" "

Related Expertise: Diversity, Equity, and Inclusion, Public Sector, Social Impact

Bridging COVID-19’s Racial Divide

By Amanda BrimmerMarin GjajaDan KahnBryann DaSilvaKedra Newsom Reeves, and Marisa Gerla

Against a backdrop of protests and a national conversation about systemic racism, disproportionate numbers of Black and Hispanic people in the US are dying from COVID-19. The prevailing view is that the higher number of deaths is due to three factors: underlying health conditions, the lack of access to quality health care, and exposure to the virus.

But our data analysis reveals that is not the case. Neither underlying health conditions nor the lack of access to quality health care has played a primary role in the discrepancy, though the latter was an issue early in the pandemic, when the lack of health insurance prevented many people from seeking treatment. Rather, the chief reasons for the disproportionate number of deaths are the greater risk of exposure to people with COVID-19 and less access to COVID-19 testing. By our calculations, these reasons account for about 85% of the disparity. Underlying health conditions, age, and the lack of access to quality health care account for the remaining 15%. (See Exhibit 1.)

The widespread misinterpretation of the data has had grave consequences. In particular, it has distracted policymakers from the immediate imperative to address the systemic racial disparities that are related to exposure as well as testing access. As a result, COVID-19 is again surging among people of color. Hispanic communities in the southern region of the US and in California are getting hit especially hard.

By contrast, Chicago’s journey toward health care equity in fighting COVID-19 shows that more can and should be done to safeguard communities of color. To reduce the spread of the coronavirus among people of color, it is vital to protect those who are essential workers and who are in high-contact occupations. It is also critical to expand the number of testing sites in communities of color and ensure sustained access to quality care.

Coming to Grips with the Data

To get a true sense of the impact of the pandemic on the lives of people of color, we analyzed data on COVID-19-related deaths, infection fatality rates (IFRs), and infection rates for Black, Hispanic, Asian, and white people in the US. We then investigated the root causes for the disparities among the different groups. (See “About Our Research.”)

About Our Research

To get a better understanding of the impact that COVID-19 has had on people of color in the US, BCG analyzed data from the US Census Bureau and the Centers for Disease Control and Prevention (CDC), including data collected by the CDC’s Behavioral Risk Factor Surveillance System (BRFSS).

Synthesizing various analyses and expert interviews, we determined the expected Infection Fatality Ratio (IFR) by age group and by the presence or absence of a relevant underlying health condition. We then applied these rates to demographic data from the US Census and the BRFSS for each racial group by state to determine the expected IFRs by group, adjusted for age and underlying health conditions.

Using these expected IFRs and reported deaths, we estimated the number and rate of infections for each racial group. The number of infections is calculated by dividing the actual number of deaths by the expected IFR.

It was also possible to estimate the root causes of the death difference between people of color and white people. We compared the infection rate for each racial group with the distribution of deaths reported through June 24, 2020, to find the portion that can be explained by age and underlying health conditions. When we repeated this analysis with data reported through July 15, 2020, the results were nearly identical, so this is likely to be a sustained trend.

Subsequent to our demographic analysis, we methodically evaluated various root causes in sequence to determine which ones were most important: underlying health conditions and age, the lack of access to quality health care, or greater risk of exposure to COVID-19 and less access to testing.

The Number of Deaths. The data is sobering. For the US population, ages 18 through 65, Black people account for 32% of all COVID-19 deaths, despite making up only 14% of that population. (See Exhibit 2.) When we compared the number of COVID-19 deaths per 100,000 people aged 18 through 65 for each group, we found that the number of deaths for Black people is 5.8 times higher than it is for white people, while the number of deaths for Hispanics is 4.2 times higher than it is for white people.

Regional data yielded similar insights. Expectedly, the number of COVID-19 deaths for Black people and for white people varies significantly from state to state. But the number of COVID-19 deaths for Black people is higher than the average number for the general population in almost every state reporting this data. (See Exhibit 3.) In urban areas, the number of deaths for Black people is about 1.5 times higher; in rural areas, it is four times higher.

IFR. Epidemiologists define IFR as the percentage of people infected with the virus—regardless of whether they have been diagnosed—who are expected to die from it. In July, the World Health Organization stated that the current consensus global IFR for COVID-19 is .6%.

The IFR for a group depends primarily on the prevalence of underlying health conditions and age distribution. We estimate that for Black and Hispanic people in the US, the IFR is respectively .4% and .3%, slightly lower than the IFR for white people (.5%). For Asians, it’s even lower (.2%). The demographic base of older white people accounts for the majority of this estimate.

Infection Rates. Given US testing limitations and the high prevalence of asymptomatic cases, we have to estimate the true infection rate. Estimated infection rates depend on IFRs and the reported number of deaths per racial group. By our calculations, the infection rates for Black and Hispanic people are respectively three and two times higher than they are for white people. (See Exhibit 4.) This is important because it shows that infection rates for people of color are the primary reason for disparities in COVID-19 outcomes.

The Key Factors Influencing Racial Disparities

Why are death and infection rates so much higher for people of color than they are for white people? There are four factors to consider: the greater risk of exposure, less access to testing, underlying health conditions and age, and the lack of access to quality health care.

Greater Risk of Exposure. Of the various factors that are linked to infection, exposure to the virus is the most important because people of color are at significantly greater risk of exposure for various reasons related to social determinants of health. They are, for example, significantly more likely to live in multigenerational housing, use public transit, and work in environments where they are more exposed to the virus.

While only 16% of white people live in multigenerational households, that number is 60% to 80% higher for people of color. Nearly 30% of Asians live in multigenerational housing. In addition to multigenerational households, people of color in congregate living facilities (especially nursing homes, treatment centers, and prisons) have been hit hard by the coronavirus. To date, 60% of nursing homes that have a significant population (more than 25%) of people of color have reported COVID-19 cases, as opposed to 30% of nursing homes that have a small population (under 5%) of people of color. Prison residents are testing positive for the coronavirus at five times the rate of the general population.

Working conditions are especially problematic. Black and Hispanic people make up a disproportionate share of two groups of workers that have the highest risk of exposure.

  • Essential Workers. The largest group that has a high risk of exposure is essential workers. This group spends many hours each day in places such as grocery stores, pharmacies, and hospitals. Essential workers make up approximately 33% of the overall workforce and 45% of the workforce that is people of color.
  • Workers in High-Contact Occupations. The other group is made up of workers who have a high risk of exposure because they contend with very crowded workplaces—such as factories and hotels. Workers in these environments are disproportionately people of color. For example, people of color represent about 60% of employees in the meatpacking industry but account for almost 90% of the confirmed COVID-19 cases in this sector.

Less Access to Testing. The data on testing reflects wide disparities between people of color and white people. The inequities have resulted from requirements for testing, access to tests and testing sites, skepticism about testing, and monolingual outreach.

  • Requirements for Testing. Early in the pandemic, many states made having a prescription mandatory for getting a test. This posed a challenge to many Hispanics who do not have a personal doctor (20% report not having a personal physician or health care provider, compared with approximately 5% of white people). However, that issue largely dissipated when many states relaxed the rule about prescriptions.
  • Access to Tests and Testing Sites. On average, zip codes for areas predominantly made up of people of color have approximately 15% fewer testing sites than zip codes with mostly white residents. Moreover, the number of administered tests per capita in zip codes with mainly people of color is only 25% higher than the number given in zip codes with mainly white people. If the number of administered tests was based on infection rates, zip codes with predominantly people of color would have two to three times the number of tests than the average in zip codes made up of mostly white people. Clearly, an insufficient number of people of color are getting COVID-19 tests. Why? We believe several factors are at work. Some people with low incomes cannot afford to take time off while waiting for results, as some employers require. Additionally, recent studies suggest that people of color face much longer wait times, and communities of color are less likely to have a sufficient testing supply than communities with mostly white residents. In other instances, well-intentioned efforts, such as Delaware’s drive-through testing sites, could not be accessed by local residents who lacked a car. Regardless of the cause, however, the result is the same.
  • Skepticism About Testing. Some people have been reluctant to get tested because they mistrust the government. That may be because long-standing racial inequities in health care have made some people reluctant to give authorities access to their private data. The consequences of being tracked and traced may be especially worrisome for undocumented immigrants who are at risk of losing their jobs or, worse yet, deportation.
  • Monolingual Outreach. Hispanics who speak only Spanish may have missed out on testing simply because outreach has been primarily in English. Given that approximately 20% of Hispanics in the US don’t speak English, the number who have missed opportunities to get tested could be significant.

Underlying Health Conditions and Age. People who have chronic health conditions, such as heart disease or diabetes, or who are over the age of 65 are especially vulnerable to COVID-19. If infected, these health-vulnerable individuals have a high risk of succumbing to the disease. Since a higher percentage of Black adults have underlying health conditions than do white adults (30%, compared with 25%), it is often assumed that underlying conditions are responsible for the higher number of COVID-19 deaths among Black people. (This does not apply to Hispanics, since only 20% have relevant chronic health conditions.)

That assumption is incorrect, however, because Black adults with underlying conditions are typically younger than white people with chronic conditions, largely offsetting the influence that such diseases have on the death rate. Approximately 25% of white people are over the age of 65, but only 16% and 9% of Black and Hispanic people, respectively, are over the age of 65. Overall, higher health vulnerability is not a significant cause of the higher number of deaths among people of color; by our estimate, it accounts for less than 5% of the difference.

Lack of Access to Quality Health Care. The last factor to be considered, health care access, has three key drivers:

  • Health Care Capacity. Our county-level analysis shows that, overall, counties with a predominantly white population have a similar number of hospital beds and a similar ICU capacity per capita as counties with predominantly people of color. While there is material variance in some regions, it is not enough to suggest that health care capacity has played a major role in COVID-19 death disparities, even though it is a well-documented contributor to other chronic health outcome disparities.
  • Health Insurance. In the early weeks of the pandemic, health insurance coverage may well have limited access to health care: higher percentages of Black and Hispanic people in the US are uninsured and may not have sought out testing or treatment because of the large out-of-pocket costs. But this should no longer have been an issue by late April, when federal programs and insurance waivers began covering out-of-pocket costs for COVID-19 testing and treatment. Since there continue to be large disparities in death rates, we can rule out health insurance as a major factor.
  • Quality of Care. The number of Black and Hispanic patients hospitalized with COVID-19 per 100,000 people is four times that of similar hospitalizations for white patients. But a much larger percentage of hospitalized white people have died than hospitalized people of color (20%, compared with 13%). This suggests that the quality of the care a person receives in the hospital is not likely a primary reason for the higher number of deaths for people of color, though it appears to be a factor in some locations. Still, the significantly higher number of hospitalizations for people of color is cause for concern. It could be due to the fact that many Black and Hispanic people lack a primary care provider or access to a community clinic. (See “Chicago: A Case in Point.”)

Chicago: A Case in Point

The city of Chicago offers an illustration of how government and society can work together to turn around health care disparities and achieve more equitable outcomes.

Because it was collecting COVID-19 data by race, Chicago identified a wide disparity for the testing positivity rates between two groups: people of color and white people. In early April, the rate for white Chicagoans was 38%, but it was 23 and 36 percentage points higher for Black and Hispanic Chicagoans, respectively. (See the exhibit.)


So, the city launched the Racial Equity Rapid Response Team (RERRT), a public health outreach and education campaign, to develop recommendations, mobilize community partners, and prioritize resources for communities with minority populations. The RERRT worked with local leaders to promote hand washing, encourage testing, and hand out free masks and hand sanitizer. It also required hospitals to share race and ethnicity data about COVID-19 patients.

Since the launch of the RERRT, the testing positivity rate for Black Chicagoans has come down to nearly the same as the rate for white residents. The rate for Hispanic Chicagoans is still about 10%, which indicates that difficulties persist in some communities of color. The RERRT’s work in these communities is ongoing.

Policies for Addressing the Disparities

We believe that a robust set of actionable policies is a good starting point for addressing the enormous racial disparities around COVID-19 outcomes that exist in the US today. The following list of policies is by no means exhaustive—see our fuller set of potential policies to protect the vulnerable. We mention these because they are especially relevant for communities of color and could help lower their overall death rates significantly.

  • Distribute masks. Governments should dispense masks to all who are health vulnerable (and those who are close to them), who have a greater risk of exposure, and who suffer economic hardships. Additionally, officials should strongly encourage everyone to wear the masks. Distributing masks and enforcing that they are worn is the single most cost-effective tool that policymakers can use to fight the disease. But people of color, especially Black men, have reported harassment and a fear of being profiled if they wear a mask. Therefore, in areas with predominantly people of color, local officials should strongly encourage that people wear masks but may choose not to require it. Additional research should be done to assess the efficacy of face shields or other solutions as an alternative to masks.
  • Conduct weekly tests. Governments should test people who have a high risk of exposure at work as well as those who are in close contact with people who are over the age of 65 or who have underlying health conditions. There are three important practices to observe. First, normalize the number of testing sites and tests across regions on the basis of test positivity rates (the number of people who tested positive divided by the number of people who were tested). Communities of color should have the same test positivity rates as communities with predominantly white people. Having more-accurate numbers will provide a clearer picture of how many more testing sites and tests are needed to ensure that people of color have adequate access to testing. Second, increase awareness of the importance of getting tested. Improving outreach and multilingual guidance are key. Third, for outreach and testing sites, partner with organizations that people trust, including community groups, churches, nongovernmental organizations, and public services. Testing sites should not require IDs for testing because it could inadvertently deter undocumented people.
  • Mitigate the risk of contracting the coronavirus in the workplace. Governments should establish, monitor, and enforce strict occupational safety protections for people who work in environments where the risk of exposure is high. Penalizing noncompliance would improve working conditions in places (such as meat processing plants) where large numbers of people are at high risk. Governments can also provide funding for safety measures—such as N-95 masks, safe transportation options, and opportunities for health-vulnerable workers to be voluntarily deployed elsewhere—thereby incentivizing companies to implement them.
  • Limit the risk of spreading the coronavirus in high-risk residences. A tragic number of deaths have occurred inside congregate living facilities. Congregate living facilities must systematically implement regular testing for all residents and staff, strictly limit in-person visits, and establish safe quarantine options to apply in the event of a coronavirus outbreak.
  • Expand the number of community health centers. Some 2,000 federally funded health clinics in communities with predominantly minority populations have shut temporarily, with many more on the edge of permanent closure as their financial prospects dim. It’s critical to take the steps needed to reopen these facilities and, if possible, expand the number. Policy measures that are proven to increase trust in medical services are also important. Community health workers (CHWs) are a good place to start. Trained and trusted, they can provide valuable health coaching and social support. Hospitals around the country now depend on CHWs, who have been shown to reduce the number of unnecessary emergency-room visits and hospitalizations.
  • Make sheltering in place more sustainable by providing food, counseling, and social connections. Simply asking people of color in health-vulnerable households to stay home from work is not the solution. Many people cannot give up their means of support. Subsidized or free contactless delivery of food and other essential services will be critical to enable people to stay home if they are sick or test positive.

The pandemic’s impacts on people of color go far beyond health issues. Economically, the crisis has led to people of color being laid off or furloughed in massive numbers. Approximately 24% of Black people and 20% of Hispanics have reported losing their jobs; those percentages are nearly twice the percentage of white people (11%) who have lost their jobs. According to the US Department of Labor, the gap between the unemployment percentage for people of color and for white people is the widest it’s been since May 2015.


The long-term consequences of COVID-19 for people of color are yet to be determined. But it is clear that the longstanding racial inequities that it has exacerbated are likely to keep growing if they are not addressed. As Black and Hispanic people continue to contract the coronavirus at two to three times the rate that white people do, the disease will spread further in communities of color. This will create a downward spiral of more lockdowns and more job losses, with the attendant problems of inadequate food and housing. And whenever COVID-19 cases surge in the future, they will disproportionately afflict people of color.

COVID-19’s impact on the lives of people of color has been broad, deep, and often misunderstood. It is past the time to do something about it.

The authors thank their colleagues for their contributions: Venkat Raman, for his research, analysis, and assistance in developing this article; Aradhana Parikh, for her research and demographic analysis; and Kelsey Hayes, for her research and help in identifying the policies.