Data and public policymaking: The case of the coronavirus pandemic

Posted in: COVID-19, Data, politics and policy, Economics, Health, Science and research policy, UK politics

Dr Matt Dickson is Reader in Public Policy at the Institute for Policy Research (IPR), University of Bath. 

The current global COVID-19 pandemic has highlighted in a very stark way the importance of data for public policymaking. It’s always been the case that policymakers require accurate information to guide their choices, but this is even more important when people’s lives and livelihoods are at stake. The mantra from the World Health Organisation has been “TEST, TEST, TEST” but it could equally be rephrased  “DATA, DATA, DATA” as the key output from testing is the data on who has the virus.

Different countries have been taking different approaches to dealing with the outbreak but a common feature of those who have so far been the most successful is both the high use of testing to generate reliable incidence data, and also the utilising of as much relevant data as possible from other sources to understand the spread of the virus.

South Korea is the exemplar: they have been testing high numbers of people since they had their first confirmed case of the SARS-Cov-2 virus on 20 January. Testing 20,000 people with or without symptoms per day in rapidly assembled drive-through testing stations, the authorities then isolated those testing positive and used all the possible data available to them to trace the contacts of those infected and test them, isolating the positives and tracing their contacts, carrying on in this way to try to limit the transmission of the virus. The test data is complemented by mobilising many other sources of data from their smart cities including mobile phone location data, credit/debit card transactions and surveillance camera footage of confirmed coronavirus patients, to recreate their movements so that they can notify all people who they may have come into contact with. These contacts are informed directly by the government that they need to be tested and the (anonymised) information is also made public – an app alerts users when they come within 100 metres of a location visited by an infected person, and a coronavirus map website shows where infected people have been, allowing individuals to take precautions themselves and avoid high-risk areas for infection. In this way, despite being the epicentre of the global pandemic for a short time in late February, South Korea managed to get the virus under control without the need for a country-wide lockdown. As of 8 April there have been 200 deaths from COVID-19 in South Korea.

In contrast, the US, who also reported their first case on 20 January, were slow to get effective testing kits distributed, allowing the virus to spread uninhibited through the population for several weeks, which as we have all been learning of late, leads to exponential growth of infections with such a highly transmittable virus. Many more tests are now being carried out across the US, where almost all states are in full or partial lockdown. However, as of 8 April there have been 12,857 deaths from COVID-19 in the US. Direct comparison of countries as different as the US and South Korea is difficult, not least because the US population is that much greater at 327 million compared to 52 million. However, the difference in deaths remains stark: 39 deaths per million people in the US, just 4 per million people in South Korea.

The UK reported its first case on 31 January but, like the US, is still yet to roll out the sort of wide-scale testing of people – whether they are suffering symptoms of COVID-19 or not – that has taken place in South Korea. At present only those who are hospitalised with COVID-19, and key NHS workers, are being tested. Moreover, we have not instigated the contact tracing of those infected - that is the second element of the most successful responses to the outbreak. There are legitimate civil liberty concerns around whether the government should be accessing all of our personal data from mobile phones, digital transactions and surveillance cameras, even if we had the advanced smart cities of South Korea. However, there are more traditional methods for contract tracing that may not be as efficient but could still be applied and impinge less on our civil liberties.

In the absence of testing and tracing, the rapid and widespread transmission of the virus in the UK has meant that, like many other countries, we have been forced into lockdown with schools and university campuses closed and citizens required to stay at home, with only limited numbers of people allowed to continue to go to work. This lack of data on who has the disease and who they have been in contact with means lockdown is the only option to stop the virus spreading and limit the number of deaths.

Without widespread testing, we cannot know how many people are infected or how many people have been infected and have recovered (and are potentially now immune). Because of these two things, we cannot accurately predict the true population case fatality rate (CFR) for the COVID-19 disease. As such, the pandemic is now developing to be a data crisis as well a health crisis. Like every other  country, we do not have reliable data on the prevalence of the virus in representative random samples of the population. Until we can carry out tests on such samples and re-test them regularly, we will not know how widespread the infection really is and whether the current measures are limiting new incidence. Moreover, we only currently have the fatality rates amongst those who have been tested – but due to the current testing regime these are mainly the people admitted to hospital with severe symptoms biasing the CFR up. Without more accurate estimates of the true CFRs by age-group we cannot know how risky it is for different people to catch the virus.

This last point highlights that testing is not only the key to knowing how deadly the disease is, but it is our only way out of the current lockdown without having to wait the 12-18 months it is estimated it will take before a vaccine might be commonly available. The lockdown is a crude measure that buys time but in itself does not solve this crisis, only further data can help us to do that. Economists are, and should be, thinking about this in terms of the cost-benefits of the continued lockdown and how we might start to get out of it.

There are huge costs to the economy of shutting down all but the essential services, with estimates ranging from 6% to 20% of GDP. The unprecedented intervention of the government to pay 80% of the wages of those put on furlough alone may end up costing tens of billions of pounds. However, we also need to take into account the welfare costs of everyone having to be locked up in their homes for most of the day and of not being able to see friends and family, go to the cinema, theatre, swimming pool, play sports, watch football and many other activities. So there are huge welfare costs of the lockdown, and this is before we even consider the impacts on mental health and wellbeing.

Balanced against this are the huge costs of lost lives. The government values a single life saved at approximately £1.8m (although critics believe this is set way too low). Therefore, saving the 250,000 lives that would potentially be lost should we not lockdown, would be valued by the government at £450 billion (about 22% of GDP). However, if we are counting the welfare losses from lockdown we also need to count the welfare losses from grief, and though the psychological cost of losing 250,000 people cannot easily be computed, it is clear that by any measure it would be unbearably high for any society.

The costs and benefits of the lockdown are also changing everyday: costs are going up for certain as more firms are likely to go bankrupt, leading to further job losses, plus the welfare losses of everyone remaining at home are also increasing day by day. But again, without accurate data on the CFR and on infection prevalence we cannot accurately quantify the benefits.

So while a comprehensive lockdown is definitely the right thing to do given so much uncertainty, the key questions are: how long can we lockdown (without catastrophic impacts on the economy and on the mental health of the nation) and how do we exit the lockdown? At the moment we don’t have good enough data to answer either of these questions.

We don’t know, for example, how long people can be limited in their freedoms before large scale disobedience occurs, with potentially devastating consequences for the spread of the virus and mortality. Already we see breaches of social distancing instructions when the weather hots up, particularly at the weekend. How long will people continue to follow the guidance, especially if they think they are in a low-risk group?

On the exit strategy, we can start to think about the marginal impact of measures to ease lockdown – what happens to infection transmission and economic prosperity if we change a single variable by a small amount, and what are the risks involved? If we can find low-risk ways of reducing the costs of lockdown especially for those in lower risk categories – provided that those in higher risk categories remain well protected – we might be able to start limiting the economic damage. What might some of these things be?

One proposal from economists at Warwick University is for young people aged 20-30 and not living in their parents’ home to be ‘released’ from lockdown to get back to work – increasing the number at work by 4.2 million people whilst limiting the risk of the infection spreading to the most vulnerable. Other proposals are for businesses to re-open with temperature testing before entry to the building, and re-shaped working practices to ensure social distancing remains in place. The government is said to be giving serious consideration to the idea of immunity passports for those who have recovered from the virus, allowing them to re-enter full participation in the economy and society.

These debates amongst economists and between economists and others are likely to become much more prominent in the coming days and weeks as we seek a lockdown exit strategy that balances the risks, costs and benefits of potential actions.

The immunity passport idea undoubtedly has its issues and moral hazards but it again highlights the requirement of further widespread testing – this time for antibodies to the disease – which would allow immunity to be established.

Coming full circle we are back to where we started: what we urgently need for any of this to become feasible and public policy to be made, is DATA, DATA, DATA.

Are you a decision-maker in government, industry or the third sector responding to the coronavirus crisis? Apply now to our virtual Policy Fellowship Programme for access to University of Bath research and expertise. Learn more

All articles posted on this blog give the views of the author(s), and not the position of the IPR, nor of the University of Bath.

Posted in: COVID-19, Data, politics and policy, Economics, Health, Science and research policy, UK politics


  • (we won't publish this)

Write a response