Very early in the Covid crisis a wise journalist at the FT (I forget which one) said that scientists were going to learn to behave like economists. Slowly we are learning the truth of that. Almost everybody knows that economists deal with the unpredictable, and treat their advice accordingly, alongside other political arguments. So little is known about Covid-19 and the virus that causes it that we should treat scientists’ advice in much the same way, except when it comes to such important issues as vaccines and treatments.
I have long used this blog to argue that science has severe limitations and does not justify the faith that some put in it. The discipline has, or should have, rigorous processes by which it can express opinions (or hypotheses if you prefer) with increasing degrees of certainty. But only relatively simple propositions can be tested in this way. And the conclusions drawn from the evidence are usually only generalities. Many areas of knowledge are beyond the reach of such rigour. Economics is one example, management and organisational practice is another.
Science remains vital, of course. The search for a vaccine for Covid-19 is an example. It plays to science’s strengths. The proposition that a particular vaccine works a relatively simple one to test, and a general conclusion, rather than one that applies to all people all the time, is what we are looking for. It isn’t worth investing in a vaccine that hasn’t proved itself in the rigours of scientific process.
An example of something where science is of less use, look at facemasks. How effective these are in preventing the spread of the disease turns out to be a very difficult thing to test. There are so many variables. Does that mean we should recommend their general use by the public? British scientists seem be suggesting that the answer is no, though they generally try to avoid saying anything at all. Interestingly, in Asia the view seems to be that they should be worn; doubtless they would say that scientific evidence does not show them to be ineffective either. And the Asians do seem to be doing a much better job of managing the disease.
This is the null hypothesis problem. What do you believe, absent proof? Some scientists like to say “nothing”, as if that was a neutral answer. But there is no neutral answer if choices depend on it. This sort of muddle seems to have been behind Britain’s late lockdown, which has been so disastrous. The government says it was following scientific advice, and there is no reason to doubt their word. It was very early days in our understanding of the virus and the disease, so the scientific advice was not clear – though some claim it was. In Britain the null hypothesis was that a lockdown would be ineffective, so action was delayed. Elsewhere (Asia again ahead, but places like New Zealand too), the null hypothesis was that a lockdown would be helpful, until shown to be otherwise. This was a variation of the precautionary principle.
Now we are in a much trickier situation with regard to how quickly to ease the lockdown. The precautionary principle is a lot less use here, as there are harms on both sides. I have a lot more sympathy with the government this time, and I think they are right to be pushing the envelope, which in turn is opening debate. But they can’t appeal to science if they are looking for definite guidance.
So, if we start to look at the science more like economics, what might we suggest? What follows is based on evidence, but the hypotheses developed have not been tested to anything like a scientific standard. But no sensible public policy is going to be.
What we are learning about the disease is that it is not as contagious as feared at first, but that in some situations it is still highly contagious. Early in the crisis some people suggested that the overall infection rate was very high (over 50%), but that this was not showing up as most people were asymptomatic. In fact, mostly, infection rates have been pretty low (below 10%), where general populations have been tested. But there have been some pretty spectacular exceptions, where a single individual has spread the disease to many. A conference in Singapore was an early example; there was a ski resort in Austria too, and recently a nightclub in Korea; care homes have been a major trouble spot. This suggests to me that enclosed spaces with many people in them are the main danger, as well as residential environments with a lot of people mixing. Further evidence seems to show that hospitals, and perhaps care homes, can be made safe by the use of personal protective equipment (PPE). Hospital workers don’t seem to be suffering more than the population at large.
So if I’m right about that, we need to keep pubs, restaurants and nightclubs closed, except where they can serve customers outdoors. Outdoor activity generally is probably relatively safe, and the 2m social distancing advice here in Britain is probably overdoing it unless you are stopping to talk to somebody. Alas trying to change that advice is probably too much of a communication challenge. Intermediate environments, such as shops and schools, are much harder. Steps need to be taken to keep densities of people low, and my feeling is that facemasks should be mandatory (though that is probably impractical in schools). One particularly important question is air travel. On the face of it an aeroplane is a particularly dangerous environment. But if everybody wears masks? We need more evidence.
Meanwhile testing infrastructure needs to be improved so that we can test more people and get the results quicker. Something seems to be going badly wrong in Britain. Paradoxically I suspect that an obsession with ramping up scale hasn’t helped. Building large scale facilities first to gather samples and then separate facilities to complete the tests looks to me like a process flow nightmare. More testing will give us much more information and if it is combined with effective tracking we can understand better how the disease is passed from person to person.
A further idea worth exploring is that of “red” and “green” zones, with tighter restrictions in the former. The problem of course is of people moving from red to green. Perhaps you could allow restaurants to open in green zones, if all customers are made to leave name, address and phone details (with those living or working in red zones not allowed to go there). You could use a smartphone app to help enforce much of this, though that may be too Chinese for Britain.
Using such policies we can haltingly work our way to a new normal, until the scientists and technologists give us a vaccine or cure. Economics is no science, but it is not useless either. Just as politicians are used to dealing with economic problems, so too they help reduce the damage that this pandemic is doing to our society by using what evidence we have.