VoxEU Column COVID-19 EU policies Europe's nations and regions

Some micro/macro insights on the economics of coronavirus. Part 1: Impact assessment and economic measures

Economists have long argued that there is a conflict between the need to make individuals and companies accountable on the one hand, and the need to share risks effectively on the other. This column, the first in a three-part series, argues that in the context of COVID-19, the reluctance to share risk based on ‘moral hazard’ has no reason to exist, and discusses how the socialisation of losses could be implemented.

We are both economists – one specialised in decision theory under uncertainty, with applications to finance, macroeconomics, climate, insurance, risk prevention and public policy evaluation; the other in development economics, with applications to infrastructure, public procurement, and institutions. Of course, there isn’t any legitimate specialist on the economics of COVID-19, and we make no such claim. Amid a health crisis, there are still many unknowns on the human, medical, economic and financial levels. And yet, decisions must be made very quickly. Scientists have useful but partial and incomplete knowledge. They have a responsibility to share it with the public and with decision-makers, while acknowledging their doubts and the uncertainties that surround this knowledge. This puts the scientists involved as much as public decision-makers in a very uncomfortable situation, with the quasi-certainty of being later criticised by citizens who are often subject to hindsight bias (i.e. judging the optimality of a past decision on the basis of information that was not available at the time the decision was taken). Already today, this bias can be heard. It will be worse tomorrow. Let us anticipate a further deterioration in the confidence of the people in their elected representatives.

The COVID-19 crisis has no equivalent in modern history, neither in its intensity nor in its treatment. The Spanish flu killed between 50 and 100 million people between 1918 and 1920, but the conditions at the time did not lead to a containment strategy at the state level. Asian influenza (H2N2) is estimated to have killed 2 million people between 1956 and 1958, in a context where the international surveillance network was still poorly developed. COVID-19 appears to have a spread rate and mortality rate much higher than that of influenza. In the ‘laissez-faire’ policy invoked at one time in the UK and the Netherlands, for example, some epidemiologists referred to a scenario involving contamination of 70% of the population and a mortality rate of 2% among those contaminated, implying a death rate of 1.4% of the population. For France, this would result in an excess mortality of almost 1 million people. We are no more able than you are to judge the reliability of such an estimate, and its confidence interval is probably considerable. In the absence of certainties about the various parameters of the epidemic, we will have to continue to make choices considering these doubts and uncertainty.

Impact assessment

Of course, we hope that health policy will make it possible to significantly reduce this apocalyptic toll. Let us consider, for example, the containment policy implemented in France since Tuesday 17 March. Labour is, along with capital, the source of wealth creation. For an isolated individual (‘Robinson Crusoe’), not working means not producing and not consuming. What is true at the individual level is also true at the collective level. We cannot distribute wealth that we have not produced. Confinement leads to a degraded version of work, and often to the complete cessation of production. Fortunately, teleworking allows may to maintain a value-creating activity, but it is still very difficult today to measure its impact on the activity. 

The duration of this decline in production is also unknown. In order to fix ideas, let us start from a quick estimate ‘from the corner of the table’. Let's imagine two months of firm confinement implying a halving of economic activity. This leads to 1/12 of no wealth creation, i.e. an 8% loss of income. If we add an only gradual return to normal by the end of the year, we easily reach a loss of more than 10% of GDP due to the containment policy. The financial markets seem to share this estimate, with a fall in prices of around 40% at the time of writing. Since, in our capitalist society, the equity capital of companies is at the front of the line to swallow this massive loss of income, these two measures are compatible with each other. Again, keep in mind that these estimates of health and economic impacts are highly uncertain, although they are consistent with the historical evidence (Barro et al. 2020). We are only at the beginning of the learning curve.

A 10% drop in GDP is certainly not insignificant, but if it were fairly distributed over the entire population, it would not mean the end of the world. Above all, it will certainly be temporary. Indeed, once the confinement has been lifted, the population will return to work if economic policy is able to avoid a cascade of business failures in the meantime. In this case, we can even imagine a rebound in growth in 2021, with companies seeking to replenish their stocks that were damaged by the production stoppage this year, and consumers making the purchases they were unable to make during the confinement. By way of comparison, the Greeks have lost around 30% of their annual income over the last decade, and a return to the pre-euro crisis era is not expected for a long time. France lost 16% of its annual GDP during the 1929 crisis – 31% because of WWI and 49% because of the WWII – with the shock lasting much longer than can be anticipated for COVID-19. Thus, while the current shock is severe, it is not as catastrophic as those experienced by our parents and grandparents.

The state as insurer of last resort

Nevertheless, in the absence of public intervention, the reality will certainly be quite different from a uniform 10% drop in income this year with a return to normal next year. At the household level, it will be far harder on the most precarious, the intermittent, those whose employment is not sustainable or is crucially limited by confinement – not to mention the homeless or refugees. Similarly, some businesses (restaurants, tour operators, etc.) will lose much more, and many of them could go bankrupt. Large companies will also face major financial difficulties. These include airlines, car manufacturers, hotel chains, concert halls, football clubs, and so on. Government officials, online service providers, the food industry and traders in essential services will not lose much. The containment strategy is a necessary collective sacrifice for the common good. This effort must be shared fairly from an economic and financial point of view. It is as much a moral imperative as it is an economic one. Under the veil of ignorance, not knowing whether we are civil servants or restorers, we would all like to see this happen. Ex-post solidarity is ex-ante insurance. Only the state can set up such an insurance mechanism as a last resort. It calls for a systematic socialisation of economic and financial losses due to containment.

Economists have long argued that there is a conflict between the need to make individuals and companies accountable on the one hand, and the need to share risks effectively on the other. Strengthening insurance often means reducing the incentive to prevent risk effectively. But in the context of COVID-19, this reluctance to share risk based on this ‘moral hazard’ has no reason to exist. We are therefore in a very different situation from that of 2008 (the subprime crisis) and 2011-12 (the euro crisis), where the moral hazard argument had an indisputable empirical basis: banks determine the risk associated with their loans portfolio; governments determine the risk associated with the bonds they issue.  The socialisation of losses could be interpreted as an absolution of past bad behaviour, and a license to start again. COVID-19 and the associated containment are a combination of an ‘act of god’ and a political decision, and no actor is able to intervene to prevent them. There can be no stigmatisation of the victims of containment.

How can this imperative of socialising losses be implemented? It is a transfer of any losses linked to confinement of households and companies to the state accounts. The victims must be compensated, and only them. The deferral of social security charges and taxes may be useful in the short term to loosen the grip on the cash flow of SMEs, but it is not an instrument for the socialisation of losses. The ban on redundancies constitutes a transfer of the burden of employee containment to companies. It is viable only in the very short term because its sustainability would risk putting many firms out of business that are unable to maintain their wage expenditure in the absence of a concomitant flow of revenue. 

The situation in France is very different from that in the US, for example, because in France we have a much more efficient and generous social security system. Ideally, it should be possible to have recourse to unemployment insurance, with 100% of the salary being maintained throughout the period of confinement, at least up to a limit. The state should also compensate the loss of income of self-employed workers forced to interrupt their activity by issuing a cheque (or a tax rebate) proportional to the duration of this interruption and based on declared income from 2019. This targeting is in this respect different from the bipartisan solution of the US Congress to send a cheque to all households, which has no economic justification in the current circumstances. This ‘helicopter money’ is only useful if it is a response to a demand shock, which is not (yet) the case. Of course, there is a risk that there will be holes in the racket and profiteers. Economists working on social transfers understand that targeting is never perfect (e.g. Elbers et al. 2007, Ravallion 2016), but the urgency and intensity of the crisis justifies some losses down the line.

Compensation of losses by the state is also a way to ensure that this production shock does not additionally generate a demand shock. By maintaining the purchasing power of households, the propagation of the shock over time is cut off. You have to know how to be Keynesian when the situation requires it. The state must also ensure that the industrial apparatus remains intact by preventing business failures. A recapitalisation through temporary state participation in the capital of certain companies may prove necessary. As in 2008, if the economic rebound is confirmed, this could even be done at no cost to the taxpayer.1 This widespread and comprehensive insurance net for workers and firms is currently the absolute priority. Then, as we eventually move out of confinement, the optimal mix of interventions will shift towards less social insurance and more aggregate demand stimulus, of a form that is yet to be defined (Dube 2020).

Debt and self-insurance

Public opinion has now clearly understood the need to flatten the contamination curve. But there is a second curve that needs to be flattened – that of the fall in income following the containment that this first flattening imposes on us.

This socialization of losses will result in a massive public deficit in 2020, perhaps around 10% of GDP, and a corresponding increase in public debt, which will have to be gradually repaid. This policy of socialisation of losses and intertemporal smoothing of the shock is economically desirable. It is well known that the uncertainty that weighs on a household's income can be effectively managed by accumulating liquid precautionary savings. This safety cushion is built up in periods of high income to be used in periods of low income. This simple self-insurance mechanism has always been used by those households that can, but is only possible to a limited extent for the more modest because of the credit constraint. The latter are often forced to pass on one-to-one the loss of income in loss of consumption, with potentially dramatic consequences if they are unable to borrow.

What is optimal for a household is equally optimal for a state. Its policy of capital accumulation (infrastructure and public participation) and indebtedness reflects this need for intertemporal smoothing of macroeconomic shocks. Unfortunately, in France, the Treasury's wiggle room is reduced, and the state has few assets that it could dispose of to finance this shock. Moreover, given the level of stock market valuations (the CAC 40, for example, has lost nearly 40% in one month), it would be unwise to dispose today of Aéroport de Paris, EDF or the assets held in the Pension Reserve Fund. For decades France has agreed to let deficits slip through during recessions without ever reducing its debt in better times. In this sense, democracy is the dictatorship of the present. This indiscipline has been reinforced since the creation of the euro and the concomitant end of the mechanism of automatic punishment of this indiscipline by the devaluation that the financial markets usually inflicted on it. Let us recall the Greek crisis. If one day we reach debt levels that could tempt a future government to exit the monetary union and default, the risk is that the anticipation of such an event could lead to the country's inability to refinance its debt or cushion a new shock, with dramatic social consequences for the country. 

There is obviously no point in dwelling on this problem of financial stability and rationality of public spending in the present circumstances. From this point of view, the suspension of the budgetary discipline rules of the European Treaties is welcome. However, this problem of the inability to balance the public budget over a long period of time will have to be dealt with one day. Why should France be eternally incapable of doing what Germany has been doing for decades? This is undoubtedly one of the major macroeconomic issues that will be on the table after this crisis.

The first consequence of this smoothing of the shock by public debt concerns its measurement in terms of GDP in the short term. The public transfer policy in favour of the victims of containment will mean that the fall in GDP could be reduced to virtually nothing in 2020. The bulk of the GDP loss would be carried forward to the following years, when part of the wealth creation would be neutralised to repay the ‘corona debt’ rather than being translated into consumable income.

Fortunately, France is not at the forefront of this exposure to sovereign debt risk. Italy has the misfortune of being the European country most affected by the pandemic. It is also one of the most indebted European countries. The exogenous nature of the economic shock also removes here any problem of moral hazard and the stigma associated with it. It is the whole of the EU that should socialise the corona loss on our continent (Baldwin and Weder di Mauro 2020). If the Union does not express this solidarity in the context of this pandemic, it will lose much of its remaining credibility as a political entity. This is all the truer since the firm policy of containment in Italy has had a beneficial effect on the rest of the Union. This loss should therefore be shared by issuing a European corona-bond with joint responsibility of the member states for its reimbursement (see, for example, recent proposals by Giavazzi and Tabellini 2020 and Garicano 2020). Failing this, the ECB should ensure that the conditions of indebtedness of the states do not diverge within the Union. The rule capping purchases of sovereign debt under the European Stability Mechanism should be exceptionally suspended (Bénassy-Quéré et al. 2020). The recent widening of government bond yield spreads within the Union should be contained. The ECB's launch of the €750 billion Pandemic Emergency Purchase Programme (PEPP) at least partially addresses this concern.

In the short term, the ECB must also avoid a liquidity crisis by offering cash to all financial institutions that request it. This is so that the latter can in turn finance solvent companies facing difficult maturities. The socialisation of losses by governments should help reassure banks about the solvency of their borrowers. Trust and credibility are key here.


Baldwin, R and B Weder di Mauro (2020), Mitigating the COVID Economic Crisis: Act Fast and Do Whatever It Takes, VoxEU ebook.

Barro, R J, J F Ursúa and J Weng (2020), “The Coronavirus and the Great Influenza Pandemic: Lessons from the "Spanish Flu" for the Coronavirus's Potential Effects on Mortality and Economic Activity.” NBER Working Paper No. 26866.

Bénassy-Quéré, A, A Boot, A Fatás, M Fratzscher, C Fuest, F Giavazzi, R Marimon, P Martin, J Pisani-Ferry, L Reichlin, D Schoenmaker, P Teles and B Weder di Mauro (2020), “A Proposal for a Covid Credit Line”,, 21 March. 

Dube, A (2020), “Step One in Fighting the COVID-19 Economic Crisis.” Boston Review, 23 March.

Elbers, C, T Fujii, P Lanjouw B Özler and W Yin (2007), “Poverty alleviation through geographic targeting: How much does disaggregation help?”, Journal of Development Economics 83: 198–213.

Garicano, L (2020), “The COVID-19 bazooka for jobs in Europe.”, 20 March.

Giavazzi, F and G Tabellini (2020), “Covid Perpetual Eurobonds: Jointly guaranteed and supported by the ECB.”, 24 March.

Ravallion, M (2016),  The Economics of Poverty: History, Measurement and Policy, Oxford University Press.


1 It should be noted that the policy of maintaining demand while production and supply will fall this year may lead to a temporary surge in inflation. Such a surge in inflation would be rather welcome in Europe.

26,141 Reads