COVID vs Economy Part I
Myths & Analytical Shortcomings
Aug 31st, 2020
Foremost, a disclosure of my experience and the inherited bias when discussing epidemiological topics: though neither an MD nor an MPH, my Princeton thesis studied the AIDS pandemic in Tanzania. A 137-page analysis, it chronicled the disparity between normative treatment policies and the actual patient-level results in various hospitals. It involved two summers of field research in Dar Es Salaam, where I witnessed the intimate and widespread destruction of human life by multiple diseases. With levels of poverty, comorbidity, and medical resource shortages unseen by the developed world in generations, my exposure to pandemics has been colored by the most unforgiving realities of biological ruthlessness. It is a perspective forged in the crucible of humanity’s most dire struggles against its deadliest diseases; an attitude stripped of its privileged first-world veneer; and an epistemic position derived from witnessing a wider range of human suffering than most Americans will ever know is possible.
Therefore, my outlook regarding the current SARS-CoV-2 (aka: novel coronavirus/COVID-19) is one marked by increased sensitivity to the power of pandemics. And yet perhaps a simultaneously more sensitive and more cosmopolitan perspective is exactly what is needed during this time of crisis. Maybe the nuance provided by a less provincial mindset is precisely what can shed new light on America’s strained and tragic response to the disease.
To start with the most sobering fact: America has a mere 4% of the world’s population, but boasts more than 20% of its COVID deaths. Alarmingly, the American death count exceeds the next nearest country by an astounding 53%. Even more tragic is the fact that this has all occurred in spite of America’s vaunted status as the country with the world’s largest GDP, the most medical device patents, the largest pharmaceutical market, and indeed the highest expenditure on medical care of any country in the world. Something fundamentally doesn’t add up.
Explanatory models for this colossal mismatch between medical potential (i.e. - the ability to stop a pandemic) and medical outcome (i.e. – the grim reality of leading the world in COVID deaths) are as numerous as they are tragic. They include: widespread policy failures, inadequate medical insurance/medical access, systemic racial disparities, an increasingly obese population, among others. It is a multifaceted miscarriage of humanity.
Popular COVID Myths
As plenty has already been written on the aforementioned explanatory models, this analysis focuses on one phenomenon: the myths perpetuated by a society. Myths exist as a subset of culture, and where culture is defined as the shared set of values among a given population. Myths are perpetuated by their repetition, particularly by those in authoritative positions and on high-transmissibility platforms like social media. And unfortunately, the myths surrounding COVID have had a devastating effect on both the physical, psychological, and economic health of America.
Some of the myths are easier to spot than others, and with common sense and basic scientific knowledge one may overcome these erroneous – and dangerous – conclusions. Much as an educated sub-Saharan African should realize that supernatural witchcraft cannot prevent an HIV infection, an equally educated American should realize that ingesting bleach cannot prevent a COVID infection. And yet nearly 50 Americans in North Texas imbibed the disinfectant in August alone.
Other myths are harder to pinpoint, and may seem all the more attractive given the absurdity of competing myths. Even more troubling is that these myths are often perpetuated by smart, well educated, and/or influential people. What follows is an examination of some of these common American myths, and a combination of simple logic and empiricism to expose them for their faults.
1. The myth that “More testing means more COVID cases!” This is a particularly pernicious one that has been perpetuated by national leaders, media outlets, and otherwise intelligent individuals. However, this myth employs particularly faulty logic; specifically, the post hoc ergo proctor hoc fallacy. In short, just because one event follows another does not mean that one caused the other. A simple empirical example to disprove this myth is to examine another pandemic which was successfully controlled, such as Ebola or SARS-CoV-1. Once appropriate mitigation and suppression efforts were employed, and testing of nearby populations took place, the number of infected people never rose. Testing was the indicator to see whether or not the disease spread, not the cause of any shift in numbers. Likewise, the current number of COVID infections hasn’t risen because of the increased number of tests, but because the virus itself has spread. Perhaps the best example to highlight the absurdity of this erroneous thought process is to point out, facetiously, “If only Herman Cain never got tested for COVID, he would still be alive today!” When one cannot dissociate simple temporal sequence from actual causality, all forms of illogical myth can fester.
2. The myth of “inevitability.” A close friend and award-winning philosophy major at Princeton tried convincing me that the mass COVID death in New York was “inevitable.” As he was in the Big Apple when the outbreak unfolded, he contended that the size and density of the population made a severe outbreak unavoidable. But then what of Seoul, where I was in late January & early February at the outset of the Korean outbreak? Seoul has both a larger population than NYC (10.1mm VS 8.4mm) and a 57% higher population density (43.3k per square mile VS 27.6k per square mile), and yet Seoul has a grand total of 20 (!!) COVID deaths VS the more than 23,000 in NYC proper, as of this writing. (For the mathematically disinclined, that’s more than a thousand-fold difference.) But perhaps Seoul is a statistical outlier; an aberration in the dataset. If that’s the case, then what of Tokyo? Of Singapore? Of Shanghai? Of Lagos, or Frankfurt, or Geneva? Of Hong Kong, or Nairobi, or Taipei? If anything, New York seems to be more of the outlier than the ‘inevitable’ trend. There are many ways to describe the tragedy of mass death in New York, but “inevitable” is certainly not one of them.
3. The myth of moderation. The same friend also tried arguing that America “needs a moderate response” to the pandemic, and that there were too many extreme reactions, such as attending bars and parties on one extreme and strict social isolation on the other. He tried arguing that there’s a normal distribution to the behavioral responses to the pandemic, and the appropriate reaction was to act within a certain “standard deviation” from the norm. (His argument was a rehashing of what some call the “enlightened centrism” position.) And while appropriate in some instances, “enlightened centrism” has proven utterly inadequate and even dangerous in the American cultural response to the pandemic. The crux of the argument is also easily dismissed, both at the theoretical level and at the applied level. At the theoretical level, the notion of an appropriate response being based on a supposed ‘normal’ standard is negated by rhetorically asking, “what is normal?” Specific to America, “normal” and the resulting deviation in “appropriate” behavioral response is thoroughly negated by America’s shifting Overton Window. That is to say that “normal” has been moving evermore towards the extreme, and therefore an arbitrary center point cannot serve as an anchor for appropriate behavior. Likewise, a simple comparison to another culture’s “normal” exposes America’s “normal” for its absurdity. For example, to use Korea again, their national debate was not whether to wear masks or not (it was universally accepted that everyone should wear masks, and I was given as many as I wanted for free on the day the first Korean COVID case was discovered), but whether surgical masks or KN95 masks should be worn. In other words, one society’s “enlightened centrism” is another society’s radical extremism, which negates the notion of moderation all together. And at the applied level, the characterization of a behavioral response that fits within a ‘standard deviation’ of a ‘normal’ also assumes that the distribution itself is normal. But as any data analyst knows, normal distributions rarely appear so conveniently. America’s stark cultural divide – seen in everything from mask wearing to national legislative policy to the percentage of Americans who think the COVID death rate is “acceptable” – is all indicative of a bifurcated distribution. Mathematically speaking, deviation from any position along a bifurcated distribution will inevitably put someone in an extreme position compared to a large portion of the population. In quick summary, amid a polarized domestic culture war and a highly variable international pandemic response, reacting “moderately” is both a theoretical and practical impossibility. It’s another myth among many.
(Addendum for the philosophically inclined: yes, this is a critique of moral relativism at an applied, epidemiological level.)
4. The myth that “For every 1% rise in unemployment, 40,000 Americans die.” This has been debunked by Economics Explained. But the short version is that this was based on a 1980’s economics paper studying the effects of outsourcing and declining American manufacturing. The association between the laid off workers and deaths was primarily attributed to alcoholism and drug overdose. As many have surely heard before, “correlation isn’t causation.” That rings true here. Moreover, the correlative relationship is likely bounded, which is to say it does not extend linearly outside of a rather narrow range. If it weren't, then America’s jump from 3.5% unemployment in December 2019 to 14.7% in April 2020 would have resulted in nearly 450,000 dead Americans from drug and alcohol overdoses, which clearly didn’t happen.
Other myths persist, including widespread misunderstanding of comorbidity, unreliable certitude in what is still an evolving understanding of this pandemic, and others. But what is abundantly clear is that plenty of Americans have died during this pandemic. More than 180,000 deaths so far, with associated “excess deaths” having exceeded 204,000 back in mid-August. And yet a growing chorus of Americans continues to clamor for a rapid return to pre-pandemic activity. How can one explain this?
Death Addiction & Risk Perception
To opine for a moment on American culture as it relates to the fact that the country leads the world in COVID deaths yet repeatedly falters at employing policies which could prevent those deaths, I’d suggest a study of the idiosyncratic American ethos. Pick your euphemism: “American exceptionalism,” “rugged individualism,” or “post-modern socio-economic Darwinism.” As COVID has made painfully apparent, all are thin veils for what can only be described as a blithe insouciance to mass preventable death.
(Addendum for those who did not click the earlier link for the statistic: A majority, 57%, of those who self-identify with the dominant political party believe the current COVID death count in America is “acceptable.”)
Naturally, many Americans will disagree with that characterization. And indeed, many Americans harbor vast amounts of empathy, care, and self-sacrifice. The 900+ American healthcare workers who have died to COVID are perhaps the best example. (And for comparison, that trounces the combined 383 battle and ‘in theater’ deaths of America’s military in 1990-1991’s Operation Desert Shield & Desert Storm.)
No, the characterization of a nation which cares so little about the life and livelihood of its citizens is one that applies less at the individual level, and more at the collective, national level; at the legislative level, as codified in the official policies which govern the country, and at the executive level, which enforces those policies. Tragically, this blithe insouciance has become a part of the country’s deeply woven cultural tapestry; an indelible part of the 21st Century American ethos. From an obsessive reluctance to make healthcare an affordable right (the way every other OECD country has), to the all-too-common occurrence of gun deaths (including more than 340 students and teachers who were shot on school property in the last decade alone, not to mention the disproportionate number of black Americans being shot or extrajudicially killed by law enforcement), to the national obsession with war (since 1776, the USA has been in a declared state of war >93% of the time), to the current COVID debacle we’re all suffering through (8 months of American COVID deaths already accounts for >29% of the total U.S. military deaths in the entire 20th century) – the national attitude to mass American death seems to be, “it is what it is.” Insouciance indeed.
Empirically and tragically, few countries are as comfortable with the mass death of its own citizens as the United States of America, particularly among the wealthy OECD countries. But there’s another layer beyond the idiosyncratic addiction to mass preventable death that explains America’s COVID response, and one which harks closer to the broader discussion of human perception of risk.
In what has been dubbed “The Five Rules of Risk,” which describes the flawed risk evaluation system endemic to most human psychology, the widespread American acceptance of the COVID death count seems to fit squarely into three of the rules: Rule #2 – acceptance of risk is inversely proportional to prevalence, Rule #3 – disease is a yardstick, and Rule #5 – numbers are numbing. Without going into detail, suffice to say that COVID’s intrinsic nature facilitates its undervaluation or even outright dismissal as a serious risk. Specifically, COVID is widespread (the pandemic is global, an example of rule 2), it is a highly communicable virus (a disease; rule 3), and it has a high death count (840,000+ deaths and counting; rule 5), all of which make it a seductive psychological trap for people trying to dismiss both the risks and associated casualties.
The Economic Tradeoff: Another Myth?
Again, foremost, a disclaimer: I’m neither an economist, nor a mathematician. I merely have an affinity for social sciences and data analysis, and believe I can provide cross disciplinary insight into current phenomena.
And an exceedingly interesting phenomenon is the purported tradeoff between preventing more COVID casualties and preventing the continued destruction of the economy. It has captured the attention of famed economist & Nobel laureate Paul Krugman (“On the Economics of Not Dying: What Good is G.D.P. if You’re Dead?”) and Princeton economist Marc Fleurbaey (“Should We Save Lives – or Save the Economy?”), among others.
Yet my evolving curiosity begs the question, is it even a tradeoff? And more broadly, is the tradeoff narrative real, or just another myth? As constructed by Prof. Krugman and Prof. Fleurbaey, the tradeoff is real; it’s a compromise between lives lost and economic damage in a two-variable direct relationship. In other words, when lives are lost the economy gains strength, and when lives are saved the economy suffers. As such, there should be a theoretical ideal between biological cost and economic cost where maximum social benefit can be attained.
I’m not convinced. There’s a strong argument with growing evidence that the economy cannot fully resume activity until after the virus has been adequately contained. Therefore, as a theoretical conceptualization of the problem, the issue is not a two-variable direct tradeoff (lives vs economy), but a three-variable optimization problem (lives vs economy vs time, where time is the immutable but arguably most important variable). More on that in Part II though.
For now, it’s worth studying the two-variable tradeoff between lives and the economy, as it is a commonly accepted analysis, despite its flaws. To analyze the tradeoff, one needs a common metric to compare. And as economists are prone to do, their chosen metric is money. As such, it becomes necessary to put a dollar value on human life. (There’s a reason that economics is known as “the dismal science” which “knows the price of everything but the value of nothing.”) Philosophical problems aside, what do the data say?
As is commonly accepted, the Statistical Value of Life (SVL) is the benchmark. For Americans, the number stands at about $10 million per human life. That means the current death count equates to a little over $1.8 trillion dollars. By contrast, the US GDP contraction during the first half of the year totaled just over $2.3 trillion, which makes for a surprisingly close comparison. (Whether or not SVL & GDP are appropriate metrics is an entirely different discussion, and ultimately, I contend that they are not.) If one is seeking parity between the two figures, America has apparently stumbled on an equilibrium that is not only within the same order of magnitude, but relatively close within that order. A grievous coincidence perhaps. Deft policy implementation likely could have kept both numbers much lower, but that too is a different discussion, and one that would place the ‘tradeoff’ debate more squarely in the ‘myth’ realm, as both would then become independently associated variables.
Nonetheless, to conclude the thought experiment, it is worth noting that if there had been no shutdown and the virus had been allowed to spread unimpeded, the death toll would have likely been between 2.2 million Americans (the “Imperial College” model) and 3.3 million (assuming a ~1% mortality and all Americans were infected), which equates to $22-33 trillion dollars in lost SVL, and which exceeds the entire value of the US economy (~$21.4 trillion in 2019).
Likewise though, it is also worth noting that due to the interconnectedness of a globalized economy, even if the US had managed to keep its economy 100% open, the shutdowns in the rest of the world would have negatively impacted GDP in a significant way, and maintaining 2019’s GDP levels would have been impossible.
Clearly the methods popularized by economists are crude. From the issue of putting a dollar value on human life to the notion that GDP actually reflects the health of the people in the society, these methods are overly ripe for widespread and well-deserved criticism. Yes, the tradeoff may roughly estimate the economic impact of the pandemic and guide policy options, but it should be abundantly clear that they suffer from gross analytical shortcomings, and fail to capture the human impact of the pandemic.
In Part II, I will offer a potential workaround to the problems inherent to the economic methods, using a mathematical formula I’m still developing.
R Code For Above Calculations (Not traditional coding; moreso using R as a glorified calculator to compute numbers and retain data for easy future calculations.)
Corrected Sept. 3rd to "Imperial College," with addendums. And I have also added a "Part 1.5" as a response to this piece here.