The COVID-19 Pandemic and Human Bias


This is my first work on a topic other than professional basketball. If you have good or bad feedback after reading, let me know on Twitter.

Disclaimer: I am not a trained epidemiologist, nor do I hold any formal social science qualifications. My interest in writing about this topic came from a years-long fascination with research on behavioral psychology, and human misconceptions of reality formed from our innate perceptions. If you are interested in learning more about behavioral psychology, Michael Lewis’ The Undoing Project is a comprehensive read on the work of Daniel Kahneman and Amos Tversky, the forefathers of modern behavioral psychology. For my writing, I also referenced Cassie Kozyrkov’s post on towardsdatascience.com and Thomas Davenport’s essay in MIT Sloan’s Management Review.

Additional disclaimer: though I don’t hold an understanding of or write in detail about the virus, its symptoms, or the potential effects of contraction if you suffer from nosophobia it may be prudent to stop reading here.


EE1: The COVID-19 Backstory

The coronavirus pandemic has been a time of reflection and understanding. Trapped in isolation, as the majority of us have been at some point during a prolonged stretch in the first half of 2020, we’ve been reckoned with reflecting on the dichotomy of pre-COVID (i.e. 2019 and before) and what we can call “the COVID era”.

While different details may emerge that dispute what we know today (spoiler alert: a large theme of this essay), the prevailing conjecture is that COVID-19, or Coronavirus Disease 2019, started in central China – more specifically, Wuhan in the Hubei province. Like other pandemics before it, epidemiological experts believe the virus originated in a wet food market – a traditional Chinese custom where raw or live animals can be bartered over and exchanged.

As of this writing, there are 14 million confirmed cases worldwide, with an estimated 8 million recovered. 600,000 deaths are among these twelve million cases, equatable to 4.2 deaths per 100 confirmed cases. Because COVID-19 is now central to our vernacular, seemingly analogous terms are misappropritaed and used incongrously. So for completeness, a “confirmed case” is a person who has had a positive test result for COVID-19. A “recovery”, according to the Center for Disease Control’s guidance, is anyone who had a positive COVID-19 test, and at least 3 days (72 hours) has passed with the resolution of fever without the use of fever-reducing medications, and an overall improvement in respiratory symptoms (e.g., cough, shortness of breath); and at least 10 days have passed since symptoms first appeared.

In case you are under the impression this would be a simple topic, it’s not. Which is a fine segway into our next topic.

EE2: The God Complex

Have you ever met someone with the self-impression that they are infallible, even when drawing conclusions on subjects which they had little or absolutely no priors? The Scottish doctor Archie Cochrane noted this in his study of physicians providing care across the United Kingdom in the 20th century, and championed the need for randomized control trials in determining the efficacy of health care objectives. His work was, to put it mildly, revolutionary. Insights from his studies not only benefited the health care field but also ushered in the use of randomized control trials in a variety of other domains. (Cochrane’s book, Effectiveness and Efficiency: Random Reflections on Health Services, details his work.)

But in lay terms, what Archie Cochran’s work showed was the manifestation of the God Complex, or narcissistic personality disorder, among the physicians in London’s hospitals. English economist Tim Harford eloquently summarized Cochrane’s discovery in his 2011 TED Talk.

Video Credit: TED Conferences LLC

Thinking about the COVID-19 pandemic holistically, we could ascribe some portion of the preventable suffering (most notably loss of life and economic hardship) to the God Complex. Heads of state in varying levels of leadership around the world have admitted they could have prepared better by dedicating national resources towards a defense fund, or by taking more austere action to slow the spread of the virus as it was beginning to ravage central China in January.

In the United States, for example, a proportion of the spread of early infections on the West Coast and in New York City could have been prevented by requiring masks in public, without instituting full quarantine measures that contributed to the economic halt in March and April. (A study published by a syndicate of epidemiological professionals from the World Health Organization, The University of Hong Kong, and the Harvard School of Public Health – to name a few – found that “surgical face masks could prevent transmission of human coronaviruses and influenza viruses from symptomatic individuals”.) Again in the U.S., reluctance on the part of elected officials at all levels, from city to federal, to make adjustments in the face of increasing infections and hospitalizations shows the inertia that can come from a small dose of the God Complex.

The human inherent God Complex is not so easily distinguishable since it is, unfortunately, rather ubiquitous in twenty-first-century life. Instead, it is easily recognizable in its absence, by the refreshing skepticism, prudence, and shrewdness that takes its place. When championed, it can substitute for neuroticism in providing early warning signs that can bring forth action or thought where there is none.

Sometimes, humility is the most effective tool in toolbox.

EE3: Interpreting the Data and Confirmation Bias

We as humans enjoy being right. Whether we admit it or not, our brains actually prefer to be right more than we prefer to solve a problem or create a positive outcome. In fact, we are so good at “being right” that even when we are given evidence to support the contrary, we build up our defense to discredit the opposition. We hold on to suppositions and half-baked ideas as if they are a representation of us, as if they are our progeny.

Personally, I have found it fascinating to watch different individuals draw disparate conclusions from the same data. Even though this has been happening for millennia, and will continue to happen following this particular pandemic, it is both more salient and automatic for the COVID-19 conversation because we are required participants. There is no unsubscribing from a pandemic.

For example, with the resurgence of cases (or “second wave” as some have termed it) in the United States, increasing confirmed cases have been met by both skepticism, zeal, and paranoia all at the same time. On July 4th, America’s celebration of independence, ironically, 50,445 new positive cases were confirmed nationwide.

One might scoff at this count if they are skeptical of the fatality of COVID-19 in general, and believe that long-term economic impacts from another “shutdown” would cause just as much, if not more agony and suffering. A common rebuttal would be using the comparison between both the transmissibility and fatality of the annual influenza strain to COVID-19, postulating that we shouldn’t be worried because we don’t change our lifestyle for the flu each year.

Another individual might read the same figure and find the impending danger incomprehensible. 50,000 new daily cases with an assumed 4.2% rate of fatality (EE1) could result in 2,100 deaths a day. That number can seem staggering, given that many of us have invested 4 months of precious time capital (the only form of capital that is nonrenewable) in an attempt to modulate the virus’ spread and harm.

So how do we go about this? The best cure for confirmation bias is intentional exposure to dissenters, or those who disagree with us on a topic. And we should, as Daniel Kahneman calls it, “put our intuition on ice“. Meaning, take the time to truly construct an idea based on the conjectures of your opponent. See his or her perspective for what it is, without our priors clouding our view. Even if we do not waver in our position, we will undoubtedly learn something we did not know previously.

Regardless of our intentions, we want to see the world the way we have always seen it. It is comforting to hear that we are right and that others are wrong. Once we have made our statement (e.g. “masks are for sheep”) it is, as we see it, a part of our identity. And at this point, our personal criterion has shifted from finding a solution to the problem to ensuring that “my team” wins the fight, which can be highly toxic when attempting to solve a complex problem with requisite collaboration and human life at stake.

EE4: Hanlon’s Razor and Perceived Wrong-Doing

In a circumstance as new and unpredictable as the current pandemic climate is, we are inevitably going to make mistakes. These mistakes can be obvious or inconspicuous to all of us even months, years, or decades later. Two people can have different opinions on the necessary action, or inaction, and only later will we learn the truly correct course. There is no immediate feedback in a wicked environment like a pandemic.

The most difficult comprehension we reckon with is suffering for the blunders of others, like a requisite social game that we must play daily. If a small minority of a community practices the guidelines outlined by the institutions instructed to provide guidance on matters of public health, but the rest of the community lives life as normal, the actions of the minority will have little effect. In a way, it’s a case stuy in the tragedy of the commons (if you believe that containment of the virus is beneficial to the general public). Our own self-interest is to live our lives as we please – buy the groceries we like, go out for drinks with friends, take our kids to the park, go shopping, etc. In this social game, we all must pay the tax for those that follow their self-interest. And to make the situation even more contentious, we don’t know the materiality of that tax.

Most individuals react to these kinds of perceived transgressions with anger or spite, wishing that they could opt-out and reprimand the neer-do-wells. Regardless of your perspective on who is right and who is wrong in any given position, it is important to remember the Hanlon’s Razor aphorism: “don’t attribute to malice what can be explained by ignorance”. In other words, don’t assume malice on the part of another.

Like other human biases, what is referred to as hostile attribution bias, or our tendency to interpret others’ actions as hostile, is an innate and visceral reaction rather than a pragmatic one. Most of us, I would posit, don’t wake up in the morning hoping to find someone to be angry with. But we make an unhappy encounter with someone acting in a slightly harmful, or even in a slightly careless manner with the potential for harm, and we naturally turn into the anger emotion manifestation from Disney’s Inside Out.

In my experience, the key to practicing Hanlon’s Razor is two-fold. First, recognize that humans are flawed and appreciate that you are fallible as well. And second, be willing to give grace to those you encounter. People make mistakes often, but that does not mean that we are out trying to cause as much harm as possible.

Ignorance, especially with regards to manners of deep personal unrest like a pandemic, is almost guaranteed to breed in such a situation.

EE5: Probabilistic Thinking

It has been denoted through a myriad of social experiments and empirical trails that human beings are not equipped to understand probabilities. A lot of us, understandably, prefer to see absolutes rather than the nuance in life. It is much less taxing and discomforting to do so. In certain environments, like team sports or algebraic mathematics problems, there are only absolutes. Did you win the game or not? Did you know the answer to 2 x 5 or not? There is nothing to interpret other than the final result.

However, in most arenas the environment and our understanding are not suited to absolutes. To be truly effective we have to learn the degrees of certainty and uncertainty and be able to communicate them to one another. A lot of humans, myself included, will see a forecast with a 90% chance of sunshine and conflate this to be a 100% chance of sun. Then we curse the local weatherman when it starts to rain in the afternoon. But 90% does not equal 100% (obviously), and in fact, there is still a significant opportunity for a 90% forecast to be wrong. Rather, it is optimal to think of these forecasts as prognostications of the likelihood of an event, played out over 100 hypothetical instances. In the weather forecast example, 90 of the 100 possible scenarios result in sunshine, but 10 of those also result in rain. In these situations, we cannot attribute the result to the accuracy of the forecast.

It’s even more difficult for us to accept this reality in environments when no one is positively, certainly right, even (and especially) the pundits. Setting aside the uncertainties of COVID-19 going forward, we still lack accurate data on our predicament at the present moment. Yes, we have a rough estimate of the rate of infection and recovery or death, but we are far from seeing the complete picture. There remain a plethora of open questions. What is the real rate of infection and death nationally and globally? Where is the virus adapting and spreading most effectively (harmfully)? What factors can cause asymptomatic cases versus those with diagnosable symptoms?

Going back to the human tendency to want to be right and liked (EE3), most of us don’t like confronting the realization that we are not and cannot be omnipotent. Even if we consciously know that, of course, we are not going to be right all the time, our subconscious begs us “yeah, but why can’t we just be right one more time?” each time we’re confronted with this realization. And then “this time” repeats itself every time we find ourselves ill-equipped or short-handed in solving one of life’s mysteries.

With COVID-19, there are some who would incorrectly interpret the 4.2% fatality rate (EE1), or any of the other marginal statistic, such as the below chart from the Center for Disease Control to be equatable to zero (at least for the age groups below forty-five).

Image Credit: Center for Disease Control

This is the closest estimation, but it won’t be as effective as using probabilistic thinking to make a determination. If you are in the “25-34 years” age group, it’s true that you are not likely to die upon infection, but that does not mean you will not die. Again, it sounds obvious but for the number to be above zero, at least one of those such cases has to have resulted in death.

A better course is to think about the prevalence of the occurrence of the unlikely event as a subset of a sample of occurrences. If you were in this same age bracket, a .6% fatality rate is roughly equatable to one death in 167 confirmed cases. When analyzing the problem in this manner, the severity of the situation is better illuminated.

Which leads well into our final bias for discussion.

EE6: Availability Bias

Our inability to understand the wildly complex world in which we live requires that we use our best attempt to connect the dots. But simply having an analogous correlation or a ready example with common characteristics does not provide us with an immediate solution. When we have limited knowledge of a particular domain, as is the case in epidemiology for the majority of us, we try to substitute anecdotes for knowledge and soldier on bravely.

I would argue this is more cowardly than brave. Sometimes soldiering on with limited information can lead to disaster. Take The Map is Not the Territory fallacy, in which we take the representation of fact as truth, rather than seeing it for what it is. Such as the account, “a 6-foot tall man crosses a river with an average depth of 5 feet 6 inches and drowns”. Why? Because the majority of the river is 5 feet deep, with a sharp drop of 11 foot depth at the river’s central point. The law of averages doesn’t serve as a proxy for all data on the river.

We have to acknowledge that at the point in time we do not have complete data to form a decision, and we probably never will. We have the map and our anecdotes through friends, family, social media, etc. that forms our general understanding. If our only notion of the pandemic comes from sensationalized national news sources or state and national leaders, we substitute our own understanding for, at the very best, hypotheses, and at the very worst, biased opinions. Without forming our own initial baselines, we substitute what we’ve heard for what we know. This can be incredibly harmful.

This is not to say that what we do already hypothsize – the collection of anecdotes, data, inferences, and prognostications – cannot be useful. These discoveries, when practiced with pragmatic reasoning and probabilistic thinking (EE5), are the basis for finding a solution. But don’t ever confuse the basis with the final picture. The map is not the territory.

In return, try to view the COVID-19 pandemic as a healthy reminder that each of us flawed. We should take into practice the advice of Kahneman and Tversky, and use our pragmatism, determination, humility, and empathy to work towards the eradication of this particular coronavirus. Because the reality is, this is not the last viral complication we will encounter as a species. If we can learn from our mistakes, acknowledge our biases, and collaborate, I believe there is no problem too large or too complex for us to solve.

Leave a Reply

Your email address will not be published. Required fields are marked *