By Sara Calligaris, Gabriele Ciminelli, Hélia Costa, Chiara Criscuolo, Lilas Demmou, Isabelle Desnoyers-James, Guido Franco and Rudy Verlhac[1]
Three years after the outbreak of the COVID-19 pandemic, the implications of the massive fall in economic activity and of the associated widespread government support to businesses are still not fully understood. Yet, they are key to inform policy design and action going forward, in particular with respect to their potential consequences for aggregate productivity via labour reallocation.
Against this backdrop, a recent paper (Calligaris et al., 2023) investigates employment dynamics across firms along the intensive and extensive margins during the COVID-19 pandemic, the extent to which these adjustments were productivity enhancing, and the role of job retention schemes (JRS) in shaping these patterns. The paper relies on a combination of novel and unique data. First, in collaboration with 12 participating countries, it collects high-frequency (monthly) harmonised micro-aggregated statistics, computed using administrative data on employment and wages from electronic payroll records, linked to monthly information on policy support during COVID-19. Second, it builds a new cross-country and high-frequency de-jure indicator of JRS allowing researchers to benchmark their generosity across countries and over time.
Employment adjustments and reallocation in COVID-19 times
The analysis shows that, since the onset of the COVID-19 crisis, employment adjusted through different margins. In 2020, employment adjusted mainly along the intensive margin, with a decline in employment growth of surviving firms relative to 2019, while – at the extensive margin – survival rates remained on average stable. In 2021, employment growth of surviving firms picked up, while average survival rates started to decline. These aggregate results mask high cross-country and cross-sector heterogeneity, with increases (decreases) in job destruction (creation) rates significantly higher in low-telework sectors.
These adjustments in employment dynamics also entail a reallocation of resources across firms within sectors, which is a relevant factor affecting aggregate productivity growth. Relative to 2019, the productivity enhancing nature of labour reallocation was weaker in 2020 and 2021. Indeed, while on the extensive margin high productivity firms still showed significantly higher survival rates compared to their lower productivity competitors, the contribution of the intensive margin weakened, with employment growth of surviving firms remaining only marginally related to productivity.
Figure 1. Labour reallocation remained productivity-enhancing, though to a lower extent
Difference in total employment growth relative to the bottom quartile of the productivity distribution.
Note: The figure presents differential employment growth rate for mid (25-75) and high productivity firms (75-100) relative to the baseline group of low productivity ones (0-25) in 2019, 2020, and 2021. The estimates come from regressing employment growth (between January of each year and January of the year after) on a dummy for each productivity quantile along with country-sector fixed effects and weighting the regression by sectoral employment shares. The columns represent estimated coefficients and the green bars 95% confidence intervals.
Source: Calligaris et al. (2023).
The role of job retention schemes
Job retention schemes have been the most widespread policy instrument to support workers and firms across OECD countries. The data show that the uptake of JRS by firms in the sample varies markedly across countries, sectors and over time. At its peak, around 80% (60%) of firms received support in New Zealand (Australia), and more than 20% of firms were supported in European countries such as Denmark, Latvia and the Slovak Republic. Uptake has been by far the highest in the “Accommodation and food service activities” sector and at the time the COVID-19 pandemic hit the hardest. It was gradually reduced and the eligibility requirements to access the schemes were tightened when countries lifted mobility restrictions. There was also heterogeneity in the allocation of support across firms within sectors: In all countries, JRS support was not disproportionately directed towards unproductive companies (Figure 2.A), as uptake was relatively higher for firms in the middle quartiles and in the top quartile of the productivity or size distribution.
The analysis investigates the role of JRS using: i) firm-level information on uptake to contrast employment growth and survival rates across firms at a detailed level, as well as ii) local projections estimations exploiting ex ante differences in the generosity of support design through a novel de-jure JRS indicator to further ease endogeneity concerns. The two approaches provide complementary insights, showing that JRS contributed to mitigating the negative consequences of the crisis on employment and business survival. Specifically, following a tightening in the intensity of the pandemic, employment growth was on average lower and firm exit higher in the absence of JRS, relative to when generous JRS were in place (Figure 2.B). Furthermore, the analysis shows that when no JRS were in place, employment for mid and high productivity firms decreased significantly more than when more generous JRS was in place, while no significant difference was found for the least productive firms.
Taken together, these results suggest that government policies were effective in mitigating the effects of the crisis and did not appear to distort the creative destruction process and productivity enhancing nature of reallocation.
Figure 2. JRS mitigated the consequences of the crisis, without distorting the reallocation process
Note: Panel A plots the coefficient and related 95% confidence intervals of a regression of JRS uptake on the productivity quantiles categorical variable, on a cross-country sample evaluated repeatedly at different points in time, controlling for country by industry fixed effects. Each sector is weighted according to its average size in terms of employment over the year. Panel B. The lines represent the effect on employment growth of a change in the containment stringency index between 0 and 6 months after the change, if the JRS indicator in the month before the change in severity was equal to 0 (No JRS, light blue dotted line) or to the 75th percentile (High JRS, blue filled line). The red dotted lines and the thick light blue lines represent the 90% confidence interval around the estimates.
References
Calligaris, S., G. Ciminelli, H. Costa, C. Criscuolo, L. Demmou, I. Desnoyers-James, G. Franco, R. Verlhac, (2023), “Employment dynamics across firms during COVID-19: The role of job retention schemes”, OECD Economics Department Working PaperNo 1788, https://doi.org/10.1787/33388537-en.
[1] The analysis is the result of a collaboration project between two OECD Departments, namely the Economics Department and the Directorate for Science, Technology and Innovation. It would not have been possible without the valuable contribution of national experts from the Central Bank, the Ministry of Economy and/or Finance, Revenues and Customs, or National Statistical Office of the countries participating to the project –namely, Australia, Belgium, Canada, Costa Rica, Denmark, Italy, Latvia, Netherlands, New Zealand, Norway, Slovak Republic and United Kingdom.