We’ve Got it All Wrong: The Fatal Global Error in Assessing COVID-19 that Will Cost Many Lives Unless We Correct It—Fast
I’m sure you’ve heard something along these lines many times in the past few weeks: “It’s really sad what’s happening in ______ (Italy; New York; US).” Basically, many of us consciously or subconsciously seek out a country or a region of our own country where the coronavirus is wreaking more damage than where we live, and we feel pity for the unfortunate souls living there.
Why We Compare
As it turns out, we human beings have a strong need to form accurate, stable perceptions of our abilities and life outcomes. When it’s difficult to find objective standards with which to evaluate ourselves in terms of how successful, creative, or happy we are—or, in this case, how likely we are to die from COVID-19—we seek out similar others (called our “reference group”) and then measure ourselves by how we are faring in life relative to them. The phenomenon of social comparison was first identified by the social psychologist Leon Festinger. If we’re doing better than similar others in any specific dimension of our lives we feel insecure about, we can raise our self-esteem by comparing ourselves to them.
Conversely, if we are the ones doing worse than others similar to us, our self-esteem diminishes. It is for this reason that students are likely to have a higher academic self-concept if they are attending a high school in which most students are average performers. Once they graduate and go to a more demanding university—as Oxford University professor Herbert Marsh and his colleagues found in a longitudinal study of Hong Kong students—their academic self-concept tends to decrease.
As happens in most casinos in the world, you enter social media seeking to enhance your self-esteem, but the cards are stacked against you. The current problem is worse than ever because the casinos have now entered our homes, schools, restaurants, stadiums, and every other place you can access the Internet on your phone. Never before has the social information of similar others been so readily available for you to digest than since the genesis of Facebook and other social media sites where you can bring these folks into your network under the astonishingly ironic title of “friends.” Yet, as in the casino, your chances of winning the social comparison game—whether it’s about who has the best job or hottest body or, currently, the lowest chances of being infected by the coronavirus—are infinitesimal. It is for this reason that a study of 584 Facebook users in Germany found that the most common reported emotion among Facebook users is envy.
So that’s why we make social comparisons, and prefer to compare with others less fortunate than ourselves. In a time of crisis, our primal instincts kick in even more. Given that COVID-19 is the largest global public health crisis in most of our lifetimes (at age 52, it certainly is in mine), it’s safe to say that our need for safety and security is compelling us to engage in these acts of social comparison—scrolling through social media feeds and checking websites for the latest number of people who have been infected by or have succumbed to the coronavirus—like never before.
Yet what if how we make these social comparisons is all wrong? The following statement, which goes through many of our minds after the initial statement above—“Too bad about what’s happening in _____ (other place from which we are a safe distance)” is often, “I’m going to go play soccer with a friend,” or “I’ve still got to go in to work; it’s too difficult to work remotely from home,” or “I’m not interacting with many people, but I do want to at least see so-and-so.” In other words, the social comparison line of thinking leads us to relax the new social distancing norms.
Two Critical (Overlooked) Factors to Consider When Calculating Your Own Risk of Infection
Hold on a minute. If that’s the case, and the anecdotal stories coming my way from hundreds of people—friends, clients and students alike—seem to suggest that it is, then if we are not playing the social comparison game accurately it may cost many their health and possibly even their lives. So let’s see how our social comparisons about COVID-19 are based on false reasoning. Here’s how it happens: we find another region or country with higher numbers of infections or deaths than the city, province or country we are living in, and feel better about all this craziness around us. Sound reasonable? As I will show you, it isn’t; in fact, it’s a grave mistake.
As it turns out, if you want to compare the dangers of the coronavirus in your region to those of another, what is important is neither the current daily or total number of cases or deaths in the (other) region. That’s important information, but you are comparing apples and oranges if you compare it to the numbers in your country or region. Why? For two primary, interrelated reasons, one easy to understand and the other a bit harder.
First, the easy one: Your region or country does not likely have the same population as the region or country you are comparing it with. For the sake of illustrating the principles in this article, I will compare the United States and Mexico. (Other comparisons can be made using the same procedures below. Unless otherwise mentioned, all of the data in this article are from the Worldometer website.) As of March 31st, there have been 4,053 deaths from COVID-19 in the US and 28 deaths in Mexico. Does this mean that you are 145 times more likely to die from the coronavirus in the US than in Mexico?
Given the lack of social distancing in some regions of Mexico, one would think so. Mexico is a very social culture in which face-to-face interactions are regular, robust and vital. Many Mexicans do not own credit cards or, if they do, have never ordered anything online—especially not required daily items such as food. Marketplaces teem with people and are as social as they are economic. That’s why I love Mexico so much: there are tons of people on the street at all hours and it’s easy to feel connected with the people around you. There is a sense of community that has become increasingly difficult to find north of the border, especially as social media companies have preyed on our individualism and wedged us further apart from each other along political, economic, racial, age and gender lines.
Yet Mexico and the US have vastly different populations (129 million versus 327 million, respectively). So can we simply adjust the percentages for total population and conclude that you are 57 times more likely to die of COVID-19 in the US than in Mexico?
This adjusted comparison at least helps us to better understand the different relative chances of being affected by the coronavirus in these two countries. Numerous comparisons do not even take this vital consideration into account, leading us to very erroneous conclusions. For example, a Politico article published a few days ago states that it “used global data from John Hopkins University to explore [how] to track countries that are succeeding — or failing — in ‘flattening the curve.’” The authors then propose a metric that displays, in real time, how often an individual in each country is infected with the virus. It then states that “One person tests positive for COVID-19 every…” and shows a list of countries, beginning with the US (4 seconds) and Spain (11 seconds).
Notwithstanding non-reported infections, this information is accurate. Yet let’s consider how the average person will interpret this data: that an individual in the US is 2.75 times more likely (11 divided by 4) to contract COVID-19 than an individual in Spain. Now let’s adjust the reported data based on the respective populations in the US (327 million) and Spain (47 million). It turns out that, as per the very same data reported in this analysis, the interpretation flip-flops: an individual in Spain is over 2.5 times as likely to be infected by the coronavirus today than an individual in the US. These are the types of cognitive errors we are making every day when we look at many of the charts emerging in front of us by the second on social media.
Yet even if the Politico analysis and numerous others were to adjust for population, they would still be stunningly inadequate. Why? This is the second and slightly harder reason to understand. COVID-19 is a highly contagious virus that is not unlike a snowball rolling downhill: the longer it gains momentum unhindered, the more people it infects. Its growth, if strict containment procedures are not observed, is exponential.
A better way to compare two regions or nations, then, would be to look at the respective trajectory of COVID-19 in these two regions or countries. Since we started with the comparison between the US and Mexico, let’s continue it more accurately (with the caveat that it is very early in the coronavirus’ trajectory in Mexico and these indicators can dramatically change at any time).
First, the baseline data: the first cases of COVID-19 in the US and Mexico were January 21st and February 28th, respectively. The first death by coronavirus was on February 29th in the US and on March 19th in Mexico. As of March 31st, Mexico was experiencing its thirteenth day since its first coronavirus death. Let’s compare how far the Mexican snowball had rolled in 13 days (inclusive) versus the American snowball 13 days after the first death in the US.
In Mexico, on March 31st there were 28 deaths by COVID-19 whereas on March 12th (13 days after the first death on US soil) there were 41 deaths in the US. These two populations, however, once again are very different. To more closely compare apples and apples, let’s calculate the chances of dying of coronavirus in each of these two countries during the first 13 days following the first death in each country based on the number of deaths divided by the total population. In the US, these chances were 1.3 out of 10 million and in Mexico they were 2.2 out of 10 million. This means that if you live in Mexico, your chances of dying of COVID-19 in the 13 days following the first fatality were 73 percent higher than your chances of dying in the 13 days after the first fatality if you live in the US.
All of a sudden, the social comparison changes. “Yes, but it’s the total number of infections that’s important,” you may be thinking. “That snowball slid much further much more quickly in the US than in Mexico.” Possibly. Again, let’s try to better compare apples and apples. As of March 31st, there were 1,094 infections in Mexico. (While infection rates are probably heavily underreported across the globe, the percentages of asymptomatic cases are unlikely to be significantly different from country to country.) March 31st is 33 days (again, inclusive) after the first infection in Mexico (Feb. 28th). The relevant question, then, is how many infections were there in the US 33 days after the first case? The first case in the US was on Jan. 21st. Thirty-three days later, on Feb. 22nd, there were a total of 35 cases. In other words, in its first 33 days COVID-19 has been spreading in Mexico at a velocity 31 times faster than the rate at which it was spreading at the same point in its trajectory in the US.
Slowing the Snowball
So does this mean that there will be more infections and deaths in Mexico than in the US? Not necessarily. First, we have insufficient data and it’s too soon to make any predictions. Second, there are other factors that will influence the snowball’s trajectory, ranging from personal health, the public health system and age to population density, population dispersion, climate and other local conditions. Yet it does mean that all is not as it seems when looking at the data shoved into our awaiting mouths by social media and news outlets as we take gulp after gulp from the digital trough. And it also means that Mexicans, like their other North American friends above the border, cannot afford to take social distancing and other countermeasures lightly.
If a giant snowball is rolling down a mountain toward a village below, we must determine how to stop it or at least slow it down. Every death is tragic and we must collectively, independent of where we live on this planet, work together to stem the flow of this deadly virus. Right now, experts agree that the most effective tool at our disposal at present is social distancing. Focusing on how to be remote but not isolated is a much better use of our time than making social comparisons in a futile attempt to convince ourselves that this threat is not as serious as it is.
If you are to pander to the very human instinct to make comparisons with other regions of your country or the world, at least do it more accurately—the purpose of this article—before stepping outside. To do so, I propose that we develop and measure variables similar to the following, which will enable us to make more accurate regional and national comparisons of the coronavirus’s imminent trajectory (which, after all, is what we are trying to predict so we can develop countermeasures to slow it) than absolute numbers of cases or deaths:
- The 10-day infection rate per capita (the number of infections in a specific region/country in the ten days following the first case divided by that region/country’s population), the 20-day infection rate per capita, and so on.
- The 10-day mortality rate per capita (the number of deaths in a specific region/country in the ten days following the first death divided by that region/country’s population), the 20-day infection rate per capita, and so on
While we still won’t be comparing apples and apples due to some of the regional variations mentioned above, we will at least be making a more accurate comparison that will help us understand what is to come if we don’t practice more stringent measures such as social distancing, hand washing, disinfecting whatever we bring into our homes, and so on.
These measures can of course be refined and made more nuanced, with associated charts and graphs, to better visualize these comparisons and evaluate the acceleration/deceleration of these rates over time. Most important is that we understand how the snowball generally rolls, and the effects of countermeasures to slow it at specific points in its trajectory, so we can develop a more accurate idea of what’s coming.
The globalized nature of our lives is a large part of what got us into this mess—and it will also play a pivotal role in what will get us out. Every human being matters, and only together can we collectively rise to this challenge— by first understanding the true nature of how and at what velocity this virus spreads, and then by each doing our part and taking compelling action to prevent it from snowballing any further and doing any more harm.
I am extremely grateful to the following people for reviewing this article for empirical accuracy and conceptual refinement: Dr. Peter Harms of the University of Alabama, Dr. Chao Miao of Salisbury University, Dr. Jeremy of Murray of California State University San Bernardino, Dr. Bradley Owens of Brigham Young University, Dr. Ronald Riggio of Claremont McKenna College, Kate Perry of the Center for Social Leadership, and my wife, Karla Silard.