Every economic crisis raises the same normative question for competition law. Should decision makers be temporarily more permissive in their application of the law to private and public restraints of competition? While historical evidence suggests that this is a bad idea, most economic crises since the 1970s led to some softening of competition law.


In countries around the world, massive amounts of state aid have been injected into the economy. While such policies deserve praise in their concern for the protection of jobs, recessions have a “cleansing effect” which is desirable and can be dampened by such interventions. Recessions facilitate the exit of zombie firms that crowd out growth opportunities for more efficient competitors, and delay the diffusion of technological innovation. A case might thus be made that the current recession might be a source of opportunities for the EU economy, long trapped in a cycle of weak productivity, low economic dynamism and conspicuous absence of superstar firm creation.

The authors  of this paper, available here, argue that an optimal competition policy reaction to the crisis should reduce barriers to the exit of inefficient firms that prevent industry reorganisation. This might justify a more proscriptive approach towards State aid, and a more permissive policy towards mergers.

A first section describes the cleansing effects of recessions.

Recessions produce a cleansing effect by substituting inefficient production units with more efficient ones. Several established models describe the mechanics of the cleansing effect. The dominant theory is one of natural selection. According to the theory, less productive firms are the first to turn out to be unprofitable and scrapped in a recession. The fall in demand will raise the returns of the most efficient incumbents relative to the industry laggards, precipitate the latter’s exit, and end up increasing productivity and growth. A second influential theory finds that the cleansing effect of recessions is due to changes in opportunity costs. In times of recession, the returns generated by investments that raise short-term output decrease as a result of falling demand. Firms thus find it relatively less costly to invest time and divert resources towards organisational efforts that improve long-term productivity, instead of spending resources addressing non-existent demand. Those firms with a superior stock of human capital prosper and grow, whereas those which did not build the right skills in the good days perish when the recession arrives.

An economy can only grow effectively out of recession if exit barriers do not unduly protect inefficient incumbents. What counts from a social welfare perspective is not the number of firms that remain in the market, but the number of efficient firms that compete effectively to meet demand, regardless of market shares or concentration. However, competition law and economics continues to rely on metrics of industrial concentration, biasing competition enforcement against exit by agreement or mergers, and potentially biasing enforcement towards State aid that prevent further increases in industry concentration. However, firms should become a priority of competition policy, particularly in times of crisis.

The second section explains why these cleansing effects matter for the European economy.

The EU has a productivity problem. Since the mid-1990s, total factor productivity performance and labour productivity growth have been lacklustre compared to other advanced economies. The mechanisms that underlie Europe’s weak productivity performance are well-known. Entry and exit of firms is low, preventing the reallocation of labour and capital towards productive activities. Stringent product market regulations arguably play a key role in explaining the low exit rates of large European firms. Further, the EU has not witnessed growth in the high-tech sector comparable to the US, and missed out on “superstar firms” which raise general and sectoral productivity, as well as provide incentives to other firms to invest and to innovate through a variety of channels, e.g., corporate venture capital, knowledge spillovers, and platform leadership. Additionally, Europe has a well-documented problem of technology diffusion. While innovation levels are comparable to other advanced economies, the EU is slower to incorporate state-of-the-art innovation into the production processes of businesses. Last, the fallout of the 2008 financial crisis, including very low interest rates, has seen a cross-sectoral rise in “zombie firms” in Europe.

With this background, the cleansing effect potential of Covid-19 might provide a unique opportunity to address part of the EU productivity problem in the short term, compared to costly initiatives like labour, product market or bankruptcy reform. But a more tempting political response might consist—in protecting inefficient firms subject to a flexible application of State aid control rules, especially if they are too big to fail. This will limit the growth potential of the EU economy for years to come.

The third section considers how competition law’s reaction to the crisis can promote economic growth.

Competition policy can play a role in mitigating Europe’s productivity problem. At the policy formulation level, this requires a recognition of the consumer welfare benefits arising from the cleansing effect of recessions. At the operational level, this requires a commitment to assist the exit of inefficient firms, by allowing them to merge with more efficient firms, and by denying them the benefit of State aid when it prevents efficient industry reorganisation or liquidation. Instead, the EU seems to be granting state aid in conditions that make it hard to apply conditionality to distortive aids to zombie companies. On the other hand, the European Commission has also asked firms to suspend merger filings. This may have the unintended consequence of depriving inefficient firms of restructuring opportunities on the M&A market.

This is problematic on several grounds. Firstly, a horizontal State aid policy may be too lenient, by entrenching zombie companies with problems that predate the Covid-19 crisis and that had nothing to do with it. Relatedly, and since the amount of state aid varies by state, this runs the risk of entrenching regional differences in a way that threatens the foundations of the single market. Secondly, the increasing scepticism of competition authorities regarding the pro-competitive effects of mergers may deter the restructuring of sectors of the economy. Firms in distress may be lucky and receive government support—more likely if they belong to rich Member States—or face the prospect of liquidation in a context of reduced demand for assets, further reduced by excluding industry leaders as potential purchasers. Concerns about lax merger policy could allayed by adopting a competitive assessment that discriminates between acquisitions by frontier firms and technology laggards. Being a past or ongoing recipient of subsidies, as well as being a state-owned enterprise (SOE), should count against the merging party in the competitive assessment. Second, merger and State aid approval should be conditional on demonstrable reorganizational and managerial efficiencies, including reskilling plans and innovation incentives for the workforce at all levels.

A fourth section concludes.

In short, the authors argue that every opportunity should be taken by competition agencies to use their relatively uncostly decisional powers under the merger and State aid rules to promote allocation of resources towards firms’ policies that raise productivity, quality and innovation. The EU has the opportunity to restructure its industrial and service sectors, getting rid of many zombie firms that have dragged its productivity and constrained its growth. Protecting those firms may appear necessary to protect employment and avert a social crisis, to maintain labour income and rein in the danger of populism. However, it is not. On the contrary, policies which allow zombie firms to survive cause long-term unemployment and set a brake to productivity and thus wage growth.

Increases in industry concentration resulting from the exit of inefficient firms are a blessing, even if that means a reduction in the rivalry metric. Of course, because each firm’s story is different, this recommendation is only valid on a case-by-case basis. Non-selective rules that grant subsidies regardless of firm type or structural presumptions and other inefficiency possibility theorems against mergers are bound to be socially costly. While it is far from simple to determine when financial distress reflects firm-level inefficiency instead of external factors and bad luck, the authors are willing to bet that inefficiency explains the difficulties of many large firms now queuing to receive support from their governments. At the same time, promising start-ups which would have done well absent the Covid-19 shock will leave the market almost unnoticed without being allowed to sell their distressed businesses to other companies with the necessary capabilities and resources to scale them up.

In short, state aid and merger policy in times of Covid-19 require a “rule of reason” approach, away from rigid “per se” rules. In addition, consumer welfare analysis in times of Covid-19 requires a departure from competition law’s obsession with rivalry, and an unequivocal embrace of efficiency.



This is an extraordinary piece – not only because of the clarity with which it articulates its arguments, but because of the forcefulness with which it goes against the current consensus in policy circles. On a quick reading, I think that while the authors’ arguments have force, they may have made them too strongly – and I suspect they may have done so on purpose, knowing that their article was bound to make waves.

I think it is important to start my reaction to this piece by noting that the authors privilege a model of growth – focused on lower regulation, the promotion of larger businesses, and accepting of higher (non-structural) unemployment in the short run – that is not consensual in Europe. The reason I raise this is because, unlike one might come to believe moving solely in competition law circles, adhesion to certain models of growth lead to different approaches to competition law, at least at the margin. This can reveal itself in approaches to the risks flowing from market power, what should count as an efficiency, what type of mergers should be prohibited or allowed, etc. Less importantly, but perhaps more noticeably, this matters because adherents to different approaches (more ordo-liberal, or more social democratic, or more new-Brandeisians) are likely to shriek in horror at the proposals.

Personally – and I cannot emphasise strongly enough that this is no way reflects the views of my employer – I find the argument that the economy has too many zombie firms is supported by data. However, this is not only in Europe. Large numbers of zombie firms are present in all corners of the globe, and, arguably, the US has more zombie firms than Europe right now. I think it is also generally accepted that recessions are unfortunate events, which nonetheless have the important role of eliminating inefficient companies from the market and re-allocating economic factors to more productive endeavours – and, in this way, are important to promote economic growth and increased productivity. It is also widely accepted that recessions should not lead to state aid – they should lead to firm exit, with the (temporary) negative social consequences being addressed by social stabilisers.

However, it is at this point that I find hard to share the paper’s arguments. In short, I do not think the COVID crisis will lead to a typical recession that weeds out only the most inefficient firms. Instead, it is a brutal exogenous shock that will lead to market exit of any company that does not have the cash reserves to survive a shutdown of the economy. Furthermore, it is an asymmetric shock, in that some sectors are brutally affected and had to shut down, while others continue to operate on a more or less stable basis. I have seen informed estimates that, without support, 40% of all companies would go insolvent within three months. The rate will vary – it could be worse, it could be better – depending on the economic sector and the country.

In such circumstances, to complain that state subsidies are not targeted enough and may risk keeping zombie firms afloat is throwing the baby away with the bath water. Of course, state aid should not support inefficient firms, and such companies should be allowed to leave the market; and of course there is a risk that politically well-connected firms are likely to benefit unduly from state aid, and I think everyone in the competition community would agree that measures should be adopted to prevent, or at least limit this. However, to complain that support plans for the economy do not distinguish between efficient and inefficient firms at this stage strikes me as complaining that the fire truck is going against traffic while the house burns. Quite simply, in the present circumstances the risk of economic shock, mass insolvencies and unemployment and long-term economic hysteresis overwhelm all other considerations. I have the impression that the European Commission made a valiant effort at ensuring in a very short timeframe that the state aid is subject to at least some conditions that make it more difficult for governments to benefit politically cosy firms. Additional attention should be paid to avoid using the emergency as a cover to support politically connected and zombie firms inasmuch as possible. After the emergency is over and the economy reopens, a very large number of firms will fail after State support measures have been withdrawn – at which point the cleansing effect of recessions may well take place, and we may have a more careful discussion about whether and what types of state support may be justified.

I think this failure to distinguish between the current crisis and its aftermath also raises problems concerning the authors’ proposals on mergers, which are likely to prove the most controversial part of the paper. In short, competition authorities asked companies to delay merger filings because they lack the resources to address them properly, regardless of whether they are pro- or anti-competitive. The fact that we are in a crisis does not change the law’s default approach to mergers of a certain size – which is that they must be subject to control. Furthermore, precisely because there have been blanket guarantees, companies are less likely to become insolvent and leave the market during the crisis.

I think that the real shock will come once the economy reopens, state guarantees are withdrawn and competition authorities are able to function again. At that point, that many firms will likely have to choose between declaring bankruptcy and merging with other companies. A paper I reviewed a few of weeks ago made the argument that the best policy would be to make companies file for insolvency, instead of allowing them to merge with other companies. You may remember I had doubts about that.

Ultimately, I think that the best policy may well depend on the scope of the crisis – if so many companies are leaving the market that this will lead to the destruction of productive capacity and mass unemployment, it may be preferable to promote mergers so that companies and jobs survive. On the other hand, I also have doubts about deciding to allow mergers on the basis of the efficiency of the purchaser – it is not only unclear to me how we are supposed to assess this but, more importantly, I cannot see how this should be a relevant consideration from a competition standpoint even if it might promote economic growth. The guidance provided by the authors strikes me as being of little help. For example, excluding companies that received state support might, after the crisis, support companies that had cash in hand or very close connections to lending bodies – none of which relates to efficiency. In effect, such an approach would act against purchases from famously inefficient firms such as Tesla.

Somewhere in these debates, we forget that competition law analysis of specific conduct or mergers operates at the margin, i.e. the question is whether specific conduct or mergers promote competition and consumer welfare. These is no good reason why a purchaser has to be a competitor; or that a merger should be anticompetitive just because the target is in distress, or pro-competitive just because the purchaser is an efficient, dynamic firm. Merger control should take these matters in the round, as it has in the past, in the light of the available evidence. Further, if we have anticompetitive concerns, there are doctrines such as the failing firm defence that look precisely at whether the merger will be anticompetitive at the margin (i.e. because if the firm leaves the market, the competition counterfactual would be similar to the world where the merger goes through). Personally, my concern is that, if the shock is large enough, merger control filings and the number of failing firm defences will be of such magnitude as to overwhelm competition agencies – but this is an institutional concern, and it may well prove to be unjustified.

To conclude, I think this is an important paper – it is crucial, even in crisis, for academics to challenge the consensus. Whatever one may think of it, its arguments should be engaged with so that a more solid approach to the post-crisis is ultimately adopted.

Author Socials A weekly email with competition/antitrust updates. All opinions are mine

What do you think?

Note: Your email address will not be published