The COVID-19 pandemic shock was unparalleled in modern history. The strict containment measures, with broad economic consequences, called for a quick governmental response. European governments swiftly enacted policies to support businesses and households on a scale and magnitude never seen before. The discussion of support mechanisms and their intended and unintended macroeconomic effects has naturally found its place in economic literature.
Evidence based on data on French firms (Coeuré 2021) suggested that the pandemic government support sharply reduced the number of insolvent or failing companies. Van der Wielen et al. (2021) analysed data from an EU-wide survey among firms and confirmed that the COVID-19 support successfully reached the firms that suffered the most in terms of pandemic-induced revenue reductions and it avoided a liquidity dry-out and freezing of the corporate ecosystem.
Bighelli et al. (2021) argued that the COVID-19 government support has not been as unproductively distributed as feared. The subsidies were distributed towards medium-productivity firms, and only marginally towards the undeserving ‘zombies’, in several euro area countries. Rodano et al. (2022) confirmed for Italy that zombie firms were less likely than healthy firms to access public support measures.
The Expert Group on Productivity, Innovation and Technological Change, which includes a team of experts from the European System of Central Banks (ESCB), conducted an analysis of the short- and long-term impacts of the pandemic, including the containment measures and policy support, on EU productivity trends. Their work (summarised in Lalinsky et al. 2024) relied on the analysis of aggregate, sector and a large micro-distributed dataset including data from 12 euro area countries.
In this column, we show a subset of results of the expert group focusing on the distribution of the pandemic policy support, and how it changed over time, across firms of different productivity levels.
The reason to focus on this particular aspect is its potential implications for aggregate productivity growth: if policy support unintendedly helped the survival of low-productivity firms, and thereby contributed to resource misallocation, aggregate productivity would be negatively affected over the short and medium term.
We document declining efficiency of the allocation of employment subsidies to firms with respect to their productivity. The analysis starts with a description of the overall evolution of the support to corporates over time, continues with the distribution of the support to different productivity clusters, and finally investigates whether the probability to receive subsidies and their size changed over time.
The dynamics of the corporate support over time
Although, the spread of the COVID-19 virus and its impact differed across countries, European governments responded with support to firms on a colossal scale. The employment support, provided mostly on a monthly basis, on average followed the stringency of the policies to mitigate the impact of the pandemic on public health and society. As suggested by Figure 1 showing aggregate developments for our sample countries, both the size of support and number of supported firms peaked in the first wave of the pandemic, when the implementation of lockdowns, social distancing measures, and other measures to curb the spread of the virus resulted in significant drops in sales affecting a large share of businesses, including financially healthy and viable ones.
Although firm eligibility was not initially directly connected to the stringency of the policy measure to curb the virus, the employment support started to rise again in October 2020, shortly after the onset of the second wave of COVID-19. It peaked again in the first months of 2021. The support then gradually declined and was suspended for some time in the summer of 2021 in several countries. Further months with somewhat increased spread of the virus and stringency measures saw only mild increases in the employment subsidies, as the severity of the COVID-19 virus declined and the ability to deal with the health and economic consequences improved.
Figure 1 The scale of support and the impact of the COVID-19 pandemic
Source: Authors’ calculations based on the output of the micro-distributed exercise, Lalinsky et al. (2024) and Hale et al. (2021).
Notes: Index (March 2020=1) for allocated support and number of supported firms based on aggregate data for Croatia, Latvia and Slovakia – countries with granular data for 2020 and 2021. The stringency index is a composite measure based on nine response indicators including school closures, workplace closures, and travel bans, rescaled to a value from 0 to 100 (100 = strictest) averaged across Croatia, Latvia and Slovakia.
In what follows, we study the distribution of the wage subsidies – one of the main instruments to support firms’ liquidity – across firms of different productivity levels in three steps. First, we divide firms into quintiles based on their productivity relative to the productivity of the rest of firms operating in the same country, and analyse the aggregate values of subsidies allocated to each of the quintiles. Next, we analyse how the probability of receiving subsidies varied depending on firm characteristics, and in particular firm productivity (the ‘extensive margin’). Lastly, focusing on those firms which received support, we analyse the correlation between the amount of support granted to each firm and its productivity (the ‘intensive margin’). For details of our regressions, see Labinsky et al. (2024).
The allocation of wage subsidies across productivity quantiles
In 2020, almost one-third of wage subsidies were allocated to firms in the top 20% of the pre-pandemic productivity distribution. Firms with above-median productivity received about two-thirds of all subsidies, i.e. significantly more than their proportional share. Only a small share of subsidies went to non-productive firms, defined as firms in the lowest quintile of the productivity distribution.
The allocation of wage subsidies changed in 2021. The distribution of support shifted towards less productive firms, as shown in Figure 2. The share of subsidies allocated to high-productivity firms declined. To better understand the drivers of these developments, we apply regression analysis to disentangle the role of the extensive and intensive margins.
Figure 2 Share of support allocated to firm productivity quintiles, 2020 and 2021
Source: Authors’ calculations based on the output of the micro-distributed exercise, Lalinsky et al. (2024).
Notes: Mean values for Croatia, Latvia and Slovakia – countries with granular data for both periods.
The extensive and intensive margins: Analysis of which firms received wage subsidies and in what quantity
Di Mauro et al. (2021) demonstrated a non-linear relationship between the likelihood of receiving support and firm productivity. They found that firms around the median of the productivity distribution had the highest probability of receiving support in 2020. Our research confirms this relationship for a broader group of countries which extend beyond the Central and Eastern Europe (CEE) region analysed in their work.
Although the degree of non-linearity and the magnitude of marginal gains vary among individual countries, on average firms from the 6th decile of productivity had about a 15% greater chance of being supported than firms from the bottom 10% of the productivity distribution.
As illustrated in Figure 3, the correlation between the probability of receiving support and firm productivity weakened during the second year of the pandemic. The reason was that high-productivity firms exited the supporting schemes earlier and, therefore, the relative probability of low-productivity firms being supported increased. Note that these results are based on data for Croatia, Latvia, and Slovakia, the only three countries with available data for 2021 at the time of writing the report. However, given that results were very similar across all countries in 2020, and that results for 2021 are similar across the three countries with data, we think this development may have been similar in other euro area countries as well.
Figure 3 Firm probability of receiving support by productivity, 2020 and 2021
Source: Authors’ calculations based on the output of the micro-distributed exercise, Lalinsky et al. (2024).
Notes: Marginal effects from logit regressions representing changes in the probability of receiving wage support for a change in firm productivity decile with respect to the lowest productivity decile. The effects are conditional, the control variables for sectors and size classes were included in the model. The whiskers represent confidence intervals. Mean values for Croatia, Latvia and Slovakia – countries with granular data for both periods.
Turning to the intensive margin, as documented in Figure 4, in the first year of the COVID-19 pandemic the amount of support increased almost linearly with firm productivity. Specifically, firms in the highest productivity decile received wage subsidies that were approximately 40% larger than those awarded to firms in the lowest productivity decile.
However, this relationship significantly weakened in 2021. The difference in the size of support allocated to high- and low-productivity firms became negligible, and firms with medium productivity levels received only about 10% more support than the least productive ones. This suggests a significant shift in the distribution of support towards less productive firms.
Figure 4 Size of support by productivity, 2020 and 2021
Source: Authors’ calculations based on the output of the micro-distributed exercise, Lalinsky et al. (2024).
Notes: Effects from OLS regressions representing changes in support for a change in firm productivity decile with respect to the lowest productivity decile. The effects are conditional, the control variables for sectors and size classes were included in the model. The whiskers represent confidence intervals. Mean values for Croatia, Latvia and Slovakia – countries with granular data for both periods.
Conclusions
The main findings of the micro-distributed analysis presented in this column confirm that the allocation of pandemic-related wage subsidies in 2020 was efficient. This means that more productive firms were more likely than less productive firms to receive support; and received more support in absolute terms. However, not only the total volume of support, but also the allocation across firms evolved over time. We find that the link between firm productivity and policy support weakened in the second year of the pandemic. The reason is that more productive firms recovered more rapidly and were, in consequence, quicker to leave the government support schemes, which increased the probability of low-productivity firms receiving support. Our research, based on data for euro area firms, confirms and complements earlier findings of Bahar et al. (2021) suggesting that the support initially protected productive firms in temporary need and contributed positively to aggregate productivity, but as the economy recovered, the supporting schemes became more distortive.
Authors’ note: The results presented in this column are the result of a close and unique cooperation among several country teams from the euro area National Central Banks and the ECB within the WGF Expert group on productivity, innovation and technological change. We namely thank Konstantins Benkovskis, Olegs Krasnopjorovs, Josip Raos and Domagoj Šelebaj for their contributions. Paloma Lopez-Garcia coordinated the WGF Expert group on productivity, innovation and technological change. Tibor Lalinsky coordinated the Workstream on COVID-19 and productivity and all micro-distributed exercises.
References
Bahar, E, J Hambur and D Andrews (2021), “COVID-19, job retention schemes and productivity: From supportive to distortive”, VoxEU.org, 30 September
di Mauro, F, T Bighelli and T Lalinsky (2021), "Covid-19 government support may have not been as unproductively distributed as feared", VoxEU.org, 19 August.
Coeuré, B (2021), “What 3.5 million French firms can tell us about the efficiency of Covid-19 support measures”, VoxEU.org, 8 September.
Hale, T, N Angrist, R, Goldszmidt et al. (2021), “A global panel database of pandemic policies (Oxford COVID-19 Government Response Tracker)”, Nature Human Behaviour 5: 529–538.
Lalinsky, T, M Anastasatou, S Anyfantaki et al. (2024), “The impact of the COVID-19 pandemic and policy support on productivity: a report by the ESCB expert group on productivity, innovation and technological change”, ECB Occasional Paper Series No 341.
Rodano, G, E Sette and M Pelosi (2022), “Zombie firms and the take-up of support measures during Covid-19”, VoxEU.org, 4 May.
Van der Wielen, W, D Revoltella, L Maurin, R Pál and P Harasztosi (2021), “Firm-level policy support during the Covid-19 crisis: So far so good”, VoxEU.org, 18 November.
