An Assessment of Job Retention Support During the COVID‑19 Pandemic and its Aftermath in Belgium

1. Overview and recommendations

Abstract

This chapter provides an overview of the report, summarising its main findings on the design, uptake and effectiveness of the system, as well as its comparison to the schemes in other OECD countries. These findings offer practical guidance for policymakers to inform the refinement of job retention support for the future.

1.1. Introduction

Job retention schemes represent one of the key policy instruments to promote labour market resilience by preserving employment relationships and preventing widespread job losses during economic downturns. They do so simultaneously by reducing labour costs for firms and providing income support to workers. This reduces the risk that firms run out of liquidity and have to resort to layoffs, while at the same time preserves firm-specific human capital embedded in job matches that remain valuable in the longer term. As such, these schemes can effectively minimise the immediate social and economic costs of unemployment and might have saved an estimated 21 million jobs across the OECD (2021[1]).

During the COVID‑19 pandemic, job retention schemes also played a particularly important role in supporting labour market resilience in Belgium. Job retention support was scaled up quickly following the first containment measures, reaching an unprecedented peak of 30% of salaried employment in April 2020, well above the OECD average of just under 20%. Moreover, effective support to firms was provided instantaneously as benefits were paid directly to workers with no need for firms to advance payments to workers for hours not worked, as was the case in most other countries with job retention schemes. The provision of timely support is likely to have played an important role in preserving jobs and preventing a surge in unemployment. However, there have also been concerns that the use of job retention support was more persistent in Belgium than in other OECD countries. Indeed, at the turn of the year 2024/25, take‑up in Belgium stood around 3% of employment, whereas its use had declined to negligible levels in most other countries (less than 1%). There is a risk that the persistent use of job retention support in Belgium could slow job reallocation across firms, reinforce labour shortages and undermine the economic recovery.

To better understand the role of job retention support in Belgium, this review provides a detailed assessment of the role of the system by focussing on its design, its use and its effects during the COVID‑19 crisis as well as its immediate aftermath. After providing a detailed overview of job retention support and take‑up in Belgium, it makes key contributions. First, it provides a systematic comparison of the job retention scheme in Belgium with that of other OECD countries based on a new synthetic indicator of de jure job retention support. Second, it provides an empirical analysis of the importance of institutional differences in the design of job retention schemes for their use during COVID‑19 with a focus on worker and firm eligibility, work-sharing requirements and the generosity of support for workers and firms. Third, it provides an empirical evaluation of the effectiveness of job retention schemes in preserving jobs by exploiting differences in the generosity of support between countries (and time) and differences in the need for support across occupations within countries (based on the lack of teleworkability).

The review complements various important ongoing and recent projects in Belgium and the OECD. First, it supplements Bermudez et al. (2025[2]) for Belgium by relying on cross-country analysis and a stronger emphasis on policy design, as well as research by the Belgian administration itself (e.g. ONEM (2024[3])). Second, it builds on recent OECD work on the effectiveness of job retention schemes including a recent evaluation for Spain (OECD (2024[4]) and a cross-country analysis of the role of job retention schemes for employment dynamics by Calligaris et al. (2023[5]). It further complements academic research on the effectiveness of job retention schemes, such as Giupponi and Landais (2023[6]), Montenegro and Hijzen (forthcoming[7]), among others (see e.g. Hijzen, Jongen and Montenegro (2024[8]), Cahuc (2024[9]) and Bermudez, Dejemeppe and Tarullo (2023[10]) for an overview).

1.2. Key findings of the report

1.2.1. The Belgian job retention scheme was scaled up in response to the COVID‑19 pandemic

Belgium entered the COVID‑19 pandemic with a well-established “temporary unemployment” (chômage temporaire/tijdelijke werkloosheid) system. This job retention scheme operated through a wide array of different modalities depending on circumstances, comprising of separate sub-schemes to deal with, among others, economic difficulty (accounting for 70% of claims pre‑pandemic), force majeure events, and bad weather conditions (particularly important for construction). Notably, the scheme effectively distinguished between blue‑ and white‑collar workers, the latter of which faced stricter firm-level eligibility requirements.

The pandemic prompted Belgium to rapidly scale up its existing framework by introducing a simplified force majeure scheme in March 2020, building on the regular force majeure variant of temporary unemployment. This new scheme streamlined administrative procedures, reduced notification periods, eliminated control card requirements, removed eligibility conditions requiring workers to qualify for full unemployment benefits, and was applied similarly to blue‑ and white‑collar workers. The simplified scheme provided more generous support to workers at 70% of gross earnings plus daily supplements, compared to the standard 65% replacement rate before the COVID‑19 pandemic.

While the simplified force majeure scheme was effective in quickly adjusting to labour market implications of the health situation and containment measures, it was kept in place for a particularly long time. With the exception for a short-lived tightening to most exposed sectors and firms in September 2020, the simplified variant remained active with minor modifications until June 2022, after which Belgium returned to its pre‑pandemic temporary unemployment schemes (with small adjustments and simplifications). Unlike many other OECD countries, the crisis scheme did not incorporate training incentives.

1.2.2. Job retention support was concentrated among specific sectors, firms and workers

This report includes an empirical analysis of job retention scheme take‑up in Belgium before during, and immediately after the COVID‑19 pandemic. The analysis is based on administrative data from ONEM/RVA, which presents general trends in the use of job retention schemes, as well linked employer-employee data from the Crossroads Bank for Social Security and the Belgian National Bank. The latter provides novel insights on the selection of take‑up across different groups of firms in terms of productivity and prior use.

Belgium reached an unprecedented peak of 30% of salaried employment placed on job retention support in April 2020, well above the OECD average of just under 20%. The use of temporary unemployment tended to be concentrated in specific sectors. Prior to the COVID‑19 pandemic, it was concentrated in construction, where 10% of workers were placed on job retention prior to the pandemic, as well as manufacturing, administrative and support services. The use of job retention support during the COVID‑19 pandemic was concentrated in sectors with high exposure to containment measures, which mostly also overlap with those with elevated take‑up before the pandemic, such as construction, manufacturing, administrative support services, but also the accommodation and food services sector, which previously had minimal reliance on such schemes.

Less productive firms demonstrated substantially higher participation rates, with 16.6% of workers in the least productive firms placed on job retention support during the pandemic compared to 6.4% in the most productive firms. This pattern was consistent both before and during COVID‑19, but the degree of selection across productivity levels increased. Overall, this suggests that financial constraints, which are likely to be more common among lower-productivity firms, drove much of the uptake.

Prior experience with job retention schemes emerged as a strong predictor of usage during the COVID‑19 pandemic, with 51% of firms that had previously used temporary unemployment utilising support compared to only 14% of those without prior experience. This could either reflect familiarity with the scheme or the persistence of use for structural reasons among low-productivity firms. Firms with established knowledge of temporary unemployment schemes may have been better positioned to utilise these schemes, while structural users and less productive firms may have been more exposed to liquidity constraints during the COVID‑19 pandemic.

White‑collar workers comprised 43% of recipients at the pandemic’s peak in Q2 2020, marking a significant departure from the historical concentration among blue‑collar workers, who had previously accounted for over 90% of recipients. The use of job retention schemes varied considerably across educational groups, with low-educated workers experiencing substantially higher rates of temporary unemployment than their highly educated counterparts.

1.2.3. Job retention likely reduced separations, but dampened productivity-enhancing reallocation well beyond the acute crisis

Before the pandemic, labour market reallocation in Belgium consistently shifted employment from lower productivity firms to higher productivity firms, primarily through voluntary job-to-job mobility rather than employment mobility (workers entering or leaving employment). During the COVID‑19 pandemic, this “job ladder” mechanism continued to operate but was significantly dampened by negative changes in the contribution of employment mobility to productivity-enhancing job reallocation, particularly affecting less productive firms. The analysis in this report suggests this disruption was likely influenced by the widespread use of job retention schemes, given that less productive firms tend to be the primary users of such support and saw reduced separations rather than increased hiring.

While these patterns were in part intended to reduce employment losses during the COVID‑19 pandemic, the contribution of employment mobility to productivity-enhancing job reallocation remained negative well beyond the acute phase of the crisis, at a time when vacancy posting rose to record high levels. This persistent disruption continued to suppress productivity-enhancing reallocation into 2022 and beyond, suggesting that whilst job retention schemes may have successfully preserved employment during the crisis, their effects on labour market functioning persisted long after the pandemic had subsided, potentially dampening overall productivity growth.

1.2.4. Belgium adjusted job retention support slower than other countries after the crisis

This report develops a new synthetic indicator of effective generosity of job retention support, measuring support across three dimensions: eligibility, work-sharing (the maximum permissible reduction in working hours), and generosity (the cost covered by the government). Using this indicator, it compares Belgium’s job retention scheme before and during the COVID‑19 pandemic with the job retention schemes of other OECD countries.

Belgium’s effective generosity of job retention support ranked in the lower third of OECD countries during the peak of the COVID‑19 pandemic. This mainly reflects the modest generosity of the scheme in supporting the incomes of workers due to a low benefit cap, despite broad potential coverage. This suggests that the scheme was relatively well targeted among workers most in need for income support.

Pandemic-related adjustments to the temporary unemployment scheme remained in place for a considerable period, with the simplified scheme ending only in July 2022 and temporary unemployment returning to most of its pre‑pandemic conditions by July 2023. Unlike Belgium, most other countries began reducing the generosity of their job retention schemes as their economies recovered or phased out their schemes entirely. By the end of 2022, the Belgian job retention scheme was therefore amongst the most generous and widely used schemes across the OECD, likely contributing to the highest post-pandemic take‑up rates of job retention across OECD countries.

Further empirical analysis of the drivers of job retention scheme take‑up across countries suggests that the use of job retention support during the COVID‑19 pandemic was influenced more strongly by the stringency of containment measures than by the effective generosity of the schemes. However, amongst the design features considered, the cost to firms for hours not worked was particularly important in driving take‑up. For Belgium, such direct co-financing requirements were largely absent, potentially contributing to the recurrent use of the scheme and continued dependence on it.

1.2.5. Job retention support effectively avoided large employment losses

The report provides new evidence that shows that job retention schemes were successful in averting considerable employment losses, particularly amongst occupations most exposed to containment measures. The effects of job retention schemes on employment are analysed by exploiting differences in the generosity of support between countries (and time) and differences in the need for support across occupations within countries (based on potential to telework).

Job retention schemes proved highly effective at preserving employment at the peak of the COVID‑19 pandemic, specifically in occupations that offered little possibility for telework and were therefore most vulnerable to lockdown restrictions. Conversely, job retention schemes had negligible effects on teleworkable occupations, which were less exposed to containment measures. Overall, the analysis shows that in the absence of job retention schemes, employment would have declined by approximately 8% on average across countries at the crisis peak.

With high take‑up of job retention schemes, the Belgian scheme had particularly strong effects on preserving employment at the peak of the COVID‑19 pandemic. It is estimated to have averted an employment loss of 12.9% at the pandemic’s peak, substantially higher than the cross-country average. For every 100 workers placed on job retention schemes in Belgium, 55 jobs were preserved, resulting in a moderate deadweight effect of 45%. However, as Belgium maintained its expanded scheme provisions well beyond the acute crisis period, with minimal adjustments well beyond the acute COVID‑19 crisis, the continued and potentially structural use of job retention support subsequently had small negative effects on employment as pandemic restrictions were lifted.

1.3. Recommendations for the temporary unemployment scheme

Overall, the Belgian temporary unemployment scheme was well placed to offer a rapid and effective response to the unprecedented impact of the COVID‑19 pandemic. The simplified scheme was promptly deployed and reached broad coverage across workers and firms most exposed to the containment restrictions. As such, it averted significant employment losses. Despite the relative success of the temporary unemployment scheme during the COVID‑19 pandemic, this report identifies several options that could help to strengthen the scheme while maintaining its effectiveness in responding to an acute and unprecedented crisis.

1.3.1. Modulating the costs of job retention support to firms over time

Outside of the COVID‑19 pandemic, the Belgian temporary unemployment scheme is characterised by high take‑up compared to other OECD countries. Even in normal times, take‑up remains at 2‑3% of employment, particularly in construction and manufacturing. This is well above all other OECD countries, who averaged at 0.2% of employment in December 2022 and suggests that the scheme has gone beyond its intended function as a crisis response mechanism and, at least for some firms, use it as a tool to address structural weaknesses, distorting the normal functioning of the labour market and undermining productivity growth. By making firms that frequently use the scheme contribute more to its financing through an experience‑rating system, which links the job retention scheme use with increases in future social contributions, the use for predictable and seasonal fluctuations would be discouraged, without re‑enforcing liquidity constraints. This would reduce the risk of creating structural benefit dependency, while preserving it as a crucial safety net for unexpected economic shocks. While the current responsibility contribution disincentivises the continued placement of specific workers on job retention support, it does not prevent using it on a quasi-permanent basis at a low intensity (in terms of the share of workers placed on reduce working time), a practice that is quite common in Belgium. Instead, following the example of experience‑rating used in France or the United States as part of their unemployment insurance, could provide the necessary disincentives to frequent use of job retention support.

1.3.2. Removing crisis-related increases the generosity of support when the crisis subsides

In contrast to many other OECD countries, the Belgian system saw very minor adjustments to the effective generosity of job retention support when the immediate crisis and its impact on the labour market subsided, potentially driving some of the persistence in use well after the end of the COVID‑19 pandemic. For future shocks, it may therefore be useful to reduce the generosity of the scheme earlier and along with the changing conditions of the labour market. A part of this may be adjusting firm eligibility requirements to specific sectors particularly exposed to the crisis (as attempted in September 2020) or by re‑introducing conditions on declining turnover (or production/orders) much earlier. This may help to avoid continued and potentially structural use among firms without acute need or future viability, which can undermine the normal functioning of the labour market.

1.3.3. Simplifying the fragmented menu of temporary unemployment schemes

Beyond the issue of persistent and structural use of the system, there is some scope to simplify its fragmented nature across sectors, workers and circumstances. The successful implementation of the simplified force majeure scheme during the COVID‑19 pandemic, which applied uniformly to all workers, demonstrated that a unified system can be effective, even in times of profound crises. Furthermore, the goal of targeting support to more vulnerable workers is already achieved effectively through the scheme’s low benefit cap, which makes the support for workers on reduced hours significantly more generous for low-wage earners than for those on the average wage. A potential reform could therefore be to eliminate the blue‑ versus white‑collar distinction entirely, unifying the rules for temporary unemployment due to economic reasons. This would simplify the system for employers and align Belgium with the practices of other OECD countries, while also following the general, though incomplete, process of aligning the labour market status of blue‑ and white‑collar workers in Belgium since the law of 26 December 2013.1

1.3.4. Embedding training as an element of job retention support

Periods of job retention can be a valuable opportunity to invest in upskilling, especially for workers with lower skillsets. The high concentration of usage among low-skilled workers, with 20% of low-educated workers receiving support in 2020 compared to just 3% of highly educated workers, reveals missed opportunities for addressing skills gaps and improving long-term employability. Training can improve the viability of current jobs, prevent skills from deteriorating during long absences from work, or improve the prospects of workers of finding a new job in expanding industries or occupations (e.g. France, Spain). Currently, the Belgian system of temporary unemployment does not include any direct incentives for training and efforts to introduce such measures have faced significant problems. Future reforms could create a clear framework for training within the Belgian temporary unemployment system. By incorporating training incentives, Belgium could not only preserve jobs but also support productivity and wages. This would require a better co‑ordination between national and regional authorities. Collaboration with existing vocational education infrastructure, sectoral training funds as well as social partners would also be essential to ensure workers receive high-quality training that is relevant to market needs. To avoid signalling that jobs are not viable, the focus should be on upskilling for current roles and future sectoral needs, rather than generic retraining that may suggest a job is at risk.

References

[2] Bermudez, N. et al. (2025), “Le chômage temporaire en Belgique : un outil efficace et bénéfique, à condition d’en maîtriser l’usage”, Regards économiques, Vol. 190.

[10] Bermudez, N., M. Dejemeppe and G. Tarullo (2023), “Theory and Empirics of Short-Time Work: A Review”, GLO Discussion Paper, Vol. 1348, https://www.econstor.eu/bitstream/10419/279649/1/GLO-DP-1348.pdf.

[9] Cahuc, P. (2024), “The micro and macro economics of short-time work”, in Handbook of Labor Economics, Elsevier, https://doi.org/10.1016/bs.heslab.2024.11.011.

[5] Calligaris, S. et al. (2023), “Employment dynamics across firms during COVID-19: The role of job retention schemes”, OECD Economics Department Working Papers, No. 1788, OECD Publishing, Paris, https://doi.org/10.1787/33388537-en.

[6] Giupponi, G. and C. Landais (2023), “Subsidizing Labour Hoarding in Recessions: The Employment and Welfare Effects of Short-time Work”, The Review of Economic Studies, Vol. 90/4, pp. 1963-2005, https://doi.org/10.1093/restud/rdac069.

[8] Hijzen, A., E. Jongen and M. Montenegro (2024), “The effectiveness of job retention schemes during economic crises”, in Handbook on Labour Markets in Transition, Edward Elgar Publishing, https://doi.org/10.4337/9781839106958.00021.

[11] Knoops, S. (2021), “Equality without improvement? A case study of the impact of the Belgian unified status among blue-collar and white-collar workers on their occupational pensions”, European Journal of Social Security, Vol. 23/3, pp. 211-231, https://doi.org/10.1177/13882627211031932.

[7] Montenegro, M. and A. Hijzen (forthcoming), “Job Retention at Scale”.

[4] OECD (2024), Preparing ERTE for the Future: An Evaluation of Job Retention Support in Spain During the COVID-19 Pandemic, OECD Publishing, Paris, https://doi.org/10.1787/a70bf8ec-en.

[1] OECD (2021), OECD Employment Outlook 2021: Navigating the COVID-19 Crisis and Recovery, OECD Publishing, Paris, https://doi.org/10.1787/5a700c4b-en.

[3] ONEM (2024), Le chômage temporaire après la crise du coronavirus, https://www.onem.be/file/cc73d96153bbd5448a56f19d925d05b1379c7f21/56b3ce93b30eb832bcabe25e6f5d83cdbf42ebd2/2024_11_21_etude_ct_post_corona_fr.pdf.

Note

← 1.  After decades of historical labour market discrimination between blue‑ and white‑collar workers, the law of 26 December 2013 prescribes a harmonisation of the labour market status between the two by 2025, eventually extended to 2030 by agreement of the social partners (Knoops, 2021[11]). This primarily affects notice period, sick pay, and occupational pensions.