Greek Business File, June-July-August 2020, No 126


Inevitable as it was, the lockdown bought out precious time for the exit strategy from the dangerous and unprecedented spiral in the economy and the society caused by the pandemic. Apart from a massive, rapid and eff ective government intervention, the next step is to cleverly test the traces of the virus using different yet reliable methods. Andrea Galeotti and Paolo Surico both Professors of Economics at London Business School, in this interview for GBF, present their economic analysis of health policies.


We have got to know the Covid-19 disease as a (WHOdeclared) pandemic. But isn’t it rather a diff erent case at each country? Can you briefly tell us what is happening to the World Economy?

Is anybody really confident worldwide about the economic implications of the Covid-19 pandemic even at a tentative level? Do you think that the supply side (dislocation of supply chains, closingdown of production) or the demand side (emotional eff ects, decline in confidence, and thus in spending) will suffer the main economic shock?

Covid-19 can be seen, initially, as a shock in supply. But this shock feeds into household panics generating a contracting demand, which then reinforces the supply shock, creating unemployment – therefore contracting household income and, as a result, demand. It is a very dangerous and unprecedented spiral. It is a spiral that generates cash fl ow problems, to certain important segments of the economy: small/medium fi rms and most of the households. If it persists, it can spread further in the economy.

New business applications have fallen by more than 50%, unemployment claims excess 22 million and GDP contracted almost 5% on an annual basis in Q1 in the US. And this is just the economic costs associated with social distancing and lockdown measures. The toll is very likely to exceed 10% unless government policies manage to break the cash flows short circuit between small firms, their employees, consumption and product sales.

Massive, rapid and effective government intervention

A liquidity-enhancing approach or a structural-aid one? What is the best way forward as far as public policy is concerned to deal with economic disruption?

It is clear that the only way to help the spiral effects described above is a massive, rapid and eff ective government intervention that sends cash to the affected economic actors.

Given your ongoing work, some with real-time data covering millions of UK households, can you tell us what is happening now?

The fall in consumption preceded the social distancing policies, suggesting that part of this could be behavioral in a mixture of panic and precaution. Services have dropped by 50% relative to April 2019, driven by restaurants, retail and transport. Online groceries and DIY/home equipment are on the rise. Many borrowers struggle to honor their payments and the mortgage market has frozen. Income has dropped for many households, with a median decline around 30%. Consumption and income inequality have dramatically risen, as the most economically vulnerable groups experienced the largest expenditure and income drops.

A strict lockdown, the only alternative

What have been the best policies so far on the health-policy front?

For European countries, it seems that it was inevitable to enter a strict lockdown, as countries were not prepared for this epidemic. When governments started to act, the virus had already substantially spread in the population. At the same time, hospital capacities were limited. The only alternative was a strict lockdown. There are exceptions, like Sweden, but these cases are mainly dictated by a much less dense population and other aspects that may allow Sweden to act differently.

Prepare for the exit strategy

Although inevitable, it was clear that the lockdown’s main point was to buy out time in order to get ready for the second phase: the reopening. So, to us, the best way to evaluate the “best policy” across countries is to see the way in which diff erent countries used this time to prepare for the exit strategy.

In this sense, we can see that Germany has, since the beginning, ramped up testing, and this allowed to direct tests not only to symptomatic individuals, but also to asymptomatic individuals. This has allowed Germany to have a much more comprehensive picture of how the virus has diffused in the population, and to stop the spreading much faster. Having good and reliable information is  what it allows to make choices in the exit strategies which are more effective.

Italy provides a very diff erent example. Despite testing is now more pervasive, there is a clear lack of coordination across different regions, and there is no clarity on what is the strategy that Italy will use to keep the infection low. The only strategy seems to keep the lockdown as much as possible.

It is difficult to say what is happening in UK, because at the moment there is not much information about what is guiding policies and what are the principles behind what and when to reopen.

The only way is to test cleverly

Is it right or even useful to discuss better or worse policies when the second wave or further waves of Covid-19 are to be expected, since less-infected countries due to stringent lock-downs will have a higher percentage of their population exposed and less of their population with immunity? When is the right time to do an evaluation?

We think it is important to make comparisons across countries of what policies are working and what policies are not. Eventually, this process of infection will only end when the population has reached herd immunity. This can be because of a vaccine. Or because people will get eventually infected. This is an epidemiological fact. So, the point is to understand what the best way is to shape this fl ow of infection in a way that countries have the capacity to deal with it (in particular, their respective national health systems).

It seems clear already now (and probably it was clear also at the beginning of this pandemic) that the only way to do so is to cleverly test to fi nd where the virus diff uses, and then intervene in a targeted way through isolation. To achieve this you need: 1. technology that helps the tracing, 2. clear information to people so that they engage with the technology and the strategy, 3. clear incentives for people to engage with the technology and the strategy, and 4. centralized decision-making that helps coordination of all those actions. South Korea is the main example of a successful strategy. The South Korean model can be replicated in Europe, it is a political choice to do or not to do so.

Looking at practices followed, can we really say that a test-test-test policy is the best approach? And is there consensus in the scientific community about the way testing should take place? For example, shall it be focused or random?

In an ideal world, we would have a test that each of us can take in the morning and that is fully reliable. That would be (almost) as effective as a vaccine. But we have neither. Tests are a scarce resource, there are different types of tests that provide diff erent information, tests are not fully reliable, and the outcome of a test may inform the policy maker for different objectives. Each of these elements determines the best way to utilize a portfolio of tests. So, there is not a universally optimal way to test. But, for a given problem, we can develop what is the best way to use the available tests.

For example, if the objective is to learn about the stock of infection, that is the number of people that have been infected since the beginning of the epidemic and possible correlations with characteristics like age, sex and alike, then one should use serological tests, because those will reveal the presence of antibodies. Ideally you would test all the population. But since that is unfeasible, you want to test in a way that you can use the collected data to do inference about all the population. For this case, a properly constructed representative sample of the population seems to be the best alternative.

If, in contrast, you use testing for contact tracing, then, testing should far from random. You want to test people that have been close to people that have been found infected. This sample of people you want to test is almost the opposite of a random sample. Again, diff erent objectives will impose different strategies.