The government has said it will “do whatever it takes” over the coronavirus outbreak. That might have been thought to be mere rhetoric but the government’s unprecedented actions – amounting almost to a national house arrest – suggest that they actually mean it.

But is this the right approach? I don’t mean has the government reached the right conclusion, but is it even asking the right questions?

Famously in economics there are “no solutions, only trade-offs” (Thomas Sowell); everything has a cost, and the costs and benefits of an action need to be weighed. Unfortunately this is a process that governments – striving for the headlines of having solved a problem – seem to be particularly poor at, even in easy times.

What is the correct balance to strike over this latest virus outbreak? Is the government waging a successful war against a deadly enemy, or is it making far too much fuss over the pandemic?

There is no right answer, but worryingly there is no sign that the government is even asking the question; no suggestion that they are making the required trade-offs, balancing costs against benefits.

The latest announcement, that “lock-down” could go on for six months, illustrates the government’s approach of trying to solve the problem at all costs, doing whatever the public health experts say is needed to reduce the number of deaths, rather than applying a broader intelligence to balance the benefits of action against the disadvantages.

As Michael Gove might – last year – have said, this shows why we should not just give free rein for national policy to be set by experts in a particular field.  Yes, we need the input of expert advice from those how understand how contagious diseases spread (although at this point their models are, necessarily, based on speculation, for example the predictions coming out of Oxford are much less alarming than those of Imperial College London (ICL)).

The problem is not that the experts might be wrong, but that even if they are correct, they are looking only at how to solve the problem that is within their area of expertise, without balancing that against the wider problems that will be caused by following their recommendations.

Epidemiologists know how to restrict the spread of a contagious disease, but they do not know how to deliver food supplies to 66 million people when most of the population is confined to their homes. Nor can they value the lost opportunities while we are in lockdown, or the cost in jobs and education, or even the mental health effects of isolating people.

Although there has been some discussion over whether the experts might be wrong, the bigger concern is whether we should follow all of their advice even if they are right – are the disadvantages of doing so worse than the advantages? There is no sign that the government is even engaged in this vital balancing act.

On the simplest level, we can try to make a direct comparison of whether the government’s action saves more lives than it kills, for undoubtedly there will be deaths from such radical action, even if they are not visibly linked. Some elderly people may die because family and neighbours are unable to check up on them. The restrictions on shopping will (despite the best efforts to counter it) cause some people to not eat enough, not eat enough fresh food, or perhaps even die from food poisoning after keeping stocked up food for too long. Many have built up businesses only to see them collapse under the restrictions, and sadly some of them may commit suicide. There may even be murders amongst families penned in together for too long.

Looking wider, government and mental health campaigners have told us for years that “loneliness kills” – by some studies a 30% increase in the risk of premature death. Isolation has its costs. A rise in unemployment – or even a fall in GDP – leads to more deaths. Divorce lawyers are expecting an increase in business as a result of couples being cooped up together – and it is known that divorce has a huge effect on life expectancy (estimated at a 23% higher chance of early death, including a significantly increased risk of suicide).  Some of these effects may not come out for some time after the crisis is over.

At the moment, with UK deaths only in the hundreds, it is possible that the government actions could, in these ways, end up killing more than they save.

However even if, on a simple counting up of lives, the restrictions save more than they cost, it is still possible to say that the cost to society of doing so is too great.

This is not “putting the economy before lives”, because the government’s massive restrictions have a social cost as well as a purely financial one, which needs to go into the cost-benefit trade-off. Even the economic cost is largely a reflection of many personal costs; the social and emotional loss of the meals and evenings out that are now banned; the family holidays not taken; the gifts not given; the missed educational, cultural and developmental losses.

Nor is this an unusually callous calculation; such decisions are routinely made, particularly by the NHS in every day on the ground decisions by doctors about how far to take healthcare.

How should we judge this? One approach, arguably the best, is to see what risks people are willing to accept (using revealed preferences, i.e. how people actually behave rather than what they might say in answer to a survey).

On this basis, it seems the government’s “whatever it takes” attitude is unbalanced; in ordinary circumstances people are clearly willing to accept a quite high number of deaths for ordinary life to continue.

In a typical year, around 20,000 people in Britain die of an ordinary winter flu, but we do not impose quarantines or “lock down” the country to prevent that.  Even though such measures would undoubtedly save lives every winter, we do not consider it to be a proportionate response to 20,000 deaths.

Potential deaths from coronavirus would therefore have to be considerably higher than 20,000 to justify such drastic restrictions on people’s lives. This is where the difficulty of prediction complicates the issue; although some predictions of potential deaths are indeed much higher, there is no consensus amongst the experts and many people are feeling that we have been here before. Huge death tolls were predicted by the experts (in some cases the same experts) for swine fever, CJD, avian flu, SARS, foot & mouth, ebola, and others, but the forecast doom did not actually occur for any of them.

People’s recent behaviour clearly indicates that there is no broad consensus that the government’s measures are proportionate, with “hundreds of thousands” heading to the beach, and Snowdon experiencing its busiest day “in living memory”. Our actual behaviour – in economic terms our “revealed preferences” – showed that there is a widespread view that the cost of the restrictions outweighs their benefits.

Indeed if the government’s approach did have general support, there would be no need for its advice to be backed by the force of law. The fact that they have needed to order pubs, museums, gyms and other places to close, and that police are using new powers to enforce the confinement of healthy people to their homes, shows that there is significant opposition.

It is said that isolation is needed to save others, but this does not seem to be a situation where people going about their ordinary lives would put others at significant risk rather than themselves (what economists call “externalities”, where we might be willing to accept risky actions because a large part of the cost falls upon others).

Yes, younger people are highly unlikely to be badly affected by the disease (except for a relatively small number of high-risk people), while the elderly and vulnerable face a notably higher risk of death if they catch it. But this does not necessarily mean that normal social or business interaction by younger people puts the vulnerable at risk, because the vulnerable and their households can self-isolate. The externalities, the danger to others, of low-risk people leading their normal lives can therefore be removed, without having to force compulsory isolation upon those whose risk is minimal.

That is why the government’s early actions were commendable; they gave a choice, enabling those people who had concerns about catching the disease to self-isolate, and helping them to do so by protecting their jobs and income.  Ensuring that workers receive sick pay, even when they were not ill, when they chose to isolate themselves to avoid the risk of disease, and measures to prevent them being laid off if they chose to do so, were important to allow people to feel that they could self-isolate if they felt they needed to.

That was a liberal approach to the problem, enabling people to protect themselves if they felt that was the best approach, and removing them from the risks caused by other people going about their normal lives. People were able to make that trade-off, that balance between the advantages and disadvantages of isolation, for themselves. However the new approach of “lock-down” – an enforced, rather than voluntary, isolation – is an illiberal imposition that prevents people from voluntarily taking a risk that, for most people, is a small one.

As we saw the weekend before the compulsory lockdown, many people do think that the risk is one worth taking. In most cases that is a reasonable choice, and that should be their choice to make, so long as measures are in place to help those who think otherwise to isolate themselves for their own protection, which they were.

Nor is it justification for the lockdown to say that the closure of pubs, shops, gyms, museums and suchlike, or the ban on groups of people getting together, made more people willing to isolate themselves, by reducing the attraction in going out. That again is a driven, top-down approach, the government making a decision for everyone. If people’s revealed preference is that they will only give up going out if everyone else is forced to do so as well, then that preference is clearly not a very strong one (rather like the people who claim that they would be willing to pay a bit more tax, but only if other people were forced to pay a lot more).

The loss of social interaction and other pleasurable or beneficial activities is part of that balancing of advantage and disadvantage, the weighing up of the increased risk from the virus against the loss of amenities. Unless we want to live in a totalitarian society, with the decisions about our lives being made and enforced by the government, it is essential that people make those decisions for themselves.

Only individuals can assess whether they would prefer a risk of illness to the certainty of lost opportunities to do other things, and different individuals will make different decisions, based on their own circumstances and preferences.

For those individuals who decide that, on balance, they prefer to isolate themselves, there may be a role for the government to help them do so whilst still being fed and without fearing losing their homes or jobs. That is the beneficial action that the government took initially. But to go further, to bring in compulsory home isolation, for the government to take that decision away from people and families, is not the approach of a liberal democracy. The government cannot get that trade-off right for everyone, and they do not even seem to by trying.

There seems to be an alarming “dog in the manger” attitude, an insistence on an equality of misery, that because some high-risk groups should isolate themselves, everyone has to. That is a view that a few people always seem to have – since we cannot give everyone a Ferrari, no-one should have one – but it is not a view that is generally accepted in a liberal society.

Individuals’ assessment of the trade-off is therefore, in many cases, that the social, personal and economic costs of isolation outweigh the potential benefits. In a liberal society, that should be their decision to make. But even in a bureaucratic society such as Britain, that cost-benefit decision should still be made.

It is sometimes claimed that something is worth doing even if it saves just one life, but as a society we clearly do not believe that. There are around 1,750 deaths on British roads every year; banning private cars would save most of those, but hardly anyone thinks that would be an appropriate response.

So we instinctively believe that a level of deaths is an acceptable price to pay for what we see as normal life. There is a cost to save lives that we do not, as a society, think is worth paying. But what is that cost? Even for those who do not accept people’s revealed preferences, is there a way of judging how much damage it is worth causing to save a life? Can science or economics guide the government’s decisions?

Actually this is a well-established, everyday process. One method often used is the “£ per QALY” – the acceptable cost per Quality-Adjusted Life Year saved (where a year’s extra life in perfect health is worth more than a year’s extra life in chronic pain or with significantly reduced mobility). This, or something similar, is routinely used by governments around the world, not only in making healthcare decisions but also, in various forms, by other government departments, judging for example whether health and safety regulations are justified by a cost-benefit analysis.

In Britain NICE (the National Institute for Health and Care Excellence, the UK government body that determines whether drugs and medical procedures are suitable for use in the NHS) puts the range at £20,000 to £30,000 per QALY.  This means that a new drug or treatment that costs less than £20,000 per QALY saved would normally be acceptable, but one that cost above £30,000 per QALY would normally not be (as the cost rises, justification would require an “increasingly stronger case”).

That range is still the one used by NICE (they confirmed in 2019 that they are not intending to revise it in the near future, although they do use £50,000 in some end of life cases where the cost, although high per QALY, would be unlikely to continue for long). Other countries adopt similar policies, for example the USA often uses $50,000 and some American private sector organisations use $90,000 or $100,000, but although the values used differ, the approach is the same and even the values used are reasonably similar.

How does the coronavirus lockdown look on a QALY-based cost-benefit analysis? The overall cost of the government’s actions is not known, but the Treasury’s estimated direct cost of the aid package to deal with it is £350 billion. At that level, based on a mid-point of £25,000 per QALY, the government’s actions would have to save 14 million life-years to meet the standard NHS cost-benefit criteria.

Since the evidence from Italy and elsewhere is that the virus almost only kills those who already have major health problems (99.2% have other illnesses), and that most of them are elderly (average age in Italy 79.5), the average number of QALYs saved per patient will be very low (even allowing for the small number of younger, healthy people who may succumb). Saving 14 million QALYs from coronavirus would probably mean 5 to 10 million lives – and this is for Britain alone.

That level of deaths is not realistic, even with no government intervention.  The highest estimate, the shocking ICL claim of a potential death rate of 250,000 if the government did nothing, is still only a fraction of that.

Even if we accept the ICL figure, and even if the government’s actions would save all of those lives (which they won’t), and even if the restrictions won’t cause any other deaths themselves (although they will), the cost that would be considered reasonable under NICE’s guidelines would be only £7 billion.

The government is planning on spending fifty times that – and that is without counting the wider economic and social costs of the lockdown.

It is impossible to make this calculus precisely. In terms of the number of lives saved it is difficult to know even how many deaths have already occurred because of coronavirus. On the one hand there is evidence from Italy of a surge in overall death rates in the very worst affected villages, far higher than the official coronavirus death toll. But that does not seem to be repeated in other countries; overall figures for weekly deaths in Britain have not significantly increased, indeed the increased death rate could be lower than the number of deaths attributed to coronavirus.

On the cost side, it is true that significant costs would have been incurred anyway, even without the government’s compulsory shutting of businesses and national enforced isolation. However against that there will also be many other costs – for government, private sector and individuals – that are not included in the Treasury’s estimate of the direct payments it is going to make.

Acceptable levels of cost per QALY vary, we do not know how many deaths will be caused by this virus, and we certainly do not know how many would have died if the government’s response had been more restricted.

However it is still a balance that we need to make, as a society and as individuals; we cannot merely hope for the benefits whilst ignoring the very real costs just because they are difficult to quantify.

Moreover the headline direct cost of the lockdown is fifty times what would normally be considered reasonable, even based on the highest estimate of potential deaths, so the prima facie case is strong, even if we allow a huge margin for error.

The government’s actions might well be justified, however the initial evidence is that they lack widespread social acceptance and even fail to meet its own guidelines for a proportionate response.

I do not know where the balance lies, but what is worrying is that the government does not seem to be making these essential trade-offs. In its determination to be seen to solve the problem, it is failing to balance the costs – social, societal, educational, even health – against the hoped for benefits.