The COVID-19 pandemic has thrown a particular conundrum to the forefront of general debate. If we have to balance economics and lives which matters more? Are potentially economically damaging lockdowns really worth it? Various pundits do different calculations based on the “cost” of a life, and the gulf between the Right, preferring wealth, and the Left, preferring lives, is exposed as a stark ideological battle-line.
Much of the debate is based on a false dichotomy: earlier and tighter lockdowns can minimise economic damage by being shorter, and by not needing to be repeated, and so therefore, in the longer term, save both lives and wealth. It could also be argued that a high level of infections would itself risk mass panic, and an unmanaged form of “lockdown”, with people staying away from places of work and shops, that would actually cause more damage, both to the economy and people’s general well-being, than a properly planned and managed lockdown. Indeed, in the UK the lockdown can be seen just as much as something forced on the Government by the actions of individuals and businesses as much as something imposed top-down by a Government seemingly very reluctant to impose any lockdown at all. So much of this debate is ill-conceived, but not all of it.
Ultimately, though, we cannot dodge the question. Sometimes we have to put costs on lives, otherwise there would be no limit on what we are willing to pay, or on how much inconvenience we might suffer, to implement, say, any safety measures. But the methodologies used to calculate those figures are generally highly questionable, and the answers produced are very wide ranging. A common methodology is to ask people how many years of life they would be willing to sacrifice to avoid losing certain sums of money (this is called Time Trade Off, or TTO, for income). There are a couple of things wrong with that, not least that that makes rich people’s lives more costly than poor people’s (generally) but, more importantly, that the person surveyed is being asked about sacrificing their own years of life, not someone else’s. By contrast, in reality the trade-offs often apply to different people: one receives the “benefit”, the other the reduced life (REF 1). The question that should be asked isn’t Time Trade Off as we would each apply it to ourselves, but, for example, would it be acceptable for, say, a small businessmen to avoid bankruptcy by being paid to do a “hit” on someone else?
The answer to that one ought to be fairly obvious. Most people (whether humanist or followers of virtually any of the mainstream religions) adhere to a system of ethics based on the “Golden Rule”, i.e. that we should respect the wants of others as much as we respect our own wants (at least provided they return the favour), i.e. that we should all try and at least not be wilfully selfish. How many people would want to die or lose a loved one in exchange for someone else not suffering a reduction in income? If the answer is normally very few then it is clearly immoral to impose a loss of life on others to save our own income.
A better way to arrive at such a “cost of a life” figure is to use the concept of opportunity costs: basically only to “cost” lives in terms of other lives. For example, if you have a finite health budget, then, to save the most lives, you’d want to spend the money to save the “cheapest to save” lives first, and then gradually, as it were, increase that threshold to take in ever more expensive interventions, until the budget was exhausted. If you did the opposite, and prioritised allocating spend to the expensive interventions first, you’d run out of money sooner and be able to save fewer lives overall.
That’s philosophically less problematic than the first approach, but comes with one hidden problem: how do you decide on the size of the health budget in the first place? Perhaps you have to apply the approach to the whole of the economy, and say that the whole point of the economy is to provide life, and then do that calculation again.
Another question we should address is if all years of life are equally valuable to the person actually living that life? In answer to this it is useful to introduce some form of quality adjusted year of life, such as the quality adjusted life-year (QALY) used by the National Institute of Clinical Excellence (NICE) in the UK. Does this mean, however, that we are then begging the question of how to compare money with lives, in that reduced income is likely to correlate with reduced quality of life? Not really, because here we are not talking about quality of life as defined by economic factors. In fact, provided freedom from want is avoided, quality of life is unlikely to correlate linearly with income. The more “money” we make, once our basic needs are met, the less that money contributes to our well-being, in a process of “diminishing returns”. This is why, if asked to notionally Time Trade Off life duration for income, rich people will put a higher value on their lives than poor people.
In fact, to break any even implied link to income and direct economic factors, we should not measure this quality adjustment in terms of Time Trade Off for income at all, but Time Trade Off for a direct improvement in life quality. For example, someone suffering from chronic pain might be willing to accept a shorter life expectancy in exchange for that life being pain free. Alternatively, someone who is childless might accept a reduced life span in exchange for being able to have children. In this way, it is possible to convert pure quality improvements, that don’t extend life span, also into QALYs or their equivalent.
So now we have the concept that our whole quality adjusted years of life “purchasing” budget is not just our “health” budget, or even our “road safety” budget, but the whole economy. Is this reasonable? In a way it ultimately depends on what we regard as the “meaning of life.” Nevertheless (as explained in “Questions and some answers: how to live with yourself and others”, REF 2) we seek meaning in order to find some form of fulfilment, or “quality”, in our lives. Therefore, it is indeed entirely reasonable that, in the final analysis, the purpose of the economy is indeed to provide quality years of life. Otherwise, at the very deepest level of justification, what is the economy’s point? (As explained in “What is economic justice and how can we create it?”, REF 3, the purpose of the economy is to serve people, not people the economy).
Then we have to repeat a calculation similar to the one we did above for the health budget, but now in terms of quality adjusted years of life. To be able to “buy” the most quality adjusted years of life, you’d want to buy the “cheapest” QALYs first, and then gradually, as it were, increase that threshold to take in ever more expensive “purchases”, until the purchasing power of the economy is exhausted. That allows you to “buy” the most QALYs: if you did the opposite, and prioritised allocating spend to the expensive QALYs first, you’d run out of resources sooner and be able to “purchase” fewer QALYs overall. The threshold you reach when you have run out of resources is the maximum allowable spend on a QALY, effectively the “cost of life”.
Not only that, but as “respect the wants of others as we respect our own wants” implies that we respect the wants of everyone equally (provided they return the favour at least), then at each stage of that calculation, i.e. for each resource allocation, we have to ensure that we don’t accidentally increase inequality by making whatever is the worse off section of the population even worse off. In other words: it’s a very difficult calculation to do!
It is undoubtedly difficult, and impossible to do perfectly. However, as this “costing QALYs in terms of the lost opportunity of other QALYs we could have bought instead” approach is arguably the only moral way to start even notionally putting costs on lives, or at least QALYs, even a very imperfect working out of this calculation would likely be superior to all of the other methods for putting a cost on life, especially as, in practice, those methods normally amount to putting a cost on someone else’s life. How we could go about performing this calculation is discussed in REF 4, below, where we also show that the result of this calculation could be extremely useful, for example potentially allowing us to depoliticise the setting of the size of the health budget.
For the time being though, we lack even the basic data to input into that calculation (though again, as also explained in REF 4, it would be possible to collect data that was “good enough” to make the calculation feasible).
So perhaps, right now, all we can is to ask people how much of our economic output we as a society would be willing to “spend” to “see off” different scales of threats. If we can’t actually put that question directly, then we can assess how, in actual practice, provided our society was democratic and reasonably well informed about the given threat, it has answered such questions in the past. In the case of COVID-19, if it was left unchecked, it could quite conceivably wipe out roughly 500 thousand people in the UK: for example, the Imperial College study (REF 5) estimates UK deaths from an unmitigated pandemic at 510,000. Total UK casualties in WWII by comparison were 449,700. Though that was fighting the Nazis, it isn’t unreasonable to guess that a not dissimilar number of Jews, Gypsies, the disabled, Socialists and Trade Unionists would have been murdered by the Nazis if we’d just surrendered. In fact, if we take the case of the Netherlands, and scale up to the size of the UK population at the time, it is not unreasonable to suppose that losing to the Nazis would have cost roughly 600 000 British lives (see REF 6, below). Of course, it could be countered that such a comparison does not take sufficient account of the “quality of life” costs to our way of life of Nazi occupation or overlordship, but, should naturally acquired immunity or a vaccine be unable to induce effective, reasonably long lasting, protection from COVID-19, then any society then that has allowed community transmission to be sustained is going to be facing not only indefinitely elevated mortality rates but enduring changes to its quality of life, from widespread grief and anxiety, through continuing social distancing to even the emergence of a surveillance state with protracted and enhanced emergency powers.
On the face of it, the scale of the threat from COVID-19 and the Axis powers therefore seems very roughly comparable. If so it is valid to ask how much the UK “paid”, in economic terms, to counter the Nazi threat. At REF 7 we estimate that figure to be, in modern prices, approximately 2.6 trillion pounds. Simplistically, we can divide that by 600 000, our rough estimate of casualties avoided, to produce a “cost of life” of 4.34 £m per life saved. (Some slightly less simplistic alternatives to that calculation are also explored at REF 7).
Of course, if a society was organised so that its economy already maximised quality adjusted years of life, within the constraints imposed by the need to also optimise equality, then any contraction of the economy would unavoidably reduce the quality adjusted years of life that the economy could provide. However, in practice contemporary societies are far from this point. Even the richest nations on Earth struggle to provide the basic needs of their citizens for homes, healthcare and sometimes even food, not because of any absolute shortage of resources, but because of problems with inequality. For example, the USA has no system for universal healthcare and, though the UK does, it tends to spend a lower proportion of its GDP on health than other comparable nations, and, to sustain profits, there is mass manipulation of all classes of society to engage in wasteful and environmentally damaging spending.
In such a situation there is plenty of scope for coping with economic contractions without suffering reductions in “quality adjusted years of life”, by better addressing inequalities and minimising the “invented needs” of consumer society.
It is extremely likely that, when this over, countries such as the UK or the USA will have enough food, clothing, housing and even healthcare to go around. There will be no need, even after a huge contraction, for genuine deprivation of any kind. So addressing the economic consequences of this ought to be simple: just redistribute properly. It’s only the greed of the rich and powerful (for example, in the UK the Tory donors and backers) that creates the economic problem in the first place: just as it did before COVID-19, with mass homelessness, “generation rent”, precarious employment, life expectancies flatlining and falling in poorer areas, tens of thousands of austerity-related deaths and even some starvation.
The terms of this debate are rigged by the Right (as usual). Solve the problem of greedy elites and there need be no real economic crisis, meaning we can concentrate on saving lives.
Notes and Sources
REF 1
For example, when it comes to setting limits on the maximum cost of medical interventions to be paid for out of general, and at least partially progressive taxation, the richest have an incentive to set the limit as low as possible, because they pay a disproportionate share of the partially progressive taxes and also, should they need to do so, can easily afford private healthcare, where as the general population tend to have the opposite incentive, and tend to favour a higher threshold in order to stay as healthy as possible for as long as possible. Similarly, as another case in point, when it comes to COVID-19, the richest often have the means to protect themselves more effectively than the poorest, and also have the most wealth to potentially lose as a result of economic damage. Furthermore, the disease itself does not have equal effects across different groups of people, being, for example, more dangerous to public-facing workers, the old, the poor and males. Therefore, again, the benefits and costs of measures to reduce the spread of COVID-19 will not be borne equally across a population, and, even worse, people can make reasonable inferences of where their own interest is likely to lay even while those measures are still being decided. (It could be argued that the poorest have the most to fear from economic contraction causing further deprivation, but unless the contraction is so severe it causes an absolute and critical shortage of resources, that question really reduces to one that, rather than being about the economic impact of measures to fight the pandemic, is really about a society’s willingness to redistribute its resources, in an emergency, to eliminate deprivation).
In such real-life situations TTO methodologies are completely invalid, as the person receiving the benefit is different from the person receiving the shorter life, and that difference is appreciable even before decisions are made. The sophistry of using TTO style methodologies to answer these types of society-wide conflicts of interest, especially as it is often favoured by neoclassical economists, could itself even be seen as just another weapon in the armoury of the legitimising ideology of capitalist systems based around huge inequalities of economic power.
REF 2
The deeper philosophical underpinnings of this article, and “Golden Rule” style morality, are explored here: https://gezwinstanley.wordpress.com/questions-and-some-answers-how-to-live-with-yourself-and-others/
REF 3
How to try and make the economy and markets work for people rather than vice versa is explored here: https://gezwinstanley.wordpress.com/what-is-economic-justice-and-how-can-we-create-it/
REF 4
In order to calculate the maximum amount we, as a society, should be willing to pay to “buy” a quality-adjusted year (QALY) of life, in terms of what the economy as a whole could afford until we reached the point that buying more expensive quality adjusted years of life would reduce our ability to buy cheaper quality adjusted years of life, and therefore reduce the total quality adjusted years of life we could afford overall, what do we need to know?
We need:
A knowledge of the demand for each possible QALY acquisition
A knowledge of the cost for each possible QALY acquisition, including any hidden costs such as environmental impact or any negative economic side effects.
A knowledge of the QALY value of each possible QALY acquisition.
A knowledge of the value of total economic output.
At least a rough understanding of the potentially redistributive effects of each QALY acquisition: to try and avoid any that make the already worst off section of the population even worse off.
We’ll call this a calculation of the Total Economic Equality-corrected Opportunity Cost Threshold per QALY (for lack of anything snappier).
Of course, knowing about every single potential QALY acquisition in an economy is going to be impractical, so, to make the problem tractable, we would have to sample the data and then aggregate upwards from those representative samples.
We could accomplish the above by a series of surveys sampling QALY acquisitions across the economy (purchases and direct allocations, such as free at the point of use public services, e,g medical treatment). Include things that have been refused, e.g. medical treatments refused or that couldn’t be afforded. Include only acquisitions essential for sustaining life, or for which, in exchange for which, people would actually be willing to sacrifice some duration of life. To make the process more practical, survey categories of similar QALY acquisition over a fixed period of time, rather than each actual or potential individual acquisition.
To do the calculation, once the data is gathered:
Rank each possible category of QALY acquisition according to its cost-effectiveness.
Notionally, do the cheapest QALY acquisitions first, until the surveyed potential demand is met, until the entire QALY acquiring power of the economy is exhausted.
The most expensive QALY that is still affordable is then the maximum allowable cost of a QALY without incurring an opportunity cost, i.e. without meaning that by spending too much on expensive QALYS we can afford fewer cheaper QALYS and therefore cannot maximise total QALYS.
But how should we perform the “equality correction”?
Before adding each further category of QALY acquisitions to the stack assess would those acquisitions make whatever is the most deprived section of the population even more deprived? In other words, does this step have some redistributive effects that would exacerbate inequality? (For example, perhaps the production of certain drugs causes environmental damage to poorer areas). If so, are there other QALY acquisitions that could at least compensate for this effect? (In our example, cleaning up the drug factories). If not these QALY acquisitions amount to the better off “taking” from the already worse off and therefore violate Golden Rule style obligations to respect the wants and needs of others, the “fulfilment” of others, as our own. So this step has to be skipped.
To do this calculation perfectly is impossible. But we don’t need perfection, we only need it be “good enough”. It wouldn’t have to be very accurate to morally be far superior to the arbitrary, widely divergent and philosophically dubious ways we assess the maximum acceptable cost of QALYS at the moment.
This threshold is also very useful in decision making. For example, we have seen how setting a budget for universal healthcare that is funded from progressive taxation can pitch richer against poorer, and therefore quickly becomes a matter of political conflict, with the various interest groups then wrapping themselves in ideological flags of convenience. Could we depoliticise the setting of the health budget by working backwards from the threshold, i.e. taking the maximum allowable QALY cost, and modelling buying all medical interventions, for which there is an apparent demand, that cost that as much or less, until we derive our total health budget, according to fundamental moral principles rather than class interest or feigned ideology?
REF 5
The Imperial College paper, in fact that one claimed to have finally induced the UK Government to authorise a lockdown, is here: https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf.
REF 6
The problem here is how to use historical data to try and estimate how many British lives British resistance to the Axis powers potentially saved. Of course, it could be argued that British resistance also saved non-British lives, but non-British lives and treasure were also sacrificed in that struggle (not least those from what at the time was the rest of the British Empire and Commonwealth), so here we considering only UK sacrifices compared with likely avoided UK losses.
We are going to use the example of the occupied Netherlands to provide at least a hint of what Britain avoided by successfully resisting Nazi occupation or overlordship.
Using this source:https://ww2db.com/country/Netherlands we’ll take Dutch casualties as a result of purely occupation, rather than active war-fighting, as 106 000, i.e. the estimated number of Dutch Holocaust victims.
To derive a comparable figure for the UK, we’ll scale up to the size of the UK population in 1939, using population figures for the Netherlands in 1939 and the UK in 1939 from this source:
https://en.wikipedia.org/wiki/List_of_countries_by_population_in_1939
That gives the UK population in 1939 as 47,760,000, and the Dutch population as 8,729,000, so the UK had a population 5.47 times that of the Netherlands..
We will therefore estimate the UK deaths that would have resulted from capitulating to Nazi occupation or overlordship at 579, 970. These figures are extremely rough so we will round that to 600 000.
Of course, we know that the risks of dying from COVID-19 increase significantly with age, especially from about age 50 onwards. Therefore it is even more likely that the QALYs that would have been lost as a result of Nazi occupation would have exceeded the QALYs lost as a result of an unmitigated UK COVID-19 outbreak. However, we haven’t factored in other complicating factors, such as the long term health effects on those surviving COVID-19, and, in the broad brush, the analysis still holds.
REF 7
So how much did the UK spend?
The National Debt dipped to 110 percent of GDP in 1940 before soaring to 220 percent of GDP after the close of World War II in 1945. (Sources: https://www.ukpublicspending.co.uk/debt_history and Slater, M, The National Debt, London: Hurst & Company, 2018, figure 2).
As those are debt to GDP ratios we need to know how 1945 GDP compares to 1940 GDP in order to have a actual handle on how much the debt itself grew.
Using this source: https://www.measuringworth.com/datasets/ukgdpir/ we can calculate that 1945 GDP is 1.0268 times 1940 GDP, so 1945 debt to 1940 GDP debt ratio is 1.03 * 2.20 = 2.27 or 227%.
So debt increased from 1.1 1940 GDPs to 2.27 1940 GDPs . In other words 1.17 * 1940 GDP was added to the debt.
So we could say the approximate cost of the war was 1.17 * the GDP.
Current GDP: (https://www.ons.gov.uk/economy/grossdomesticproductgdp/timeseries/ybha/pn2)
2019 2,214,888 £m in current (2019) prices
We multiply that by the 1.17 we just calculated to derive an equivalent “war cost” = 2,591,418.96 £m in 2019 prices.
Dividing that by 600 000 we end up with about 4.34 £m (2019 prices) per life saved.
We can go further here, if we want to, in order to refine the analysis, though we should never forget that it is never going to be more than extremely rough and ready.
For example, it is also arguable that Britain expended resources fighting the war to avoid economic damage, especially Nazi confiscations and extraction of “tribute”.
So we have to ask how much economic damage the Dutch suffered during their occupation. Graph 3 of this source:
https://www.researchgate.net/publication/231849982_Did_the_German_Occupation_1940-1945_Ruin_Dutch_Industry#pf2 implies that, at its peak, Nazi occupation reduced national income by 55% from 1939 levels.
Therefore so let’s assume that we saved 55% of 1940 GDP by not capitulating, and deduct that saving from our spend.
So the pure life against treasure saving cost was (1.17 – 0.55) or 0.62 times GDP or 1,373,230.56 £m in 2019 prices (deducting the cost we would have incurred anyway, even if we had capitulated).
Dividing by 600 000 gives 2.29 £m per life potentially saved in 2019 prices.
But we also paid a cost in lives, of 449,700 lives. So we could argue that the 1,373,230.56 was spent to save 600 000 – 449 700 lives instead, or 150,300 lives.
1,373,230.56/ 150,300 = 9.14 £m (in 2019 prices) spent (after deducting from total war costs what we would have lost even if had capitulated) per life potentially saved in excess of lives lost.
That deals with deaths and economic damage, and all the effects there-of. Of course, as we discuss in the main text, there would be quality of life costs beyond those arising from deaths or economic damage from both defeat by the Nazis or COVID-19, but those are extremely difficult to quantify or even compare sufficiently to judge on which side those costs would be greater, and way well be comparable within the wide margins of error we are having to deal with here.
gezwinstanley 