We’ve been trapped in our houses for six weeks now. The economy is suffering; so are we. It’s (possibly) affecting our mental health. And people aren’t visiting the NHS for routine – or not so routine – appointments. Lockdown is coming at a cost.
Yes, the coronavirus is deadly. But so is the lockdown — it will undoubtedly kill people. The restrictions are also having a very real effect on our happiness, health and wellbeing, from joblessness, loneliness and all the myriad knock-on effects. So which is worse?
About 50 years ago, Gordon Smith, then dean of the London School of Hygiene and Tropical Medicine, said: “The essential prerequisite of all good public health measures is that careful estimates should be made of the advantages and disadvantages for both the individual and the community.” 1
That is, you should be sure that you’re not doing more harm than good. I recently made a Radio 4 documentary that attempted to work out whether, with Covid-19, the cure is worse than the disease.
I’ll cut to the chase: I concluded that yes, there’s a lot of uncertainty, but even (or perhaps especially) under that uncertainty, the lockdown is probably worth it. There is one aspect of it, in particular, that was discussed in the documentary and which I wanted to flesh out a bit more: the economic impacts, and the loss of life (or rather years of life: impacts are measured in “quality-adjusted life years”, QALYs) we can expect from that, and how we compare it to the cost of the virus.
In short, we need to work out what the cost of the virus would be, if left unchecked; then we have to work out what the cost of our response to it would be; and then use those two factors to decide whether the lockdown is worth the cost. The trouble is, we don’t know either of those things
First, the impact of the lockdown. The Chancellor, Rishi Sunak, and the Office of Budget Responsibility suggest that we’re facing a really major blow to the economy. It’s really important, by the way, that we don’t get wrapped up in the idea that it is “the economy” vs “human life”. The economy consists of people’s lives, in a very direct way: if you stop people working, you make their lives worse; their businesses go under, they fall behind on rent or mortgages, they can’t afford to buy the things they want or need.
More than that, a growing economy seems to keep people alive. The correlation between GDP per capita and life expectancy isn’t exactly 1:1, but it’s pretty strong, by the standards of most social-science findings.
It’s worth noting a caveat. Jonathan Portes, a professor of economics at Kings College London and an adviser to the Brown government during the 2008 crash, pointed out in interviews for the documentary that it might not be a straightforward causal relationship – longer life expectancies could improve productivity, very plausibly; or some third factor might cause both. But he felt it very likely that much of the increase was indeed due to economic growth.
You might, therefore, conclude that a recession will kill people. Somewhat weirdly, that doesn’t seem to be the case. As Portes has written, the counterintuitive but apparently well-established finding of health economics is that in recessions, the death rate goes down and people usually live longer. There are more suicides, but those extra deaths are swamped by a decrease in accidents at work and on the roads, cardiovascular diseases, cirrhosis of the liver, and various other factors.
Long-term economic decline, though, does correlate with shortened life expectancy, according to Portes. So the impact of the lockdown on our life expectancy will depend on how long the lockdown is, and how rapid the recovery afterward. The OBR says that the dip will be significant but the recovery will be swift. Portes and Sam Bowman, an economist at the International Centre for Law and Economics (to whom I also spoke for the documentary), tend to agree – as long as proper measures are taken to deep-freeze and then reanimate the economy.
Philip Thomas, a professor of risk management at the University of Bristol who works with economic modelling and has modelled the impacts of the lockdown, was more wary. But they all agree the length of the lockdown is key, so unless we know how quickly the economy recovers, we can’t be sure how many lives will be lost to it.
But there’s another point. The economic damage caused by the lockdown — and therefore the time required to recover from it — does not scale linearly with the length of that lockdown; or, as Portes put it, “a four-month lockdown is not twice as bad as a two-month lockdown”. Instead, it’s probably many times as bad. The economist Tim Harford says that, like the virus, the economic damage is an exponential process; the OBR’s assessment relies on a lockdown of no more than three months.
All this seems to imply that we ought to come out of lockdown as soon as possible, in order to prevent the sort of long-term damage to the economy that would cost more QALYs than would the virus itself.
The trouble is that so far we haven’t yet looked at the costs of the virus itself – and it’s really, really hard to know what they are. We don’t know how many people have had it; we don’t know how easily it spreads; we don’t know how likely it is to kill any given person. (For the record: we can make increasingly good estimates as more data comes in! It’s not that we don’t know anything. But so much is still hugely uncertain.)
So comparing the QALYs lost to the lockdown (which you don’t know) to the QALYs lost to the virus (which you also don’t know) is extremely difficult.
But it’s even harder than that! The two aren’t independent. For one thing: you have to consider the counterfactual. The severity of the virus (and the public response to it) affects the impact of the lockdown. You’re not comparing our lockdown economy to business as usual – you’re comparing it to what the economy would have been with a deadly virus flying about.
You can see that to some extent in Sweden, which — as my colleague Freddie Sayers has written — is a sort of control group for Covid-19 policies. It hasn’t enacted a lockdown, but instead has allowed schools, bars and restaurants to remain open, with social distancing rules in place. But it, too, has seen a “historically deep” economic slump, if less severe than in other European countries.
This shouldn’t be surprising. If you know that going out, going to work, going to bars and museums etc significantly raises your risk of catching or spreading a dangerous disease, you’ll be less likely to do those things. The economy will take a severe hit even if we don’t lock down. (As Bowman has written elsewhere, that’s part of what Toby Young got wrong in his much-discussed article about the tradeoffs involved.)
And it’s even more complicated than that. Sure, it’s bad if we keep the lockdown on for too long – but, according to Portes, it would be even worse if we came out of lockdown too early. The early Imperial College model suggested that lockdown measures could be imposed, relaxed, reimposed as the virus came and went. But for Portes that would be worse, economically speaking, than a single, long lockdown: “That seems to me to be, from an economic perspective, quite simply mad … you cannot possibly expect the economy to return to anything like normal if people if businesses are anticipating that another lockdown might be re-imposed.”
So coming out of lockdown before we’re ready to impose other measures — contact tracing, vaccines, treatments, masks — which can keep the virus at bay, meaning that we have to go back into lockdown, would be a catastrophe.
Perhaps it will turn out that the virus is less deadly than we think, which could theoretically mean that the impact of the lockdown is worse than the disease it’s meant to combat. But right now, despite the confident claims of some sceptics, we don’t know that. So you can forgive governments and policymakers for erring on the side of caution – hoping for the best but preparing for the worst.
After all, if the disease is in fact worse than we think – and there’s no more reason to assume we’ll be wrong in one direction than the other – then not locking down could lead to vastly greater harms. It’s better to lock down when you don’t need to, than not lock down when you do need to.