Permanent 2.5 Year Drop in US Life Expectancy - It's Just Like the Flu, Brah!
There is a three in a million chance that a Boeing 737 MAX won’t arrive at its destination in one piece. At the end of the day, this isn’t that big of a deal – as late as the 1980s, this was the average for the commercial airline industry, and risks were twice as high in 1970. But people don’t tolerate such numbers such risks these days, as the value attached to human life has gone up. As a result, this model has been grounded across the world, with attendant consequences for Boeing’s bottom line.
But while it may not be that big of a deal, it is still probably not a great idea to take 500 flights in a Boeing 737 MAX within a year if one can possibly help it. Why 500 Boeing 737 MAX flights? Because even though it is a disease that overwhelmingly affects the elderly, that happens to be the equivalent risk of dying from COVID-19 for people in their 30s. Moreover, when you board a plane, you are only risking your own life. People with a cavalier “iT’S JuSt lIkE ThE FlU” are presumably more likely to spread it to elderly people, for whom a brush with COVID-19 is equivalent to a round of Russian roulette (mortality is ~1/6 for over 80 year olds). Moreover, it would even be reasonable to pay money to avoid such risks, even if it involves some inconveniences.
For this novel coronavirus threatens to fundamentally degrade the global demographics of human mortality, the effects of which may last years or decades.
This graph shows q(x), or the probability of dying at any age “x”. It is calculated by taking a hypothetical cohort, usually fixed at 100,000 at the age of 0, and dividing the number of deaths by the number of survivors by age group.
The green line represents the probability of dying in the US as of 2017.
The other lines represent the effects of various epidemic shocks: An approximate doubling in severity of the average flu season (yellow); a 10% COVID-19 infection rate (orange); and a 70% COVID-19 infection rate (red).
These figures were obtained by taking the percentage chances of dying from the flu/COVID-19 and adding them to the q(x) percentages for the US in 2017 at the mortality.org database.
The mortality stats for the flu were taken from the CDC, as reported in Business Insider. They also helpfully compare the age-specific death rates to COVID-19 mortality, as derived from an investigation earlier this month by Russell et al. based on numbers from the Diamond Princess cruise ship (where everybody was tested). The extrapolated total CFR (case fatality rate) was pegged at 1.1%, but note that this applied to situation where quality healthcare was readily available (ventilators, IV drops, antibiotics, etc.). In situations where the epidemic overwhelms the healthcare system, things are going to be much worse.
Note from the outset the near insignificance of flu as a cause of mortality; under 65’s are basically two orders of magnitude as likely to die from COVID-19 as from the flu. In other words, for the younger generations, “the flu” is just 2-3x Boeing 737 MAX flights per year, as opposed to 500x for COVID-19. While the absolute numbers for the elderly are horrific, the disparity between flu and COVID-19 mortality for them is actually considerably less – just about a single order of magnitude – though even so, that’s still the difference between a ride on the Space Shuttle (flu) versus a round of Russian roulette (COVID-19). I’d rather take the Space Shuttle, thank you.
In another study by Riou et al. 2020 analyzing data from Wuhan, a total CFR of 1.6% was estimated “among all infections”, with a larger sample allowing for more precise figures for each age group (see above). As such, I will be using the numbers from this study to adjust q(x) in the different COVID-19 scenarios. Apart from that, the Wuhan scenario is likely to be more typical than the Diamond Princess scenario, if we are talking about large-scale outbreaks that partially overwhelm the capacity of healthcare cities.
There isn’t much evidence that we can hope for substantially lower mortality rates, even in developed OECD countries; contra Western stereotypes, medical care in Wuhan seems to have been highly sophisticated, with dozens of people being ventilated in the average hospital, and complex procedures such as extracorporeal membrane oxygenation (“removing blood from a person’s body and oxygenating their red blood cells”) through ECMO machines being available in cases where ventilation didn’t work. Consequently, it can’t be excluded that mortality in most of the rest of the world – even in the OECD – may well end up higher than in China. For instance, England only has 28 and the US has 250 of these ECMO machines, whereas even provincial hospitals in China have been reported to have 5 of them each.
The healthcare system in Lombardy – one of the most developed regions in the world – is already on the cusp of collapse. Unless there are draconian quarantines implemented right about now, most of the rest of Western Europe and the US seem set to join it in its misery in another 10 days to two weeks. Cost-cutting “optimization” in healthcare has drastically reduced the number of hospital beds per capita throughout the West in the past two decades. At this point, I would certainly not wager on “the West” mounting a better or more competent response to COVID-19 than the Chinese.
Another cardinal difference between “the flu” and COVID-19 is that the latter is far more contagious. The standard measure of how many other people each person with a given disease infects in turn, r0, seems to be ~4 under “normal” conditions, versus just 1.3 for the flu. Moreover, as a novel coronavirus, people do not have any preexisting immunity to COVID-19 that might mitigate its virulence, and it has far greater contagiousness. Consequently, professional epidemiologists have predicted that as much as 70% of the world population may eventually become infected with COVID-19, a number has been repeated by Angela Merkel and the British government in recent days. As such, I will be modeling a 70% COVID-19 infection rate – which presupposes millions of deaths – as a “worst case” scenario.
One final “blackpill” about COVID-19 is that, should we fail to control it, many epidemiologists expect it to become a new seasonal disease – that is, a fifth endemic coronavirus, just like the common cold. But far deadlier. The flu infects about a tenth of the population every year. What would be the impact if COVID-19 was to reach similar intensities?
This graph shows l(x), or the number of survivors at any age “x”. It can be calculated by recursively applying the aforementioned q(x) to the initial, hypothetical cohort of 100,000 newborns.
As before, we can see that even doubling the flu season – adding mortality from an average flu season to the existing probability of dying – barely nudges the curve.
However, even a 10% COVID-19 infection rate moves the curve visibly left, and the change is extremely traumatic once you get to 70% infection rates – the sort of numbers that multiple European governments are now bandying about.
This graph shows the changes in life expectancy at different ages. It is calculated from two values derived from the above data: The total number of person-years lived by any particular cohort, or T(x), divided by the number of survivors, or l(x), in that cohort. T(x) is the sum total of person-years, or L(x), lived by any particular cohort up until all its members have died. That, in turn, is given by the following formula: L(x) = l(x+1)*d(x)*a(x), where l(x+1) refers to the quantity of that cohort’s survivors in the next year, d(x) refers to the number of deaths during that interval (or, in other words, l(x+1) – l(x) ), and a(x) is a constant that is usually equal to 0.5 (except in the very first and the very last year of life).
Here is a summary of the results:
- US life expectancy at birth was 78.86 years in 2017 (via mortality.org). There is a minor discrepancy with the official CDC figure of 78.6 years.
- Modeling a typical flu epidemic “on top” of that (so, in practice, a ~doubling of the flu season severity) would reduce US life expectancy to 78.63 years, translating to a reduction of ~0.25 years (three months).
- Modeling a 10% COVID-19 infection scenario with Riou et al. (2020) age-specific mortality rates – the sort of numbers we may expect should it become endemic – reduces US life expectancy to 76.15 years , translating to a reduction of ~2.5 years.
- Modeling a 70% COVID-19 infection scenario with Riou et al. (2020) age-specific mortality rates reduces US life expectancy to 66.79 years, translating to a reduction of a cool ~12 years.
Now this is not the end of the world, as I make sure to emphasize by including the historical mortality profiles for Russia in 1994 and Sweden in 1751 across all three of these graphs.
The year 1994 marked the single worst time for Russian mortality in its post-1956 history, when rampant alcohol abuse, violence, and the despair of the 1990s reduced life expectancy to a local minimum at 63.93 years; during that time, middle-aged male mortality was equivalent to that of Imperial Russia and Sub-Saharan Africa. This is probably the worst mortality profile ever observed in a major industrialized nation outside of wartime.
Mortality rates in the preindustrial world – Sweden has the earliest comprehensive records dating back to 1751 – jumped wildly year to year, depending on the state of the harvest and the virulence of the bugs going around in that particular year. The biggest difference relative to industrialized societies, though, even ones as collapsed as Russia in the 1990s, is that deaths during infancy and childhood were mundane, not freak occurrences. Hence why life expectancy actually goes up as children live through (survive) their infanthood.
As we can see, in terms of mortality, a serious COVID-19 epidemic should be broadly equivalent to living in 1990’s Russia – and for people under the age of 50, it would be notably safer than living in a preindustrial society, such as 18th century Sweden. It will be a shock relative to current expectations colored by more than a century of “Pinkerian” progress in safety and survivability, but there were people who lived their entire lives under similar or worse mortality profiles, and that didn’t prevent many of them from finding joy and meaning in them.
However, even though the pandemic “shock” will pass, if the epidemiologists are correct and COVID-19 becomes an endemic, seasonal disease, then we may permanently lose the equivalent of about 25 years worth of progress in raising life expectancy (American life expectancy was last below 76.15 years in 1996). In this scenario, the graph of future US life expectancy may look like something above, dipping sharply this year and stabilizing at a new, lower normal in subsequent years.
In the long-term, there may be even more years lost – perhaps 3 years – in many West European countries, and perhaps in developed East Asia as well, should this pandemic veer out of control and make it impossible for them to preserve their current achievements at checking COVID-19 (I assume that even disciplined East Asian societies cannot maintain Corona-suppressing “social distancing” behaviors indefinitely). That is because, thanks mainly to America’s opiates epidemic, the West European countries now have substantially better mortality profiles than the US, so the extra “shock” of COVID-19 will depress their life expectancy to a relatively greater extent. Though, curiously enough, most of these same countries will “lose” fewer years of progress relative to the US, since American life expectancy has basically stood still for the past decade due to the opioids epidemic.
Meanwhile, industrialized countries with worse mortality profiles, such as Russia, will not actually see as big of a drop in life expectancy as the US; as of 2014, the last year for which I can find life tables for Russia, a 10% COVID-19 infection scenario translates to a 1.7 year fall in Russian life expectancy (US: 2.5 years), and a 70% infection scenario translates into a drop of 8 years in life expectancy (US: 12 years). However, due to strong gains in Russian life expectancy since 2014 – it has risen from 70.9 years in 2014 to 73.4 years in 2019 – the effects of COVID-19 will actually now be stronger (if still not as strong as in the US).
(Reminder: This is all assuming that both infection rates and the age-specific mortality rates from COVID-19 are the same across these countries – this will almost certainly not be the case due to local specifics).
Moreover, there will be multiple other factors that will either ameliorate or depress the above estimates:
- COVID-19 is going to kill off the frailest people in this current wave, in which up to 70% of people may be infected; but this will soften its long-term impact, since you can only die once.
- In subsequent years, when ~10% annual infection rates may become the new norm, healthcare systems will adjust and everybody should receive adequate care, lowering CFR from the ~4% currently observed when healthcare systems are overwhelmed, to the 0.5%-1.0% rates seen in South Korea and Chinese provinces outside Hubei, which have managed to keep on top of cases.
- Conversely, people who are intubated now may suffer permanent, long-term insults on their health, making them more vulnerable to subsequent COVID-19 infections in future years.
- Needless to say, there may well be changes in COVID-19’s contagiousness and virulence in the future.
I am not even going to attempt to model any of this. But the bottom line stands. This virus has the capability to deal a traumatic shock to the world’s population, especially to the older societies of the Global North. In the longer term, it may also permanently depress global life expectancy by about 2 years, robbing millions of future people of their planned retirements and time with their grandchildren.
There are political factions that cynically, and unironically, pray for Corona-chan to do her magic. The Chapo Trap House folks bask in the idea of COVID-19 killing off Drumpf-voting boomers who are keeping them from electing Bernie, instituting M4A, and rescuing the planet, while elements of the Alt Right anticipate the West rediscovering its youthful vigor in the wake of the “boomerpox”. But I would caution both factions against premature Schadenfreude – political trends rarely work out the way anyone expects them to. They may get more than they bargained for.
OK, summing up: We should really, really try to avoid COVID-19 veering out of control and becoming endemic.
China has demonstrated that Corona-chan can be contained; its r0 has plummeted by an order of magnitude from 4 to just 0.32, even if had to bring its economy to a near standstill to do it. As Steve Sailer notes, China hasn’t merely “flattened the curve”. It has crushed it. This means that its success should be replicable, at least in First World nations with epidemics on the scale of peak Hubei, as in Italy.
Even more encouragingly, the nations of East Asia – Japan, Taiwan, Singapore, Hong Kong, even middle-income Thailand – have all managed to bring COVID-19 under control at its earliest stages without resorting to China’s drastic measures. As Tomas Pueyo explains, they did this by carefully filtering infectees’ contacts at the earliest opportunity and putting them under quarantine. The main reason that South Korea failed is because its “Patient 31” happened to be a religious “super-spreader”, yet even so, even there, the epidemic is currently under control.
But all their efforts would be in vain if just a few (or even one) defeatist, incompetent, or plain stupidly-run countries decline to take the necessary steps, and thereby cut two years off global life expectancy into the indefinite future.
This stupidity and incompetence takes different forms. In Western Europe, it is the Left’s fundamentalist commitment to open borders, accompanied by bizarre claims that quarantines do not work. In the US, it is the Right’s fundamentalist commitment to free markets, as exemplified by $5,000 copays for coronavirus tests, lack of sick leave, and Trump’s “iT’S JuSt lIkE ThE FlU” mantras to appease Mammon. Meanwhile, in what is perhaps the most “powerful” move of them all, the United Kingdom has set up a cyber-unit to combat “Russian” Corona-chan shitposters while basically admitting that it has no interest in combating, like, the actual coronavirus. Hopefully the British boomers croaking in their deathbeds in another two months’ time will be understanding of HMG’s priorities.
One is almost tempted to wish a pox on all their houses.
Anyhow, while I still hope for the best, I do not expect it.