Don’t Racialize the Pandemic
Over the past couple weeks, the national news media and many Democratic politicians have been calling attention to the significant number of African Americans who have perished from the novel coronavirus. An April 7 op-ed by the New York Daily News even went as far as to call the virus a “black plague.”
The underlying logic of racializing the pandemic is that African Americans make up a disproportionate number of those who have perished from the disease in several major American cities (many are not closely tracking racial data, meaning that we have an incomplete picture). Disproportionate, in this context, means that the percentage of deaths from a group exceeds its percentage of the population.
For instance, in Houston, Texas, where African Americans make up around 22 percent of the population, seven out of eleven deaths through early April were African Americans, prompting the city’s health authority, Dr. David Persse, to note that there is a “disproportionate number of African Americans who seem to be suffering the worst consequences of the virus.”
We are still in the early stages of this pandemic, and the research on this topic is as novel as the virus itself — it’s difficult to draw hard conclusions about the causal factors that explain who gets the virus and who succumbs from it.
Persse hazarded his own guess, noting that there have been factors such as health conditions, including diabetes, hypertension, and heart disease, that appear associated with more severe coronavirus infections. He also noted that these conditions, as well as disparities in medical care, may explain some of the difference in the death toll.
Ironically, on the same day the New York Times published an article focused on African Americans that explained that the “coronavirus is infecting and killing black people in the United States at disproportionately high rates,” New York City revealed age-adjusted numbers showing that Latinos, not African Americans, had the highest age-adjusted coronavirus death rate.
Perhaps this will bring about calls that the Times have should have reframed its article around Latino death rates rather than black ones. Maybe another newspaper op-ed will call the virus a “Latino plague,” at least in New York City.
The city’s data also showed that, despite claims from New York City’s Public Advocate that “New Yorkers of more color” are disproportionately at higher risk, the Asian-American death rate is actually lower than that of whites.
One way to look at this data would be to scramble the racial hierarchy being assembled by liberal activists and the news media — Latinos actually have it “worst,” and Asians, another ethnic minority — have it “best.” But that would simply replace one form of distorted thinking for another.
Most Americans don’t have statistical training, and have a hard time parsing out the data when presented this way. The racial data presented by the city of New York, the New York Times, and many other media and political actors are what we call univariate, or rotating around one variable: race.
The problem with univariate data is that they don’t use controls. With only one variable being presented, you only have one explanation for a phenomenon.
Of course, practically nobody is arguing that African Americans or Latinos have a greater genetic predisposition to catching or dying from the novel coronavirus. That kind of scientific racialism is rare in the United States today (although there were some unfortunate rumors on social media suggesting black people can’t contract the virus).
But when we present our data with only one variable, we can end up flattening the image of an entire group and misinterpreting a social problem.
We’ve seen this story before, with crime. In many parts of America, African Americans are overrepresented versus their share of the population in crime statistics. For years, racist demagogues used this fact to claim that criminality was a feature of African Americans. But eventually, social scientists discarded simple univariate analysis and learned to correlate factors such as social inequality, the prevalence of the drug trade, and low-quality policing as the true culprits that drive crime — among blacks, whites, and virtually every other group. The news media, wisely, moved away from focusing on the race of criminal perpetrators, coming to understand that race is merely a social fiction, not useful for correlating to complex problems that impact people of all backgrounds.
We should remember that when we think about the pandemic.
A young, healthy African-American with great health-care coverage who telecommutes every day and has groceries delivered to his door is not automatically at higher risk of contracting the virus or succumbing to it than an elderly white man who rides the subway to work every morning to work at an essential business where he interacts with hundreds of people a day.
The causal variable here is almost certainly not race. Although the virus is not perfectly understood, it is much more likely that factors such as underlying health conditions drive death rates, not race. There’s also the reality that a virus spreads person to person. It will take detailed analysis by epidemiologists and others to understand how the virus spread from neighborhood to neighborhood. In the south Georgia town of Albany, for instance, we know that the outbreak emerged primarily from a crowded funeral. Why is there a much more severe viral outbreak in Detroit than Baltimore, despite some overlap in demographics? Univariate analysis can only tell us so much.
Some would argue that because the prevalence of these risk factors is different among different racial groups, we should still target racial groups in response. But doing so would fail the test I offered above: You’d end up over-delivering social concern and care for the person who faces little risk and under-delivering social concern and care to the person who faces great risks.
Our response to the virus should be as comprehensive as possible. We wouldn’t want to miss, for instance, the possibility of a tragic and destructive outbreak in the county that in 2016 held the title for America’s lowest male life expectancy due to its high concentration of health problems — that was West Virginia’s McDowell County, a locale that is around 90 percent white.
Rather than trying to draw a circle around a racial approximation of who has it “worst,” we should marshal our resources and sympathies to ensure that nobody squeezes through the cracks. It may very well be true that in many locations, the majority or a disproportionate number of the people who hold the actual causal risk factors are African American or Latino. But if we care for those people who hold the actual at-risk factors rather than trying to approximate them with racial generalizations, we should be able to respond in a way that encompasses everyone — whether they be black, Latino, white, Asian, Native American, or mixed-race.
If a group is disproportionately impacted, a comprehensive and universal response will make sure they are disproportionately aided. But such a response will aid everyone.
Furthermore, the progressives who are pushing racial narratives around the virus should think about the political costs of doing so. In many ways, the pandemic has united Americans, encouraging the kind of mass solidarity needed for social distancing and supporting emergency measures, such the over $2 trillion of relief spending Congress just authorized. Research shows that implicit associations between African Americans and poverty decreases support for aiding the poor — because people wrongly think that helping the poor with government assistance means helping blacks over whites. Universal, comprehensive policy doesn’t face that obstacle.
In some sense, the racialization of the coronavirus is part of a larger racialization of social and political issues in the United States. I fear that this racialization is giving life to racial categories — pretending that they are material, rather than a coincidental feature of a person.
Disproportionate statistical realities are often used to justify racialization. “X group is disproportionately impacted by Y,” this logic goes, therefore Y is a problem primarily for X group. This formulation is often politicized, which is something you can tell because of how unevenly it is applied. We don’t, for instance, discuss suicide as a “white problem,” even though age-adjusted white suicide rates are astronomically higher than black and Latino suicide rates. Prison isn’t a “male problem,” even though upwards of nine out of ten people in prison are men.
I’m not arguing that discrimination — both historic and contemporary — is not at least a part of the reason we have disproportionate outcomes in all these areas. Discrimination is a real problem we must confront to have the post-racial society we dream of.
But highlighting disproportionality is not an excuse to argue that some group has it “worst.” If you were to pick a white person and a black person who both had coronavirus, you would have to know more about their individual profiles to know who has it “worst,” if such a thing can even be argued. Given the thousands of factors that decide a person’s health outcomes, we are likely to see wide variations in how people within racial groups are impacted by the virus — which will paint a much more complicated picture than a simple, single-variable correlation between race and deaths.
People go through life as individuals, not as racial groups, and describing their lives with the language of racial groups is always, at best, an approximation. At worst, it leads to ugly stereotyping and racial essentialism. As University of Pennsylvania political scientist Adolph Reed Jr., a man of the Left, writes:
Racism is the belief that race is a category that defines and encapsulates natural populations. It does not. A claim that black people are especially vulnerable, as black people, to COVID-19 or any contagion is as preposterous as a claim that unicorns are especially vulnerable.
