover. In the Coronavirus 2020, working time and all its components follow a clear Recession, cyclical employment declined by 1.4% from cyclical pattern. However, while all these components 2019q4 to 2020q2 and a further 0.4% by 2021q1. It then contributed to the safeguarding of employment during started to recover over the remaining quarters of 2021. the financial crisis (Herzog-Stein et  al. 2018), this is no In contrast to the economic developments observed in longer the case in the Coronavirus Recession. Germany, in the United States external flexibility bore the brunt of adjustment in response to the economic shock 3.2.1 Shor t‑time work (STW) as a consequence of the Covid-19 pandemic (Panel D). In terms of safeguarding jobs, STW has two important From peak to trough, cyclical employment decreased by dimensions: the number of workers in STW and the 13%. However, in contrast to the cyclical development in intensity of STW, i.e., the number of reduced working Germany, employment then started to recover. This also hours per short-time worker due to STW. Comparing contrasts the cyclical behaviour of employment in the the development of STW in both recessions, two aspects Great Recession when it continued to decline beyond the stand out particularly. First, policy makers reacted fast trough quarter. and made the use of STW more attractive for establish- Average hours worked per employee decreased by 1.0% ments immediately after the outbreak of the Covid-19 in the Coronavirus Recession, while labour productivity pandemic at the end of the first quarter 2020. This had cyclically increased by 2.1% from peak to trough. This the effect of introducing STW on a uniquely large scale in also contrasts with developments in the United States both dimensions of STW. In April 2020, the month with during the Great Recession, when internal flexibility the highest incidence of STW in the Coronavirus Reces- from cyclical reductions in working hours per worker sion, almost 6 million, or 17.9% of all employees subject and changes in labour productivity together accounted to social security contributions were in STW. The average for about a quarter (− 1.2%) of the labour market adjust- loss of working time for a short-time worker was nearly ment relative to the cyclical decline in real  GDP (Panel 50%. In employment equivalents this corresponded to C). In the Great Recession, from peak to trough working 9.1% of all employees subject to social security contribu- time per employee decreased cyclically by 1.7%, a larger tions (Fig. 5). decline than during the Coronavirus Recession. But in Although the number of employees in STW declined the latest contraction the speed of the working-time steadily after April 2020, there were still more employees reduction was faster than in the Great Recession. How- in STW in October 2020 than at the peak of the Great ever, the major difference in the cyclical labour-market Recession. As a result of the second wave of the Covid- responses between the two recessions lies in the develop- 19 pandemic, the number of workers in STW rose again ment of cyclical labour productivity in the United States. from November 2020 and peaked in February 2021 Cyclical labour productivity behaved slightly pro- to acy- before declining again. clical in the Great Recession and anticyclical in the Coro- Consequently, in the Coronavirus Recession there navirus Recession. was a rapid cyclical reduction in average working time Overall, this section has shown that the German and per worker of 2.4  h already in 2020q1 alone (relative to US labour market reacted quite differently during the 2019q4). This is comparable in its magnitude to the cycli - last two recessions. In Germany, internal flexibility domi - cal working-time reduction induced by the use of STW nated labour market adjustment, while in the United from peak to trough in the whole Great Recession of States it was external flexibility. This finding fits with the 3.3 h per worker—of which 3.1 h were reduced in the first descriptions of the policy responses outlined above. two quarters of 2009 relative to the last quarter in 2008. Second, while the immediate response in STW was 3.2 G ermany: internal flexibility already comparable to the Great Recession, at the trough In Germany, several instruments of internal flexibility are of the Coronavirus Recession in the second quarter 2020, available at the establishment level to temporarily adjust STW reduced the average working time per worker by the number of hours worked per employee in response 18.4 h compared to the peak quarter 2019q4, more than Germany and the United States in coronavirus distress: internal versus external labour market… Page 9 of 22 11 Fig. 4 Components of cyclical changes in working hours per employee per quarter (2005–2021). The term ‘cyclical’ refers to the difference of actual and trend changes for each working‑time instrument (if the series shows a trend). STW and WTA show no trend. The trend is constructed applying the Hodrick‑Prescott filter with  = 1600 . All components are measured in working hours per employee per quarter. Sources: Institute for Employment Research (IAB) working time calculations; own calculations Fig. 5 Short‑time work and employment equivalents (2008–2020). Proportion of short ‑time workers (realised numbers or employment equivalents) in total employment subject to social security contributions. Sources: Federal Employment Agency; own presentation 11 Page 10 of 22 A. Herzog‑Stein et al. Fig. 6 Share of recipients of short‑time allowance, average working time reduction, and employment change by economic sector. B: Mining and quarrying; C: Manufacturing; D: Electricity, gas, steam and air conditioning supply; E: Water supply; sewerage, waste management and remediation activities; F: Construction; G: Wholesale and retail trade; repair of motor vehicles and motorcycles; H: Transportation and storage; I: Accommodation and food service activities; J: Information and communication; K: Financial and insurance activities; L: Real estate activities; M: Professional, scientific and technical activities; N: Administrative and support service activities; O: Public administration and defence; compulsory social security; P: Education; Q: Human health and social work activities; R: Arts, entertainment and recreation; S: Other service activities. Short‑time work refers to the share of recipients of short‑time allowance by economic sector in May 2009 and April 2020 respectively (columns). The intensity of STW refers to the average reduction in working time of a short‑time worker (in %) due to STW (dots) and is calculated by dividing the employment equivalent by short‑time workers. Change in employment (diamonds) refers to the sum of employment subject to social security contributions (seasonally adjusted) and exclusively marginally paid employees by economic sector from March to April 2020. Data on marginally paid employees by sectors are only available since 2020. Hence no seasonally adjusted data are available. Given that employment is not provided in each economic sector, employment changes of the sectors B, D, E, L, M, O, U, R, S, T are approximated by the corresponding average employment changes by the sums of B + D + E, L + M, O + U, R + S + T. Sources: Federal Employment Agency; own calculations five times the working-time reduction due to the use April 2020 was nearly twice as high as in May 2009. The of STW in the Great Recession. On average, STW is average intensity of STW use was particularly high in the accounting for around 89% of the total reduction in hours services sector, exceeding 70% in sections ‘Accommoda- worked per worker from 2019q4 to 2020q2. tion and food service activities’ (I), ‘Arts, entertainment In the two recessions, employees subject to social and recreation’ (R), and ‘Other service activities’ (S). In security contributions were affected differently by STW the past, STW intensity of 100% was not common. In the in the individual economic sections (Fig.  6). A compari- Coronavirus Recession it was used only modestly, despite son between May 2009 and April 2020, the months with the severity of the crisis. According to Kruppe and Osi- the highest incidence of short-time work in both down- ander (2020) using information on the individual STW turns, shows that this time not only was the number of intensity from a survey in May 2020, 24.1% of all STW- short-time workers significantly higher, but in the econ - workers reported a loss in hours of 100%, but still more omy as a whole STW was used more heavily (columns in than half a loss in hours of only up to 50%. Fig. 6). While more than 80% of short-time workers were employed in manufacturing during the Great Recession, With respect to employment subject to social security contributions the it was only about 31% during the Coronavirus Recession. employment structure of the German economy has changed only moderately since the Great Recession. The employment share in manufacturing (section In the Coronavirus Recession, STW is also used more C) has decreased from 23.3 to 20.8%. In turn, the employment share in the intensively across all economic sections (dots in Fig.  6). services sector (sections G-N) has risen. In the total economy, the intensity of the use of STW in Germany and the United States in coronavirus distress: internal versus external labour market… Page 11 of 22 11 In contrast to the importance of internal flexibility and time from 2019q4 to 2020q2 on a quarterly basis (− 0.3 h especially the use of STW, external flexibility—unlike in vs −  0.2  h per quarter) is similar to that in the Great the United States (see Sect.  3.2.1)—hardly played a role Recession (− 0.2 h and − 0.2 h per quarter), but together in Germany between March and April 2020. The overall accounting only for less than 5% of the total working- change in employment, measured by the sum of employ- time reduction per worker during that time period, in ees subject to social security contributions and workers contrast to nearly 20% in the Great Recession. only marginal employed, was only about − 1% on average (diamonds in Fig.  6). Only in section H (Transportation & Storage) there is a substantial drop in employment of 3.2.4 R egular working time − 7.5%. Unlike in the Great Recession, there is not really a cycli- cal response in regular working time to reduce working hours in the Coronavirus Recession. The cyclical com - 3.2.2 Working‑time accounts (WTA) ponent of regular working time per worker even slightly Together with STW, they were the most important increased average working hours per worker by on aver- instrument of internal flexibility during the Great Reces - age 0.3  h from 2019q4 to 2020q2. Overall, this observa- sion. Like STW, the use of WTA at that time reduced tion might be explained by the dominance of STW, which the average working time per employee by 3.3  h in total made further adjustments to working time unnecessary. or 0.7  h per quarter from peak to trough. In the Coro- navirus Recession, the contribution of WTA to the tem- 3.2.5 Summary porary reduction in average hours worked per worker In conclusion, although external flexibility again was of is this time much smaller than in the Great Recession. minor importance and instruments of internal flexibility From peak to trough, WTA contributed 1.7 h, or on aver- played a crucial role in the safeguarding of employment age 0.8  h per quarter, to the reduction in average hours in both the Great Recession and the Coronavirus Reces- worked per worker in the latest downturn. sion in Germany, a closer look at various working-time At first glance, this is unexpected, as WTA became components shows marked differences between the two more common over time and 56% of all employees had recessions. While in the Great Recession several instru- WTA in 2016 (Ellguth et al. 2018). However, one possible ments contributed markedly to the temporary decline in explanation could be the respective economic dynamics hours worked per worker, in the Coronavirus Recession in the boom periods before the two recessions. STW is the instrument that has contributed by far the In the upswing before the Great Recession, WTA were most to the reduction in working hours (Fig. 7). filled, providing firms with a considerable working-time- In the Great Recession, STW and WTA contributed account buffer for the following downturn. In contrast, in equally to the cyclical reduction in working time from the long boom period before the Coronavirus Recession, peak to trough (− 3.3 h each). Paid and unpaid overtime working time was closer to its long run trend with smaller and a temporary reduction in regular working hours cyclical variations. As a result, opportunities to increase both reduced cyclical working time by an additional two the balances in the WTA were more limited than in the hours. In contrast, while most instruments responded as boom period before the Great Recession. Therefore, the expected in the latest downturn, in absolute and in rela- working-time reductions due to WTA account only for tive terms STW was by far the main driver to safeguard 8% of total working-time reduction in the latest recession employment in the Coronavirus Recession (−  18.4  h). from 2019q4 to 2020q2. WTA was again the second most important instru- ment of internal flexibility used. However, its quantita - 3.2.3 Overtime tive importance was smaller, reducing average working In general, paid and unpaid cyclical overtime vary hours per employee by 1.7  h. The same is true for paid between ± 1  h per quarter over the business cycle. and unpaid overtime which together reduced average Unpaid overtime was most important at the beginning of working hours by another 0.9  h. Reductions in regular the considered period (Fig. 4). After the minor economic working time do not contribute to the cyclical reduc- slowdown in Germany related to the so-called Euro Cri- tion in working time in the Coronavirus Recession. The sis from 2011q3 to 2013q1, it lost its relevance for cycli- observed dominance of STW in the attempt to safeguard cal fluctuations. Interestingly, unlike unpaid overtime, employment in the Coronavirus Recession is in line with the cyclical variation of paid overtime continues after the the made discretionary policy changes governing the use Great Recession and can also be observed in the Covid- of STW. It is conceivable that the extended and simpli- 19 pandemic. fied use of short-time work “crowded out” to some extent In the Coronavirus Recession, the contributions of paid the use of other measures like e.g., WTA since already in and unpaid overtime to the cyclical reduction in working 11 Page 12 of 22 A. Herzog‑Stein et al. Fig. 7 Contributions to the cyclical working‑time reductions in the Great Recession and the Coronavirus Recession. The term ‘cyclical’ refers to the difference of actual and trend changes for each working‑time instrument (if the series shows a trend). STW and WTA show no trend. The trend is constructed applying the Hodrick‑Prescott filter with  = 1600 . All components are measured in working hours per employee. Sources: Institute for Employment Research (IAB) working time calculations; own calculations March 2020 no negative balances on WTA were required employers (Gallant et  al. 2020), was the prominent tool anymore as eligibility criteria for the use of STW. for dealing with the crisis. Finally, the important impact of the use of STW on While temporary unemployment has been between unemployment respectively employment is best seen 0.4 and 1.2% throughout the years and even during the by looking at the seasonally adjusted inflow rate from economic and financial crisis it played no prominent role employment into unemployment and the exit rate from with respect to the overall increase in unemployment, the unemployment into employment on a monthly basis share of workers on temporary layoffs jumped to 11.5% in (Fig. 8). From February to April 2020 the inflow rate from April 2020, accounting for almost 80% of all unemployed employment into unemployment increased from 0.5 persons (Fig.  9). While unemployment declines slowly to 0.7% and declined than quickly back to 0.5% in June, during an economic recovery, the work-finding rate for while the exit rate from unemployment into employment the temporarily laid-off unemployed is usually twice as decreased from 8.3 to 4.7% from February to May 2020 high as for the unemployed. Accordingly, unemployment and did not reach its pre-pandemic level until the end of fell faster this time than in previous recoveries (Hall and 2021. This can be seen as some indication that the mas - Kudlyak 2022). Thus, the rate of temporary unemploy - sive use of STW was able to prevent large and prolonged ment halved from April to July 2020, while the jobless flows from existing employment into unemployment. unemployment rate increased by 1.2 percentage points. External flexibility via temporary lay-offs was the 3.3 Unit ed States: external flexibility main means of the labour-market adjustment in the Although STW programs exist in about half of the U.S. states and STW utilisation was much higher than in the past, the use of STW has overall not played a major role The job-finding rate is often used to describe the labour market tightness. in the United States (Krolikowski and Weixel 2020). Here, Hall and Kudlyak (2022) use the work-finding rate instead in the context of the focus was rather on external flexibility. However, for the Covid-19 pandemic given that unemployed on temporary layoffs are not obliged to look for a job while waiting to be recalled. Some of them might the first time, the use of temporary lay-offs, i.e. laid-off nevertheless take a new job rather than waiting to be recalled by the previous individuals who expect to be recalled by their former employer. Germany and the United States in coronavirus distress: internal versus external labour market… Page 13 of 22 11 Fig. 8 Monthly unemployment flows (2009 to 2021). The monthly inflow rate from employment into unemployment (national definition) is defined as the number of inflows from employment to unemployment in month t relative to the employment level in month t ‑1. The monthly exit rate from unemployment into employment is defined as the number of outflows from unemployment (national definition) to employment in month t relative to the unemployment level in month t‑1. The numbers are seasonally adjusted. Sources: Federal Employment Agency; own calculations Fig. 9 US unemployment rate: temporary layoffs and jobless unemployed. Jobless unemployed comprises job losers not on layoff, job leavers, reentrants to labour force and new entrants to labour force. Sources: U.S. Bureau of Labor Statistics (BLS), own calculations 11 Page 14 of 22 A. Herzog‑Stein et al. Fig. 10 Proportion of persons in the United States unable to work due to lost business in the coronavirus pandemic. Supplemental data measuring the effects of the Coronavirus (COVID ‑19) Pandemic on the labour market. Persons unable to work at some point in the last 4 weeks because their employer closed or lost business due to the Coronavirus pandemic by receipt of pay from their employer for hours not worked and employment status. Sources: U.S. Bureau of Labor Statistics (BLS) Release “Eec ff ts of the Coronavirus (COVID ‑19) Pandemic on the Labor Market”, https:// www. bls. gov/ cps/ effec ts‑ of‑ the‑ coron avirus‑ covid‑ 19‑ pande mic. htm; own calculations Coronavirus Recession, but complementary to this, com- The majority of them reported that they were not com - panies also used some measures of internal flexibility by pensated by their employers. Only less than a quarter reducing the working hours of their employees. While received some compensation. Therefore, the reduction unlike in Germany there are no detailed information on respectively loss in working hours in the United States the average number of working hours lost per worker due took place in a way that is quite different from the short- to the Covid-19 pandemic in the United States, an addi- time allowance in Germany. tional survey conducted by the BLS beginning in May Overall, these information on unpaid as well as 2020 as part of the Current Population Survey provides on compensated temporary working time losses fit a good insight into the extent to which workers in non- together with the macroeconomic evidence for the agricultural industry were affected by the crisis. In this United States presented in Sect.  3.1. As shown above survey, workers were asked whether they had been una- there was some cyclical reduction in the average work- ble to work due to the pandemic in the last four weeks ing time per employee from peak to trough of 1.0% in and whether they had received any kind of payment the Coronavirus Recession. If we take into account that from their employers. Unfortunately, no information was the Coronavirus Recession was much shorter than the asked about the form of compensation paid or the exact Great Recession, the average individual working-time number of hours lost as a consequence of the pandemic. reduction per quarter was stronger this time. Given u Th s, it is also not known whether workers who were that “job-losses have disproportionally hit the low-wage compensated for working time lost due to the pandemic workforce” (Bateman and Ross 2021) it is also likely received any payments via one of the short-time work that the reported working-time losses were concen- programs at the state level. trated among the low-skilled. Since low-skilled workers In May 2020, 20% of workers reported that they were generally have a lower hourly labour productivity, this affected by some kind of loss of working time, and in Given that the survey asks about the employment situation in the previous June and July 2020 still more than 15 respectively 10% of four weeks, the May survey probably covers most of the employment situation workers experienced some loss of working time (Fig. 10). in April 2020, the month with the highest crisis impact. Germany and the United States in coronavirus distress: internal versus external labour market… Page 15 of 22 11 -10 -20 -30 -40 -50 Change in employment Unable to work Unable to work (received pay) (did not receive pay) Fig. 11 Proportion of persons in the United States unable to work (with and without compensation) due to lost business and change in employment by industry in April 2020. NAICS classification. 21 = Mining, quarrying, and oil and gas extraction, 23 = Construction, 31 = Durable goods manufacturing, 32–33 = Nondurable goods manufacturing, 42 = Wholesale trade, 44–45 = Retail trade,48–49 = Transpor tation and warehousing, 22 = Utilities, 51 = Information, 52 = Financial Activities, 54 = Professional & Business Services, 61–62 = Education & Health Services, 71–72 = Leisure & Hospitality, 81 = Other Services, 92 = Public administration, N.I. = Nonagricultural industries. Proportion of persons unable to work are from May 2020 which refers to the previous four weeks. Data for the change in employment refers to the monthly change from March to April 2020. Sources: U.S. Bureau of Labor Statistics (BLS) Release “Eec ff ts of the Coronavirus (COVID ‑19) Pandemic on the Labor Market”, https:// www. bls. gov/ cps/ effec ts‑ of‑ the‑ coron avirus‑ covid‑ 19‑ pande mic. htm, Current Employment Statistics (CES), own calculations would explain the anticyclical increase in labour pro- imply that firms do not use measures of internal flex - ductivity observed in the Coronavirus Recession (see ibility, too. The information from the economic sectors Fig. 4D). indicates that economic sectors that were hit hard by the More details about the American way of dealing with Covid-19 pandemic relied on external as well as on inter- the Coronavirus Crisis are revealed by looking closer nal flexibility in response to the Coronavirus Recession. at the change in employment and the share of work- There is a strong positive correlation between layoffs ers affected by working hours lost due to the inability to and working-time reductions with a correlation coeffi - work in different economic sections of the US economy cient of 0.9: in economic sectors with a larger reduction (Fig.  11). Given that for economic sectors no data for in employment there is also a larger share of workers temporary layoffs are available the change in employ - with a loss of working hours due to the inability to work. ment is used instead to indicate the intensity with which However, this positive relationship is driven by the posi- employers were hit by job losses across economic sectors. tive correlation between employment reductions and As in Germany, the economic sections have been affected working-time losses without compensation; there is no differently by the Covid-19 pandemic, the service sectors correlation between the magnitude of employment losses more than the industry. and the size of the share of workers with renumerated Interestingly, a combination of layoffs and reductions working-time losses. This suggests that, in contrast to the in working hours dominates in all sectors of the econ- experience in Germany, in the United States the burden omy. Hence, the dominance of external flexibility in the of labour market flexibility in the Coronavirus Recession labour market adjustment in the United States does not is borne primarily by workers. 32-33 44-45 48-49 61-62 71-72 N.I. 11 Page 16 of 22 A. Herzog‑Stein et al. Fig. 12 Contributions (in percentage points) of the components of US unemployment in the Great Recession and the Coronavirus Recession. Jobless unemployed comprises job losers not on layoff, job leavers, reentrants to labour force and new entrants to labour force. Sources: U.S. Bureau of Labor Statistics (BLS); own calculations In conclusion, the United States have relied again heav- points were due to temporary layoffs and only 0.7 per - ily on the use of external flexibility. However, there are centage points to workers who lost their jobs. Interest- also major differences in its response compared to the ingly, fewer employees seemed to leave their job of their Great Recession as a breakdown of the change in the own accord, and no change in labour force entry was unemployment rate in the two crises reveals (Fig. 12). observable. Most of the change in the unemployment rate is deter- mined by the number of employees losing their jobs, usu- 4 Blind spots of the chosen strategy ally without being on recall. During the Great Recession, Germany and the United States pursued different goals the unemployment rate rose by 4.5 percentage points. with their Coronavirus Crisis responses. Germany Only a small part of 0.5 percentage points was due to focused on employment protection via mechanisms temporary layoffs. Another 3.1 percentage points of the designed to promote internal flexibility. The United increase in the unemployment rate was due to workers States focused on a mixture of external flexibility and losing their jobs. The proportion of job leavers among the income protection. As strategy objectives were differ - unemployed is hardly influenced by the business cycle ent from one another, the blind spots are likely to differ and lies typically in a range between 0.5 and 0.6%. Hence, between the countries. Therefore, we discuss the chal - its contribution to the change in unemployment over lenges each strategy poses in this last section. time is negligible. Almost one percentage point of the increase was due to re-entrants and new entrants into the 4.1 Germany’s employment protection labour market during the Great Recession. For the success of Germany’s chosen strategy, the main As shown in the analysis above the labour market goal was to secure existing employment relationships to response during the Coronavirus Recession was extraor- save firm specific human capital, prevent unemployment, dinary and very different to the one observed in the Great and to reduce future training costs. Recession. For the first time, temporary layoffs played a The burden of job losses in the Coronavirus Reces - prominent and dominant role in the United States. The sion was unevenly distributed in Germany. Job loss unemployment rate rose by a total of 9.4 percentage rates of employment subject to social security contribu- points from peak to through, of which 8.8 percentage tions, which enjoys the protection of the STW scheme, Germany and the United States in coronavirus distress: internal versus external labour market… Page 17 of 22 11 were less severe than the losses of marginal employment employment as well as self-employment are not pro- (Minijobs). tected by the STW scheme. Employment subject to social security contributions, the backbone of the German welfare state, decreased by 4.2 Differences in income protection in Germany around 450,000 jobs or 1.3% between March and May One advantage of STW besides securing existing employ- 2020 (see Fig. 13A). In line with their growth trends dur- ment relationships and firm specific human capital is that ing the long boom before the outbreak of the Coronavi- the income decrease during STW is less severe than with rus Crisis, the recovery of employment subject to social an immediate fall back to unemployment benefits. How - security contributions was more dynamic than that of ever, since income replacement rates of STW are compa- total employment in the summer and fall 2020. rable with the initial replacement rates of unemployment Workers in marginal employment (Minijobs), who benefits, the impact of STW on income is dependent on overwhelmingly work in the services sector were severely the previous wage and the reduction in working time. hit by the economic crisis due to the Covid-19 pan- This is also why Germany’s focus on employment protec - demic. In the months March to May 2020 the percent- tion via STW also helped with income protection. The age decrease in marginal employment (−  7.5%) was five widespread use of STW not only safeguarded employ- times as large as in employment subject to social security ment, but also secured part of household income for contributions (see Fig. 13B). households whose members were affected by STW. Even after accounting for the different growth trends One challenge of STW during the Coronavirus Reces- of these two employment forms, marginal employment sion was that the average short-time worker was very dif- was more severely hit by the Coronavirus Crisis. Fur- ferent from the average short-time worker in the Great thermore, while both employment subject to social secu- Recession. The massive use of STW in other economic rity contributions and marginal employment started to sections than manufacturing as well as the more intensive recover in the summer months, marginal employment use of STW in general during the Coronavirus Recession declined again with the second wave of the pandemic. have immediate income effects. There are two obvious reasons for these remark - In Fig. 15 the average income losses due to STW in each able employment patterns during the current Corona- economic section are plotted for the Great Recession and virus Recession, which are interlinked. First, this time the Coronavirus Recession. Interestingly, whereas in the the services sector was much more hit by the economic Great Recession there was no clear negative correlation crisis than during the Great Recession. Furthermore, in between the level of earnings and the percentage earn- the Coronavirus Crisis the necessity to temporarily lock ings loss due to STW (blue dots), the situation during down and interrupt parts of economic activity to prevent the Coronavirus Recession is completely different. We the spread of infection cannot be overcome by stimulat- observe a clear negative correlation with a correlation ing aggregate demand. Rather, economic policies must coefficient of −  0.7 (orange dots). The lower the average try to sustain businesses, and hence employment, during earnings are in an economic section the higher is the per- these periods of (partial) lockdown and interruption of centage income loss due to STW. The difference is likely production processes. to be even more pronounced if we were to consider addi- Second, STW, the major pillar of the government’s tional supplements of the short-time work allowance due strategy to safeguard employment is not applicable to to the employer which is more often paid in jobs with marginal employment. This left more jobs unprotected higher earnings (Pusch and Seifert 2020, Table 3). Pusch in the services sector, where marginal employment con- and Seifert (2020, Table  3) present the share of employ- stitutes a larger share of total employment. In services ees who receive a supplement to STW allowance. There sectors like accommodation and food service activities is a positive correlation: The higher the average earnings (section  I), or arts, entertainment and recreation (sec- in an economic sector, the higher the share of employees tion  R), more than 40% of all employees were either who receive a supplementary STW allowance from their working in marginal employment as their only or as their employers. This implies that while the initial income loss second job (Fig. 14). was strongest in sectors with the lowest incomes, addi- Overall, since the services sectors were more severely tional supplements to STW were concentrated among affected by the Coronavirus Recession than by the Great sectors with the highest incomes. Recession, some weaknesses in the approach to safeguard As pointed out before, not only short-time work- employment became visible. In contrast to employ- ers suffered income losses due to the loss of work. In ment subject to social security contributions, marginal general, all types of employment, who lost (temporary) part or all of their work, or even became unemployed, Marginal employment exhibits a marked negative growth trend since the suffered income losses. Groups like the self-employed introduction of the general legal minimum wage at the beginning of 2015. 11 Page 18 of 22 A. Herzog‑Stein et al. (A)EMPLOYMENT SUBJECT TO SOCIAL SECURITY CONTRIBUTIONS (B) MARGINAL EMPLOYMENT Fig. 13 Different types of employment (2008–2021). Level (line, left scale) and change (columns, right scale) measured in 1000 persons (seasonally adjusted). Sources: Federal Employment Agency; Bundesbank; own presentation Germany and the United States in coronavirus distress: internal versus external labour market… Page 19 of 22 11 Fig. 14 Composition of employment according to economic sectors (March 2020). B: Mining and quarrying; C: Manufacturing; D: Electricity, gas, steam and air conditioning supply; E: Water supply; sewerage, waste management and remediation activities; F: Construction; G: Wholesale and retail trade; repair of motor vehicles and motorcycles; H: Transportation and storage; I: Accommodation and food service activities; J: Information and communication; K: Financial and insurance activities; L: Real estate activities; M: Professional, scientific and technical activities; N: Administrative and support service activities; O: Public administration and defence; compulsory social security; P: Education; Q: Human health and social work activities; R: Arts, entertainment and recreation; S: Other service activities. Sources: Federal Employment Agency, own calculations or workers in marginal employment were not entitled and the focus is on re-employment, this approach has to STW or unemployment benefits. Walwei (2021) the potential to target individual’s particular in need of stresses that, in contrast to the previous crisis, employ- income stabilization. ment associated with weak income security (marginal Temporary layoffs were the dominant component employment and solo self-employment) were particu- of the unemployment increase (Fig.  9). However, larly hard hit in the Coronavirus Crisis. they were rather unevenly distributed, and concen- To conclude, STW did not protect employment and trated among high wage employees. While there was income of everyone equally. This is why further discre - a rebound of employment for high wage earners, there tional measures were taken in a later stage of the Cor- were persistent job losses among low wage earners onavirus Crisis to reduce the blind spots of the STW (Cajner et al. 2020; Chetty et al. 2020). instrument. These additional discretionary policies as As discussed in Sect. 2, the CARES Act covered vari- in example the one-off payments for children and an ous instruments to secure income of individuals. Three increase in STW replacement rates for long term recip- policies to stabilize income directly or indirectly are ients helped to cushion the income loss (Christl et  al. of particular interest: The PPP, the Federal Pandemic 2021). Unemployment Compensation and Stimulus Checks. While the initial idea of the PPP was to stabilise the 4.3 D ifferences in income protection in the United States income of individuals otherwise losing their jobs, stabilise There are two objectives for the success of the United companies’ financial flows, and support business owners States’ chosen strategy. First, the job loss should be through direct support for small business, the actual ben- temporary and not permanent. Second, the income of efits had little to do with employment protection. Evalu - individuals should be protected despite the temporary ations overall indicate that the program was untargeted job loss. As long as unemployment is only temporary and inefficient (Chetty et al. 2020; Autor et al. 2022). Even 11 Page 20 of 22 A. Herzog‑Stein et al. Q F Q J 0500 1000 1500 2000 2500 3000 Average gross monthly earnings Fig. 15 Earnings loss due to short‑time work by economic sections. Blue (orange) dots indicate average income losses in the Great Recession (Coronavirus Recession) in %. For the comparison we use the information about the average amount of working time lost due to STW in the two crises which is provided by the Federal Employment Agency (Employment equivalent/number of short‑time workers). Given that STW is paid on the basis of net earnings losses we calculate the impact of STW on average gross monthly earnings (SOEP) in Euro by using the Kurzarbeiterrechner on the assumption that the short‑time worker is single, in tax class 1, without children. Sources: Federal Employment Agency, SOEP, Kurzarbeiterrechner (https:// www. netto lohn. de/ rechn er/ kurza rbeit ergeld. html), own calculations when it safeguarded employment, this was concentrated Germany, in contrast, labour hoarding was encouraged among employees in the top income quintile (Autor et al. through STW programmes, which maintain employment 2022). However, as pointed out by Autor et al. (2022) this relations between workers and firms. As a result, in the mainly results from the trade-off between a timely ver - United States the unemployment rate (which is largely sus a targeted intervention under limited administrative driven by temporary layoffs) and the non-employment capacities. rate rose sharply, as have the take up rates of STW in In contrast, the Stimulus Checks and the Federal Pan- Germany. demic Unemployment Compensation proved to be an Put differently, initial income support was stronger in efficient tool to secure income. While large earning the United States, but despite high levels of temporary declines were more likely for low wage workers, both layoffs, lead to a higher share of permanent job losses instruments together outweighed the otherwise result- (Barrero et  al. 2021), which in turn required more job ing income losses (Larrimore et  al. 2022) and stabilised search and reallocation activities. This was particu - income particularly at the lower end of the income distri- larly true for lower wage earners. It remains to be seen bution (Autor et al. 2022). Ganong et al. (2020) show that whether these reallocation effects will be beneficial in the the majority of workers eligible for unemployment ben- succeeding recovery. Germany, in contrast, secured exist- efits between April and July 2020 had replacement rates ing employment relations at the expense of temporary above 100%. This lead to only a modest increase in pov - income losses without protecting workers in marginal erty rates until the unemployment supplements expired employment. While unemployment also leads to a per- (Parolin et al. 2020). sistent decline in wages, Giupponi et al. (2022) point out that Germany’s previous experience with STW indicates 4.4 Comparison of the strategies that wages of secured workers adjust to its pre-crisis The United States decided to insure workers’ incomes level. The German approach therefore allowed to secure against the costs of job losses by increasing the generos- employment relations and human capital in the short run ity, eligibility criteria and eligibility period of unemploy- and secured income in the long run. ment benefits and other income support measures. In Average income loss due to STW Germany and the United States in coronavirus distress: internal versus external labour market… Page 21 of 22 11 Overall, both countries protect individuals in weaker However, a closer look at the blind spots of the positions on the labour market (atypically employed and chosen strategies in Germany and the United States low wage earners) to a smaller extent through the chosen showed that despite the differences in the respective strategies. This is an indication of the segmentation of the approaches, people in weaker labour market positions labour market by wages and working conditions (Reich were less well protected by the chosen strategies. In et al. 1973), which also reflect their lack of protection by Germany, marginally employed workers who lost their labour market policies during the crisis. jobs were not protected by either STW or unemploy- Additionally, regardless of the labour market measure ment insurance. Moreover, low-income earners for chosen, STW as well as the unemployment insurance whom the short-time allowance was not sufficient were carry the risk of moral hazard for labour market actors additionally dependent on basic income support. In and are associated with social costs. Regarding STW, the United States, the Stimulus Checks and the Federal Cahuc et al. (2021) document moral hazard problems on Pandemic Unemployment Compensation proved to be the firm side by showing that firms with relatively low an effective instrument to secure income, especially for revenue shocks tend to reduce their employees’ working low wage  earners. But they were even less protected hours without actually safeguarding their jobs. In con- from job losses and suffered disproportionately. trast, too generous unemployment benefits may induce Acknowledgements moral hazard issues if they reduce search effort for new We thank Lukas Lehner, Janine Leschke, Fabian Lindner, Camille Logeay, Hart‑ jobs (Schmieder et al. 2016). Despite the discussed blind mut Seifert, Sabine Stephan, two anonymous referees and the editor Britta Gehrke for valuable comments. spots and social costs, Giupponi et al. (2022, 50) conclude “…that short-time work can be an efficient and expedi - Author contributions ent way to attenuate the social costs created by “excess” All four authors contributed in equal shares to the paper. All authors read and approved the final manuscript. layoffs in recessions”. Furthermore, since they argue for instance  that both instruments cover different types of Funding workers, STW and unemployment insurance can be val- Not applicable. uable complements. Availability of data and materials The data that support the findings of this study are available from the 5 Conclusion Hans-Böckler-Stiftung and the Research Data Center of the Socio-Economic Panel but restrictions apply to the availability of these data, which were used The global Covid-19 pandemic hit the economies of Ger - under license for the current study, and so are not publicly available. Data many and the United States  hard. Both countries expe- are however available from the authors upon request and with permission of rienced an economic downturn of similar magnitude. the Hans-Böckler-Stiftung and the Research Data Center of the Socio-Economic Panel. Interestingly, Germany and the United States pursued very different economic strategies to minimise the impact Declarations of the Coronavirus Crisis on the labour market. While Germany focused on safeguarding existing jobs through Ethics approval and consent to participate the use of internal flexibility measures, especially STW, Not applicable. the United States relied on a mix of external flexibility Consent for publication and income protection. This fact allowed us to examine Not applicable. more closely the German strategy of securing employ- Competing interests ment through internal flexibility and the German labour The authors declare that they have no competing interests. market development during the Coronavirus Recession by contrasting it with the chosen strategy and the labour Author details Macroeconomic Policy Institute (IMK), Georg‑Glock‑Str. 18, 40474 Düsseldorf, market development in the United States. 2 3 Germany. University of Koblenz‑Landau, 76829 Landau, Germany. K iel Uni‑ Our analysis has shown that Germany responded to 4 versity, Wilhelm‑Seelig‑Platz 1, 24118 Kiel, Germany. University of Muenster, the economic shock with a massive temporary cyclical Schlossplatz 2, 48149 Münster, Germany. reduction in working hours, mainly through STW, on Received: 24 November 2021 Accepted: 11 July 2022 a historic scale and unemployment rose only moder- ately. In the United States, on the other hand, unem- ployment rose at an unprecedented rate, but unlike previous recessions, the nature of unemployment was References quite different, being driven mostly by temporary lay- Autor, D., Cho, D., Crane, L.D., Goldar, M., Lutz, B., Montes, J., Peterman, W.B., Ratner, D., Villar, D., Yildirmaz, A.: The $800 billion paycheck protection offs. This allowed for a very fast recovery of unem- program: where did the money go and why did it go there? J. Econ. ployment in the United States—much faster than in Perspect. 36(2), 55–80 (2022). https:// doi. org/ 10. 1257/ jep. 36.2. 55 previous recessions. 11 Page 22 of 22 A. Herzog‑Stein et al. Balleer, A., Gehrke, B., Hochmuth, B., Merkl, C.: Autonomes Fahren statt stop Schmieder, J.F., von Wachter, T., Bender, S.: The effect of unemployment ben‑ and go: Vorschläge zur effektiven Gestaltung der deutschen Kurzarbeit. Z. efits and nonemployment durations on wages. Am. Econ. Rev. 106(3), Wirtsch. 68(3), 252–260 (2019) 739–777 (2016) Bateman, N., Martha, R.: The pandemic hurt low‑ wage workers the most—and Steffen, J.: Sozialpolitische Chronik. Die wesentlichen Änderungen in der so far, the recovery has helped them the least. https:// www. brook ings. Arbeitslosen‑, Renten‑, Kranken‑ und Pflegeversicherung sowie bei der edu/ resea rch/ the‑ pande mic‑ hurt‑ low wage‑ w‑orke rs‑ the‑ most‑ and‑ so‑ Sozialhilfe (HLU) und der Grundsicherung für Arbeitsuchende—von den far‑ the‑ recov ery‑ has‑ helped‑ them‑ the‑ least/ (2021). Accessed 3 May siebziger Jahren bis heute.” Portal Sozialpolitik. www. portal‑ sozia lpoli tik. 2022 de (2021) Barrero, J.M., Bloom, N., Davis, S.J., Meyer, B.H.: COVID‑19 is a persistent realloca‑ Walwei, U.: Erwerbsformen in Krisenzeiten: was folgt aus Corona? WSI‑Mit ‑ tion shock. AEA Papers Proc. 111, 287–291 (2021) teilungen 74(2), 151–159 (2021) Cajner, T., Crane, L., Decker, R., Grigsby, J., Hamins‑Puertolas, A., Hurst, E., Kurz, C., Yildirmaz, A.: The U.S. labor market during the beginning of the pan‑ Publisher’s Note demic recession. Brookings Papers on Economic Activity, Summer 2020, Springer Nature remains neutral with regard to jurisdictional claims in pub‑ 3‑33 (2020) lished maps and institutional affiliations. Cahuc, P., Kramarz, F., Nevoux, S.: The heterogeneous impact of short‑time work from saved jobs to windfall effects. IZA Discussion Paper No. 14381 (2021) Chetty, R., Friedman, J. N., Hendren, N., Stepner, M., The Opportunity Insights Team: How did COVID‑19 and stabilization policies affect spending and employment? A new real‑time economic tracker based on private sector data. NBER Working PaperNo. 27431 (2020) Christl, M., de Poli, S., Hufkens, T., Peichl, A., Ricci, M.: The role of short‑time work and discretionary policy measures in mitigating the effects of the COVID ‑ 19 crisis in germany. CESifo Working Paper No. 9072 (2021) Committee for a Responsible Federal Budget (CRFB): COVID Money Tracker. Committee for a Responsible Federal Budget. https:// www. covid money track er. org/ (2021). Accessed 19 July 2022 Ellguth, P., Gerner, H‑D., Zapf, I.: Flexible Arbeitszeitgestaltung wird immer wichtiger. IAB‑Kurzbericht 15/2018 (2018) Federal Ministry of Finance (BMF): Corona‑Unternehmenshilfen—eine vor ‑ läufige Bilanz. Monatsbericht des BMF (November 2021). Analysen und Berichte (2021) Ganong, P., Noel, P., Vavra, J.: US unemployment insurance replacement rates during the pandemic. J. Public Econ. 191, Article 104273 (2020) Gallant, J., Kroft, K., Lange, F., Notowidigdo, M.: Temporary unemployment and labor market dynamics during the COVID‑19 recession. Brookings Papers on Economic Activity, Fall 2020, 167‑216 (2020) Gehrke, B., Hochmuth, B.: Counteracting unemployment in crises: non‑linear effects of short ‑time work policy*. Scand. J. Econ. 123(1), 144–183 (2021). https:// doi. org/ 10. 1111/ sjoe. 12395 Giupponi, G., Landais, C., Lapeyre, A.: Should we insure workers or jobs during recessions? J. Econ. Perspect. 36(2), 29–54 (2022) Hall, R., Kudlyak, M.: The unemployed with jobs and without jobs. NBER Work‑ ing Paper No. 27886 (2022) Herzog‑Stein, A., Zapf, I.: Navigating the great recession. ILR Rev. 67(3), 891–925 (2014) Herzog‑Stein, A., Lindner, F., Sturn, S.: The German employment miracle in the great recession: the significance and institutional foundations of tempo ‑ rary working‑time reductions. Oxf. Econ. Pap. 70(1), 206–224 (2018) Herzog‑Stein, A., Nüß, P., Peede, L., Stein, U.: Germany’s labour market in Coronavirus distress. New challenges to safeguarding employment. IMK Working Paper No. 209 (2021) Houseman, S.: IZA COVID‑19 crisis response monitoring: United States (Janu‑ ary 2022). Country Report (2022) Krolikowski, P.M., Weixel, A.: Short‑time compensation: an alternative to layoffs during COVID‑19. Economic Commentary No. 2020‑26. Federal Reserve Bank of Cleveland (2020). https:// doi. org/ 10. 26509/ frbc‑ ec‑ 202026 Kruppe, T., Osiander, C.: Kurzarbeit in der Corona‑Krise: Wer ist wie stark betrof‑ fen?—IAB‑Forum. https:// www. iab‑ forum. de/ kurza rbeit‑ in‑ der‑ corona‑ krise w‑er‑ ist wie‑‑ stark‑ betroen/ ff (2020). Accessed 11 May 2022 Larrimore, J., Mortenson, J., Splinter, D.: Earnings shocks and stabilization dur‑ ing COVID‑19. J. Public Econ. 206, Article 104597 (2022) Parolin, Z., Curran, M., Matsudaira, J., Waldfogel, J., Wimer, C.: Monthly poverty rates in the United States during the COVID‑19 pandemic. Poverty and Social Policy Working Paper (2020) Pusch, T., Seifert, H.: Kurzarbeit in der Corona‑Krise mit neuen Schwerpunkten. Wirtschafts‑ und Sozialwissenschaftliches Institut ( WSI). WSI Policy Brief 47 (09/2020) (2020) Reich, M., Gordon, D.M., Edwards, R.C.: A theory of labor market segmentation. Am. Econ. Rev. 63(2), 359–365 (1973)

Journal

Journal for Labour Market ResearchSpringer Journals

Published: Dec 1, 2022

Keywords: Working-time reduction; Safeguarding employment; Unemployment; Internal flexibility; External flexibility; Short-time work; Temporary layoffs; Great recession; Coronavirus recession; Covid-19 pandemic; E24; E32; J08; J20

References