Riding the waves: Adjusting job retention schemes through the COVID-19 crisis

 Introduction

Job retention (JR) schemes seek to preserve jobs at firms experiencing a temporary decline in business activity by lowering their labour costs and supporting the incomes of workers whose hours are reduced. They may take the form of short-time-work schemes that directly subsidise hours not worked or wage subsidy schemes that subsidise hours worked but can also be used to top up the earnings of workers on reduced hours. A crucial aspect of all JR schemes is that employees’ contracts remain in force even if their work is fully suspended.

As the COVID-19 crisis took off, nearly all OECD countries took steps to ensure that JR schemes provided timely and broad-based support for firms and workers that were affected by social distancing restrictions. This contributed to record high levels of take-up by firms of such schemes and, importantly, helped prevent a surge in unemployment and rising financial hardship among workers and their families. Preliminary estimates suggest that during the initial period of the COVID-19 crisis, JR schemes may have saved up to 21 million jobs across the OECD (OECD, 2021[1]).1 However, as the health and economic situation has evolved and the need for JR support has declined, many countries have adjusted, scaled down or even phased out their JR schemes. Others have kept temporary measures in place in the face of high uncertainty over the ability of governments to prevent recurrent waves of infections.

This document takes stock of the different approaches taken by OECD governments in adjusting the temporary measures taken in response to the COVID-19 crisis to scale up JR support. It starts by documenting how the use of JR support has evolved since reaching its peak in the early phase of the COVID-19 crisis. It proceeds by documenting how governments have adjusted their JR schemes as the crisis evolved by targeting support more strategically to jobs that need support to survive but still remain viable in the medium term. The present document complements an earlier one that documented what measures governments had taken to scale up JR support at the start of the crisis (OECD, 2020[2]) and a chapter in the OECD Employment Outlook 2021 that provided a preliminary assessment of the role of JR schemes during the COVID-19 crisis (OECD, 2021[1]).

 The use of job retention support has declined significantly but remains elevated in some countries

The use of JR support has declined sharply since reaching its peak during the first wave of the COVID-19 crisis in 2020 (Figure 1). From reaching an unprecedented peak of 20% of employment in April/May 2020 on average across the OECD, it fell to 1.3% of employment in November/December 2021, the most recent date for which cross-country data on take-up are available.

The significant decline in take-up since its peak in April/May 2020 reflects to an important extent the reduced weight of social distancing restrictions on economic activity and the subsequent need for public support.2 However, it is also likely to reflect to some extent the adjustments that countries have made to the design of JR schemes as the health and economic situation evolved and the need for JR support declined. Indeed, most of the temporary schemes that were introduced in response to the COVID-19 crisis are no longer operational (e.g. Australia, Canada, Denmark, New Zealand, the United Kingdom), while others are scheduled to end their operations in the near future (e.g. Ireland, the Netherlands). In many other countries the schemes have been increasingly targeted to ensure their continued cost-effectiveness (as discussed in detailed below). The significant decline in the use of JR support suggests that its unprecedented use was by and large temporary and that JR support has not been used on a massive scale to support firms with structural problems.

The use of JR support remains, however, elevated in some countries. Take-up is particularly high in Ireland and the Netherlands, which still operated a temporary wage subsidy scheme in November 2021, and to a lesser extent, Belgium and Italy, two countries with pre-existing short-time work schemes. Apart from the weakness of the economic recovery, this may also reflect the specific design of JR schemes in those countries. Deadweight effects that arise when jobs not needing support are supported represent a particular concern in countries with wage subsidies due to the tendency to use them across all workers in the firm irrespective of whose working hours are reduced (OECD, 2020[2]). The absence of co-financing requirements on firms for the costs of hours not worked in Belgium and Italy (see Figure 3) raises potential concerns about displacement effects, i.e. the possibility that support is going to jobs that have become permanently unviable. Keeping workers in unviable jobs not only adds to the fiscal costs of JR support, but also could undermine the recovery by slowing reallocation from less productive to more productive firms and reinforcing labour shortages.

 

Figure 1. The use of job retention support has declined significantly but remains elevated in some countries

Percentage of dependent employment

Note: Take-up rates are calculated as a percentage of all dependent employees in 2020 Q1. *Latest data refer to October 2021 (Luxembourg), August 2021 (Portugal) and September 2021 (Sweden). † Australia, Estonia, Hungary, Latvia, Lithuania, New Zealand, Poland, Slovenia, United Kingdom and Turkey: Scheme no longer operational or not widely available. #Japan, the Netherlands: estimates based on the total use during the reference period and the assumption that support is provided for no more than three months during this period. # Italy: latest data is estimated based on the number of authorised hours. The United States: Refer to short-time compensation benefits. No information on take-up available for Colombia, Israel and Iceland. No scheme present in Costa Rica and Mexico. Sorted by latest available data.

Source: National sources.

 

Box 1. The use of job retention has almost returned to pre-crisis levels

This box documents the evolution in the use of job retention schemes since the start of the global financial crisis (2007 Q1) to the most recent period for which data are available (November 2021) for four selected countries with pre-existing short-time work schemes, i.e. France, Germany, Italy and Spain (Figure 2). This provides the following key insights:

• Job retention support during the COVID-19 crisis peaked at a much higher rate than during the global financial crisis. On average across the four countries considered, the peak during the COVID-19 crisis was nine times higher as during the global financial crisis. The difference is particularly striking for Spain and France, where job retention support played a minor role during the global financial crisis. In both countries the use of job retention schemes has been greatly simplified in response to the COVID-19 crisis.

• The use of job retention support has declined significantly since reaching its peak in all four countries, irrespective of their experience with such schemes during the global financial crisis. As of November 2021, take-up tended to be only somewhat higher than at the onset of the COVID-19 crisis. The decline in take-up not only reflects the reduced need for support but also its increasingly targeted nature. The use of work resumption bonuses in Spain may further have contributed to the reduction in take-up.

• The synchronised reduction in take-up across countries during the COVID-19 crisis stands in contrast with the experience during the global financial crisis when country differences in the persistence of take-up tended to be more pronounced. Whereas in France, Germany and Spain, the use of job retention support returned within two years from reaching its peak to its pre-crisis level, in Italy this took around ten years.1

While significant differences in the design of the job retention schemes remain, they have tended to converge in the sense that their use can be scaled up quickly in response to a sharp economic downturn and be phased out as the economy recovers (OECD, forthcoming[3]). For Spain in particular, the crisis presented an opportunity to fulfil its ambition since the global financial crisis to establish a scheme which could be scaled up quickly during economic downturns. As a result of the widespread use of job retention support in Spain during the COVID-19 crisis the rise in unemployment was much smaller than during the global financial crisis, despite a larger reduction in economic activity.

 

Figure 2. The use of short-time work in historical perspective

Percentage of dependent employment, January 2007 – November 2021

Italy: Data before 2018 are based on the number of authorised hours (estimated number of employees using the ratio of the total hours authorised under the quarterly average hours worked by employee), and spliced using the actual number of participants from January 2018. Spain: Data are not available from October 2010 to February 2020.

Source: OECD (forthcoming[3]), Preparing ERTE for the Future, Chapter 2.

1. The persistent use of STW in Italy most likely reflects a combination of factors, related to the absence of significant restrictions on the use of STW and a certain reluctance on the side of employers to lay off workers in a context where support for such workers was limited (e.g. economic dismissals were not entitled to severance pay, a universal unemployment insurance system did not exist). This situation changed with the Jobs Act of 2015. This restricted the use of STW by shortening its maximum duration, particularly for workers on zero hours, and increasing the use of co-financing by firms for the costs of hours not worked. At the same time, it may have increased the social acceptability of economic layoffs by introducing tenure-dependent severance pay and paving the way for the establishment of a broader and more modern unemployment insurance system. This is consistent with the evolution of take-up of STW, which shows that it declined more quickly after 2015 than during the preceding period.

 Support has become increasingly targeted to jobs in firms that are most affected by the pandemic

As social distancing restrictions have become more focused on specific sectors and the need for support has declined, several countries have increasingly targeted JR support towards jobs that are most likely to need support either by maintaining more generous support or restricting eligibility for firms whose activities have been most affected by social distancing restrictions (Table 1)

Countries that differentiated the level of support across groups of firms typically maintained generous crisis measures for firms that are directly affected by health-related economic restrictions (“crisis JR”), while they gradually introduced less generous provisions for the regular use of JR support in the case of a temporary shortfall in demand or disruptions in the supply chain (“regular JR”). The provision of more generous crisis support measures to firms whose activities are directly affected by economic restrictions may reflect the possibility that the risk of moral hazard in such firms is more limited. Yet, many of these hard-hit firms operate in sectors that tend to be characterised by relatively low levels of firm-specific capital, which may mean that the value of labour hoarding also tends to be more limited (e.g. hotels and restaurants). Another line of argument could be that the burden of administrative restrictions imposed for the benefit of all citizens should be shared collectively and that this is achieved more effectively through broad-based working-time reductions supported with JR subsidies than by laying off workers and supporting them with unemployment benefits.

Countries have increased the targeting of support towards firms affected by administrative restrictions by linking support to: i) the type of activities or areas that are subject to restrictions, and/or ii) the decline in business activity of firms. Traditionally, JR schemes limit deadweight effects, i.e. the risk of supporting jobs that do not need support, by conditioning eligibility on having an economic justification based on the reduction in sales. As the crisis has evolved, a number of countries have increased the economic threshold for eligibility in general (e.g. Ireland, Portugal) or introduced specific thresholds to determine eligibility for more generous crisis support (e.g. Austria). Others have directly linked support to administrative restrictions with respect to certain economic activities or geographic areas (e.g. Luxembourg, Spain) or combine the use of economic thresholds and the scope of administrative restrictions to determine the level of support available (e.g. Canada, France, Japan, Latvia).

 Financial incentives for firms have been strengthened to prevent support from being used too widely

Even when eligibility is limited to firms that are most affected by the remaining restrictions, it does not ensure that support goes towards jobs that are temporarily under pressure and not those that have become permanently unviable. Determining whether a job is viable or not, however, is very hard for governments, particularly so in the present context where social distancing restrictions remain present and market signals are weakened. However, firms and workers typically have a better sense of the viability of a job. This suggests that financial incentives for firms and workers could be used to enhance the targeting of JR support to jobs that are temporarily at risk, but remain viable in the medium term. In practice, this usually meant that the way the costs of short-time work for hours not worked are shared between firms, workers and the state was adjusted by requiring firms to cover a larger share of the cost of hours not worked.

As the health and economic situation evolved and concerns about the possible adverse impact of JR support, job reallocation and labour shortages have mounted, several countries introduced co-financing or increased its rate to enhance the targeting of support to jobs that remain viable in the medium term (e.g. Austria, France, Norway, Spain, Switzerland) or as part of their complete phase out (e.g. Netherlands, Slovenia, the United Kingdom). Requiring firms to cover a larger part of the costs of hours not worked (co-financing) provides incentives for firms to use public support only for jobs that are likely to re-start after the crisis and to resume regular work schedules as soon as possible. Consequently, co-financing is expected to play an important role in ensuring that JR schemes do not become an obstacle to the recovery by locking workers in unviable jobs and slowing the reallocation of workers across firms.3

As a result of these adjustments in the generosity of JR support, the majority of countries with JR schemes now require firms to co-finance some of the costs of hours not worked. This stands in stark contrast to the situation during the early phase of the crisis when most countries set the cost of hours not worked for firms to zero. This allowed firms to adjust labour costs in line with the decline in business activity, and reduced the risk that firms were running out of cash to pay their workers and forced to lay off a significant part of their workforces. However, by not having to share in the costs of hours not worked, firms faced no strong incentives to limit support to jobs that were at risk but remained viable in the medium term. Given the widespread impact of lockdown measures across different activities, this risk was limited during the early phase of the COVID-19 crisis. However, this risk is likely to have become more pronounced as time passed, as many of the firms experiencing temporary problems have now resumed their activities, while firms with more structural difficulties may continue to rely on JR support.

Only one third of countries that still operated a JR scheme in November 2021 do so without any form of co-financing. Among these countries, Germany re-introduced co-financing in January 2022, while Ireland intends to fully phase out its temporary wage subsidy scheme in the course of 2022. A number of these countries operate permanent schemes that also did not have any co-financing requirements before the COVID-19 crisis (e.g. Belgium, Finland, and Turkey). It is noteworthy that the average take-up rate across countries without co-financing is about three times as high as across countries with co-financing (3.4% versus 1.3%). The absence of co-financing increases the risk that JR schemes are used to support firms with structural problems and could undermine the economic and labour market recovery by reinforcing labour shortages and slowing the reallocation of workers to more productive firms.

 

Figure 3. JR support has tended to become less generous for firms

Cost for firms, % of labour cost

Note: ‡ Czech Republic: For November 2021, refer to Antivirus, regime B. During the Crisis, Antivirus, regime 3A. Canada: There are two schemes, the Canada Emergency wage subsidy, ended in November 2021, and the Work-sharing programme (indicated as ♦), ongoing. Denmark: There are two schemes, the system of division of labour (Arbejdsfordeling, indicated as ♦), that was temporary redesigned and the Wage compensation scheme (Lønkompensation), ended in June 2021. Greece: There are two schemes, the Special purpose compensation, restricted to some specific sectors and Syn-Ergasia, (indicated as ♦), ongoing. † Schemes no longer operational or not widely available. Mandatory employer contributions for private insurance are not taken into account (consistent with the OECD methodology of Taxing Wages). Norway: for the first 3 months (60 days).

Source: Country answers and ad hoc updates to OECD Policy Questionnaire on Working Time Regulation and Short-Time Work Schemes.

 Concluding remarks

As the COVID-19 crisis hit, nearly all OECD countries took prompt measures to ensure that JR schemes provided timely and broad-based support to firms and workers affected by social distancing restrictions. This contributed to record high levels of take-up by firms of such schemes at the height of the pandemic and, importantly, helped save millions of jobs across the OECD. As the health and economic situation has evolved and the need for JR support has declined, many countries have adjusted, scaled down or even phased out their JR schemes.

Adapting JR schemes to the evolving crisis has been a major challenge, due to the high degree of uncertainty about the short-term outlook and the varied effects of social distancing restrictions across different groups of firms. The uncertainty about the future evolution of the health situation has made it very difficult to plan ahead. A number of countries that started scaling back the level of JR support had to change track and revert back as the health situation deteriorated unexpectedly. Adjusting the operation of JR support too frequently may reduce the readability of the system and undermine its effectiveness. Other countries have avoided multiple adjustments by maintaining generous JR support with the risk of unnecessarily increasing their fiscal costs. Moreover, universal forms of JR that provide the same level of support to all workers and firms may not be optimal in a context where there are large differences in the impact of social distancing restrictions across sectors and regions. Indeed, by differentiating the level of support across groups of firms, it may be easier to align the level of JR support with the evolution of the health situation.

The trade-offs that one has to confront when adapting or phasing out JR schemes can to some extent be mitigated by the use of complementary policies. These policies can help to reduce the risk of maintaining support for too long by providing incentives for job mobility, for example through the use of work resumption bonuses as in Spain, or the use of training while on short-time work. Complementary policies can also be used to soften the blow of withdrawing support too quickly for firms and workers. For example, one can ensure that firms with more persistent but still temporary financial difficulties have access to credit through government liquidity-support programmes. Similarly, concerns about the consequences for workers and their families of phasing out support too quickly may be less pronounced in countries with well-developed systems of unemployment benefits, re-employment services and adult learning. In the Netherlands, the termination of its JR scheme was accompanied with an expansion of credit provisions and more active support for job losers.

An important further question is whether it is easier to modulate the optimal level of JR support by having a permanent or a temporary scheme. Several countries without pre-existing schemes have been able to provide timely support to a large number of firms and workers through the introduction of temporary schemes and to quickly phase out support as social distancing restrictions have become less relevant. However, the level of support provided by temporary schemes may have been somewhat crude both in terms of the type of support provided (e.g. only for workers whose jobs were fully suspended, lump-sum subsidies unrelated to the decline in business activity) and its modulation during the crisis (e.g. the risk of terminating support too early). While permanent schemes may have a more elaborate design that helps fine-tuning the level of support provided at different times, there may also be a greater risk that they end up being used to support firms with structural difficulties. Ultimately, the choice between having a temporary and a permanent is not an easy one and may mainly be based on the broader institutional context.

A priority going forward is to evaluate the effectiveness of JR schemes in preserving jobs and supporting job creation during the COVID‑19 crisis. A key aspect of any such evaluation should be to analyse the effectiveness of JR schemes in protecting different types of workers from the risk of job loss and supporting longer-term career paths. Job losses during the COVID-19 crisis have been highly uneven across groups of workers, and much more important for workers with low skills, younger workers as well as for workers in non-standard forms of work, raising important questions about the ability of JR schemes in protecting the jobs of the most vulnerable. Evidence from Switzerland suggests that short-time work does not necessarily imply significant differences in the effectiveness in protecting different groups of workers against the risk of job loss (Hijzen and Salvatori, 2022[4]). This may in part reflect the temporary extension of eligibility to temporary workers and the removal of co-financing in Switzerland in response to the COVID‑19 crisis.