Toby Young’s attempt to measure the costs and benefits of the current coronavirus shutdown fails on a number of fronts. He is right to recognise that lost economic output has real costs, and the trade-offs of any policy have to be considered before we can judge whether that policy is worthwhile, but he makes several important errors that lead him to the wrong conclusion.

First, Young argues that we should judge that the life expectancy of people who would die of Covid-19 is one year – roughly the difference between the average age of those who’ve died so far in Italy (79.5 years old, apparently) and average life expectancy in the UK. 

But the average life expectancy of someone who has made it to the age of 80 is more than 81 – according to the UK’s National Life Tables, someone who is 80 years old has a life expectancy of about nine more years. If this is counterintuitive, remember that average life expectancy includes people who have already died before reaching the age of 80. For these purposes we need to exclude them, which leaves us with many people who will live well past the average.

He also says that the value of a life-year in the UK is £30,000. This is the figure that is used by NICE, the National Institute for Health and Care Excellence, to judge the cost-effectiveness of a new treatment. 

But this is not the same as the value of a life year to society, and the Department of Health and Social Care itself “estimates that a QALY has a monetised value of £60,000 at 2009 prices”. The Treasury’s Green Book also uses a figure of £60,000 per statistical life year as the standard way of assessing cost-effectiveness across government. And NICE will also go up to this amount for end-of-life treatments. Other estimates have put the statistical value of a life year as high as £100,000 in the UK.

Taking these factors together, Young’s initial calculation of the benefits of saving a marginal life are up to 18 times lower than they would be if he had used the Treasury and Department of Health’s statistical life year figures, and accurate life expectancy data. Even using the larger estimate of life expectancy Young considers (though he later returns to the claim that the people whose lives are being saved have “one or two years left”), the resulting per-life figure of £660,000 exceeds the £500,000 that he reckons each life saved would have to be worth to justify the current shutdown. 

The Bristol University study that Young cites to suggest that a recession would also raise mortality is flawed: it mixes up the effects of a permanent reduction in incomes with a temporary one, and there is no evidence to suggest that temporary income reductions cause higher mortality rates.

It’s also likely that Young is overstating the additional costs that the official shutdown has created compared to a “mitigation” scenario. 

For one, he looks at money spent by the government as the true “cost” of the shutdown and, even then, he uses a muddled figure. The £350 billion he uses is from the “rescue package” announcement made by the government in March, but, as he says, £330 billion of this is not actually government spending, it is loan guarantees that only cost the taxpayer if those loans are defaulted on. 

The risk of that certainly imposes some expected cost on taxpayers, depending on the expected number of defaults so Young adjusts the figure downwards. But the figure he uses assumes half of all the loans guaranteed by the government will be defaulted on. This is exceedingly unlikely: in each year during and after the Great Recession, less than 1 per cent of small businesses defaulted on their loans. (To be fair, businesses that take advantage of these loans are probably more likely than average to go bust, but this can’t be that big an effect.)

The biggest cost to taxpayers is likely to be the government’s employee retention scheme, which allows furloughed workers to claim 80 per cent of their wages up to £2,500/month from the state. The IFS estimates that the cost will be more like £10 billion over three months for each 10 per cent of the country’s workforce that takes advantage of it. In total, the IFS reckons the government will have to borrow an additional £120 billion during this period compared to what it expected.

But remember that most of this is cash payments, and so for everyone losing money there is someone gaining the same amount. Since this is being borrowed, it will eventually have to be paid back, so it is best thought of as a transfer from UK taxpayers tomorrow to UK recipients today. 

Because these are cash transfers, not a commitment of real resources (unlike much other government spending), the cost to society depends on how much deadweight loss is created by the new taxation that will eventually be needed to pay back the borrowing. These taxes will distort and deter some economic activity, but even if the cost is as high as the 30 per cent suggested by some economists, it is much lower than the headline figures would suggest.

For the most part, the bulk of the cost of the shutdown is not the amount the government is going to spend, it’s how much economic output is lost because of people staying home and businesses staying shut. But not all of this lost output is because of the government’s shutdown, it is because of the coronavirus itself. In a world where there was no shutdown, and the virus was spreading across the population, it is likely that many people would distance themselves from others anyway, avoiding restaurants, pubs and other public places, in the same way they have now been instructed to by the government. It also seems reasonable to assume that, as cases and fatalities from the disease grew, this would increase, but on a lag that meant we got many more cases overall.

How big is the loss of output from the shutdown likely to be? Morgan Stanley projects a 5.1 per cent GDP loss in 2020, followed by a 5.6 per cent rebound in 2021, compared with previous estimates by the Bank of England of annual growth of 1.4 per cent and 1.6 per cent in 2020 and 2021. Assuming that we return to trend GDP growth after that, which is the aim of the government’s current policies, that gives a total GDP loss of about 3 per cent, or about £66 billion lost output. And if we assume that the relevant counterfactual is not business as usual, but some unavoidable level of economic slowdown caused by the virus and people’s individual attempts to avoid it, then the net economic cost, temporarily ignoring the human lives saved, is even lower than this. Both are considerably lower than the £350 billion suggested by Young.

It’s also worth noting that most of the government’s fiscal response is designed precisely to avoid a long, enduring recession after the shutdown is over. The objective is to put the economy into a sort of stasis, keeping businesses afloat even if they are not bringing in any revenue, and avoiding them having to make layoffs. The more we can keep existing economic arrangements in place, the easier it will be to return to normal after the suppression measures end, and the smaller the total economic harm will be.

Taking all this together, we can attempt a rough recalculation of whether the government shutdown is worth it, from a statistical perspective. If 230,000 lives are saved (the difference between mitigation and suppression in the Imperial model Young cites), with an average age of 79.5, that is around 2,070,000 life years saved. At the statistical life year value of £60,000, that is a total benefit in statistical terms of £124.2 billion, compared to less than £66 billion lost to the economy. We should also add the deadweight losses created by the roughly £120 billion in tax we will have to raise to pay for the additional government borrowing.

But we also have to add three additional factors that add to the case for the shutdown.

One, as discussed above, the net cost of the shutdown is less than £66 billion, since a lot of economic activity would stop with or without the government’s involvement. 

Two, in a scenario where hundreds of thousands of people were dying, the NHS would become overwhelmed and the average age of the people dying would likely fall, so the number of life-years lost is probably an underestimate, and possibly an enormous one if the average age of death fell as the NHS had to prioritise younger patients. And the misery involved would be appalling. It would mean people dying in their beds alone at home, some of dehydration and starvation alongside their pneumonia, with no palliative care of any kind.

Three, there would be a significant cost to the economy of giving serious medical treatment to the millions of people who needed it, and a serious cost in terms of  human suffering as well. Remember that while hundreds of thousands of people would die, many more would contract pneumonia and other serious conditions associated with Covid-19. The cost of this suffering is probably enormous, and does not factor in any of the above calculations.

All this suggests that, so far, the shutdown is worth it, and probably would be worth it for some time more, depending on how badly hit the economy turns out to be. Three months would seem to be well worth it if it saves hundreds of thousands of lives, or millions of life-years, and it is right for the government to spend as much as it is to to keep businesses and workers afloat so we can bounce back once this period is over, to minimise the real economic costs of the shutdown. A longer period of shutdown may also be worthwhile if necessary.

Hopefully, though, it won’t come to that. With each day that passes we are getting better at fighting the virus: we are learning more about it, rolling out new ways of tracking and controlling its spread, and building more ventilators and intensive care facilities for treating people who get sick with it. The time we’re buying ourselves is incredibly valuable, and the case for giving up is weak.