It is the unacknowledged maths

From COVID-19: making predictions by Neo.

I have gotten to the point where I no longer put any trust in the prognostications of scientists about COVID-19 and the comparison of various social policy tactics for dealing with it. These people have lost credibility – and I was already skeptical in the first place.

Some no doubt are well-intentioned and trying their best to do a good job. Others probably have a political agenda. But at this point I think even the first group simply doesn’t know. The country-by-country data gives us some information, to be sure – enough to bring me to the conclusion that there is no simple relationship between any of the rules and the outcomes claimed to be a result of those rules. But it doesn’t tell us what to do going forward.

Sweden is encouraging in terms of relaxing the rules, but the US is not Sweden. Georgia, likewise. But it still doesn’t tell us enough.

However, I also concluded a few weeks ago that we must go forward and open up the economy or the amount of human suffering as a result of the shutdown will be absolutely enormous, too. In fact, it already has been. Those at greater risk (like me) will have to make our own decisions about how and when we’ll venture forth. But the greater good requires that we end this stalemate.

We still don't know. The data is still abysmal in its incompleteness and inconsistency. The mechanics, the measures of mortality, the variability of experience between countries and regions. Our speculations are marginally more informed than four months ago, but only marginally. The only thing we know for certain is that anyone making blanket statements with great conviction is almost certainly wrong.

Since near the beginning, I have been advocating that we pay close attention to all causes mortality. Since there is such fungibility between deaths from flu, from pneumonia, and from Covid-19, in many places, what seem like disproportionate deaths from Covid-19 are almost certainly a bureaucratic feature of defaulting to Covid-19 as cause of death. You only catch that definitional shifting by looking at all causes deaths.

One of the few things we have learned of Covid-19 is its asymmetrical mortality. To the extent that there is a pattern among dramatic plagues, the norm is that the more vulnerable, the young and the old, have the highest mortalities.

One of the distinctive features of the Spanish Flu in 1918 was that it inverted that norm - those in the prime of life also tended to have the highest mortalities. Which is what also made it so economically damaging. The mass loss of life from WWI which culled the population of healthy workers had a devastating impact on national productivity. It was then immediately amplified by the Spanish Flu which took out even more of the healthiest and most productive aged portion of the population.

What we are so far seeing virtually everywhere is that Covid-19 is asymmetrically deadly. It takes the elderly and most infirm. Young children are virtually immune and even adults don't have much exposure to mortality until you hit the fifties and sixties.

The past couple of weeks my attention has moved beyond all causes mortality measures. Still important but there is another measure which seems increasingly compelling that we acknowledge. One we are deeply reluctant to acknowledge.

The additional measure is from the epidemiological and actuarial sciences - Years of Potential Lost Life (YPLL). There are a number different ways of measuring that but the central concept is a life lost at 40 has different economic and societal consequences than a life lost at 85. 65 is often taken as the benchmark. If you die at 40, you have suffered 25 years of potential lost life.

It is a most uncomfortable measure. We are primed to consider all lives as sacred and any loss of life a tragedy. YPLL inescapably feels like a revocation of that moral view. YPLL calls us to believe, somewhat similarly to the White Queen, two or more impossible things simultaneously. Two things which cannot both be true - that each lost life is a unique and equal tragedy AND that each lost life has a different consequence on everyone else. We have to believe that the loss of an 85 year old reclusive widow in poor health is equally tragic as the loss of a healthy 40 year old father of three young children with a stay at home wife. We know morally that the two losses are equally tragic given the sanctity we attach to life and we also know that they are differently consequential. It is the very definition of cognitive dissonance.

Since the median age of death for Covid-19 appears to be in the upper seventies, the YPLL is going to be exceptionally low, implying a negligible impact on the economy.

I have seen a lot of relatively interesting discussions about the trade-offs between Covid-19 deaths and all causes deaths. If we quarantine everyone we probably will drive down Covid-19 deaths and traffic fatalities. But we probably will drive up suicides and drug overdose deaths. We are banking on the fact that the net of those calculations is positive; that the avoided deaths from Covid-19 and traffic fatalities is greater than the increased deaths from suicide and overdoses.

And that will probably be true but we don't know it empirically yet because our data is so poor and tardy.

If it is, as expected, positive, it then gives a context for estimating the cost of quarantining versus the benefit of the net saved lives.

But because we do know that the median age of lost life is so high for Covid-19, and the loss of life for those under 65 is so low, the Years of Potential Lost Life is very, very low compared to a disease with a similar mortality rate but spread across all ages.

The glaring implication is that the damage to the economy in lost productivity from Covid-19 is going to be close to zero. In contrast, we are already beginning to see the quantification of lost economic activity due to economic shutdowns. Early numbers seem to indicate an impact of 5%. I am guessing it will climb, perhaps as high as 10%.

The important distinction here is the contrast between the economic cost of Covid-19 and the economic cost of our measures to combat Covid-19. The economic costs of Covid-19 itself seem likely to be close to zero. The economic costs of of the shutdown might be 10% lost GDP.

That sounds bad but survivable. Importantly though, the pain is not spread evenly. Just as Covid-19 strikes mainly the very elderly and infirm, the economic cost of the economic shutdown is borne substantially by the 60% of the population who live paycheck-to-paycheck.

Very crudely, 60% of the population is experiencing economic immiseration, lost jobs, delayed aspirations, family break-ups, financial bankruptcy, etc. in order to add a few months of life to a small portion of the very aged. Put in those cruel terms it sounds both heartlessly callous to the elderly and heartlessly cruel to the young and hale.

The final twist is that policies are being set primarily by the 5% of the population which is healthy, wealthy, and shielded from consequences - the Mandarin Class. People who can afford to signal morality by adding a few months of life to the very elderly at the expense of the 60% of healthy working people living paycheck-to-paycheck.

The debates about what should have been the right policies and when should they have been taken will continue meaninglessly for such time as we do not have good data. As will the debate about what is the right time to open back up. It is all opinions based on estimations and guesses and mixed objectives.

It is easy to look at this and see a contest between the repressive statist mindset seeking to control a free people versus a freedom mindset seeking to liberate people. And there is no doubt an element of that in the debate.

But focusing on Years of Potential Lost Life brings some compelling clarity to the argument. The percent of the population over 70 who represent the bulk of deaths are approximately 10% of the population. Among the 85% who are below 65, 60% live paycheck-to-paycheck and whose life and prospects are being ruined. For any system which is at all democratic, the math suggests a compelling force towards opening up sooner than later.

It is not necessarily the morality, or the ideology which is driving this. It is the unacknowledged maths.