On 23 June Prime Minister Boris Johnson set out further changes to lockdown measures in England. Among other things, such as pubs, restaurants and hairdressers being able to reopen, providing they adhere to COVID Secure guidelines, he announced that where it is not possible to stay two metres apart, people can keep a social distance of ‘one metre plus’. This means staying one metre apart, plus adhering to mitigations intended to reduce the risk of transmission.
Johnson and other officials and experts have used similar, but slightly different, language to explain what the risk of infection will be as a result of this easing. When he announced the reduction of social distancing from two metres to one metre, Johnson claimed that the mitigating measures would mean that the overall risk will end up being ‘broadly equivalent’ to the current level of risk.
In the 5pm briefing, Sir Patrick Vallance, chief scientific adviser to the Government, responded to one question by asserting that the risks under the new measures would be ‘equivalent’ to the current level of risk, and used the phrase ‘same risk’ in response to a subsequent question. Professor Chris Whitty, chief medical adviser to the Government, used the phrases ‘broadly similar’ and ‘broadly equivalent’.
We speak to Graham Loomes, professor of behavioural science at Warwick Business School, who researches the valuation of health and safety measures, to find out his views on the new system and what it means for those of us trying evaluate risk.
Graham, you have said that you find it hard to believe that the overall risk from Covid-19 will be ‘broadly equivalent’ to the risk before the new measures. Why is that?
I am doubting it, because when you look at the documents that the announcement last Tuesday refers to, from the Environment and Modelling Group of SAGE, what you see is that they’re pretty confident that the reduction in distance from two metres to one will do something between doubling the risk and multiplying it by ten times. They don’t know within that range what the right figure is, but they appear to believe that the distance change itself at least doubles the risk.
Then there’s this reference to mitigating measures. But again, if you read those documents, there are very few mitigating measures for which there’s much evidence of effectiveness. It doesn’t necessarily mean they’re not effective: it’s just that they have little idea whether they’re effective or not, or how effective they are. And the measures that the experts have more confidence in (they do a confidence rating on different mitigating measures) are things that are mostly already being done. So there aren’t many new measures to offset the extra risk of going from two metres to one metre. When you put all that together, the idea that the risk at the end will not really be any different from the risk as it is now at two metres, just doesn’t ring true.
Do you think the language used will impact the public’s ability to evaluate risk?
I know people find it difficult sometimes to think about probabilities and risks, but my feeling is that if you try and give the impression that the risk under these new measures is going to be about the same as it is at the moment – the ‘don’t worry your little heads about it’ kind of attitude – which is what I took from that press briefing last Tuesday, then it may mean that people are not being as wary as they should be, especially those who are most at risk in terms of the consequences.
If you don’t give people the correct, or your best judgement of, the information about risk, then you’re not allowing them to make what for them might be the appropriate decisions about how they trade-off the risks and benefits.
People make judgments and trade-offs when it comes to health risks all the time. Does the coronavirus present new challenges to evaluating risk?
I’m not sure that we’re terrifically good at evaluating risk. Often we don’t really know what the risks are. I might choose to drive from A to B rather than taking the train. I understand that the risk of being killed or injured is higher in a car journey than the equivalent train journey, but the car is so much more convenient in various ways so I’m going to take the risk. As a society, we accept that in a normal year something like 1,800 people will be killed and another 25,000 seriously injured in road accidents, because we think that’s a price worth paying for the convenience of driving our cars.
Of course, we still take steps to try to reduce those risks. And because we know a lot about them, we can make fairly good estimates of how much different measures will change the risks. By contrast, because Covid-19 is a new phenomenon, the databases are very limited and the uncertainties can be used to accommodate a convenient political message. Because we don’t have the data to do a statistical test of the net effects, it is hard to rebut the assertion that the two different policies are broadly equivalent. However, on the basis of what evidence is reported in the EMG documents, and relying on the collective wisdom of 14 members of that expert group, it really doesn’t look as if the idea that extra mitigation measures will cancel out the effect of reducing the distance is anywhere near the most likely correct one.
The more likely scenario is that mitigation measures will only partially offset the reduced distance so that the infection rates will go up compared with the present rates. And that will impact particularly on elderly people and those with certain underlying health problems – although, because the prevalence rate is now quite low, the actual numbers of those extra deaths and illnesses may not be very large compared with what we have experienced so far.
My point is twofold. First, as individuals, we need the best information, undistorted by political wishful thinking, in order to make the best judgments we can about whether to go to the pub or to visit our friends and relatives. Second, as a society we have to face up to a difficult dilemma. On the one hand, there will be tens of millions of people who will be made better off as a result of easing the lockdown. On the other hand, there will be several millions of elderly and more vulnerable people who will be potentially worse off by being more likely to become seriously ill or die from the infection. How do we balance those different groups and different impacts? How do we trade those two things off, in the way that as individuals, we might trade off the extra risk of being killed in a road accident against the extra convenience that we get from using our cars?
How do policy-makers usually make those kind of trade-offs?
There’s a thing called the UK Treasury Green Book, a guide, which basically says that in all areas of government where decisions may impact upon health and safety and life expectancy, then you should undertake some kind of reasonably formal cost benefit analysis. So in road safety projects, if you expect over a period of time to reduce the number of fatalities as a result of road safety measures, then that’s a benefit and we put a monetary value on it to weigh against the economic costs. The Department for Transport, roughly speaking, puts a money value of 1.8 million pounds on every expected fatality that’s prevented by some new road safety measure. And in the healthcare system, the Department of Health seems to be going with £60,000 for each ‘quality adjusted life year’.
These numbers are derived from a mix of judgement and past surveys of members of the population about the trade-offs that they would make between money and risk.
But as far as we can tell, the responses to the pandemic – in particular, the lockdown from 23 March and associated measures – have not so far been appraised according to those standard approaches. You might think that’s understandable, given the speed at which the disease was spreading and the fears that the NHS would be overwhelmed. But you might also think that now, when we’re considering measures for relaxation, maybe we should be going back to this guidance and see whether the new measures are justified by the same sorts of values and trade-offs in other areas of public policy. It’s very possible that, even if infection rates are increased by the new package of measures, they do pass this test, because the economic benefits of relaxing the restrictions may be huge, compared with the Green Book values when applied to the extra illnesses and extra deaths.
But what I perceived going on at the press briefing was a dodging of being upfront about these trade-offs, glossing over the more likely scenario of increased risk. And I just think that needs to be questioned. People need to think about the fact that there are real trade-offs to be made, and consider how they feel about the sort of values that are involved in making those trade-offs.
Does the behaviour of humans during the pandemic open up new areas of research?
Yes. It can be sometimes be difficult to generalise about people’s reactions to the various measures and behavioural scientists are interested in that. My colleagues have been doing some work, for example, on people’s attitudes to wearing masks in different situations. That could go one of two ways. One way it could go is if you’re wearing a mask, you think it’s safer and okay to be a bit closer to others. The other way it could go, is you say to yourself: if everybody’s wearing masks, this may be a signal of a really dangerous world, so I’m going to stay even further away. So there are two possibilities: in one you get closer, in the other you stay further away. The reality is that some people will think one way and other people will think another way and behavioural researchers are interested in seeing what factors might explain those differences.
Another question relates to the extent that people comply with messages or instructions depending on how much they trust the people giving those instructions. As reported in an Ipsos-Mori survey published in 2019, politicians in the UK – and around the world – are the least trusted profession. Does this mean that people discount or ignore messages from politicians, especially if they don’t like what they are being told to do? On the other hand, that survey showed that scientists and doctors are the top two professions in terms of trust, which may be why the politicians liked to have them alongside at the press briefings and constantly referred to their policies as being ‘guided by the science’.
If people are more likely to act on the basis of sources they trust, that is all the more reason why the people representing science and medicine should try to give their best judgments of things such as the impact of the new measures on infection rates, independently of any pressures to use forms of words that are more comfortable and convenient for politicians. On the other hand, if people perceive that the scientists are trimming their message to make life easier for the politicians, trust may be damaged and future messages may be discounted.