As the novel coronavirus pandemic continues to sweep across the world, its impact on the United States has been dramatic and disproportionate to the overall size of the country, the third most populous in the world. With 4.3 million cases and nearly 150,000 deaths as of this writing, the US has more deaths than any other single country. If the US experienced coronavirus at the same rate as the world as a whole (2,328 cases per 1 million population), it would only have a bit more than 750,000 cases.
Within the US, however, numbers of both cases and deaths also have not been evenly distributed among the population. Racial and ethnic disparities have been prevalent since the early days of the pandemic and have grown more pronounced as the virus has spread.
To help understand the overall picture of how coronavirus is impacting different populations in the US, we turned to SIS professor Rachel Robinson, a sociologist and demographer who focuses on global health. We asked her to help unpack the numbers behind the disparities we see.
Q. Rachel, the big question is: how disproportionate are the numbers of coronavirus cases among Native American, Latinx, and Black populations related to their overall size of the US population?
The impact of coronavirus is not distributed evenly across the US population. In most states, the percentage of total cases among Black, Latinx, and Native American populations exceeds the percentage of the population made up by each of those subgroups. In other words, the rate of coronavirus is higher among these groups than it should be. Thankfully, the disproportionate impact of coronavirus has received more media coverage than often given racial/ethnic disparities in health outcomes, so the general public is more aware than might usually be the case. But the way that COVID statistics are reported can still make these disparities difficult to see.
In particular, almost all reporting of COVID statistics—be it cases, deaths, or testing—are presented as raw numbers. Although it’s important to know these numbers, such figures obscure the true risk for COVID among subgroups because they don’t include denominators. For example, there were on average 70 new cases of coronavirus per day in the District of Columbia during the last week of July. Absent any context, this number loses its meaning. However, the size of the population in which those cases occurred is crucial to know. During the same one-week period, there were on average 275 new cases per day in New York City. So it seems that New York is doing worse. But that’s not actually the case, as the population of New York is 8.3 million, while that of DC is only 700,000. The rate of new cases per day in DC is 9 per 100,000 population, but only 3 per 100,000 in New York City—a threefold difference.
Those over 65 form another population disproportionately impacted by coronavirus. Eighty percent of deaths have occurred within this subgroup. Thus, a state that has more people over the age of 65 will look like it has a “worse” epidemic than one with a younger population, even if you take into account the size of the state—the denominator problem from above. Demographers have a trick for addressing such difference, called age standardization, but this is not generally being applied to mainstream reporting of coronavirus cases or deaths. Happily, people under 15 are disproportionately avoiding coronavirus.
There are numerous imperfections in coronavirus reporting – the number of cases comes from those who have reason to seek a test, because they have symptoms or have plausible exposure, and the number of deaths relies on accurate identification of cause of death. Death reporting in a country as large as the US is already complicated; it is made more so by the fact that deaths are recorded at the state level first and then must filter up to the national level. Furthermore, we know that racial discrimination plays a role in cause-of-death reporting: cirrhosis decedents are more likely to be recorded as American Indian on their death certificates, and homicide victims are more likely to be recorded as Black. Therefore, one of the best ways to identify the full extent of coronavirus mortality is to compare the number of deaths in a given week to that same week from a year prior. While this method is not perfect (for example, flu mortality also varies from year to year), it gives a better sense of what demographers call “excess” mortality. The Financial Times has one of the best sets of graphics depicting the data in this manner. This technique may be one of the only ways to understand the extent of the epidemic in countries with weak data collection infrastructure, including most developing countries.
Q. What population is the most disproportionately affected by the coronavirus? Which population follows? And third?
The rates of coronavirus deaths are most disproportionate for Black populations, followed by those for American Indians and Latinx populations. Specifically, the per capita rate of death for Black people is 2.5 times greater than for white people, and 33 percent higher for American Indians and Latinx people as compared to white people. Looking across US states, there is no state where the percentage of deaths to white people is greater than the percentage of the population made up by white people. To be clear, though, these differences are entirely the result of the social determinants of health, and not due to any biological difference that exists by race/ethnicity.
Q. As a demographer and a global health scholar, does anything about these numbers surprise you?
Sadly, these numbers are not surprising. The burden of poor health is disproportionately borne by those who face the most disadvantages in society, which in the US tend to be people who have lower incomes or come from Black, Latinx, and other communities of color.
Q. Briefly, what do we mean by “social determinants of health”?
The social determinants of health refer to conditions in which people are born, grow, work, live, and age and the wider set of forces and systems shaping the conditions of daily life. They can be economic, political, and social. So, income, gender norms, and racism are all social determinants of health.
Q. When a person understands the concept of social determinants of health, the underlying causes of these disparities become more easily understood. But not excused. What, in your opinion, could be done to lessen these disparities as the pandemic moves into the second half of 2020?
First and foremost is for key leaders—from the president to governors, senators, Congressional representatives, and on down—to develop and spread a message of “we’re all in this together.” We cannot stem the pandemic without each of us seeing every single other human being as having equal value to ourselves. Sadly, the political climate in this country and in many others discourages such thinking. When we prioritize the wellbeing of entire communities, it suggests a number of productive policy and programmatic steps that will not only facilitate the end of the coronavirus pandemic but will also decrease inequality and make our communities stronger. Examples include ensuring that all people have access to health care that they can afford and which prioritizes preventative health, basic income provisions and minimum wages high enough to actually support people, and efforts to dismantle systemic racism. Even though making these types of changes is hard, the first step is to believe in their value and importance, which is something that we can all do, and apply those values to the choices we make as we vote, pressure our elected officials, and strive to make changes in our own institutions and communities.
More useful resources