States are releasing incomplete racial data on COVID— here’s why that matters

The last couple of weeks have been filled with headlines like these: “Milwaukee’s Covid-19 spread highlights the disparities between white and black”, “Black people in NYC twice as likely to die from COVID as white people”, and “Covid-19 is hitting Chicago’s black neighborhoods much harder than others, officials say”.

As of this writing, 32 states have released racial data for COVID-19 cases, and 27 for COVID-19 deaths. These findings confirm that Wisconsin, New York, and Illinois are not alone. In Mississippi, Black Americans represent 51% of COVID cases and 63% of deaths, despite being less than 38% of the population. In Georgia, they represent 26% of cases and 52% of deaths while only 30% of the state, and in Maryland, 37% of cases and 40% of deaths while representing just under 30%. These numbers do not account for the possibility that these gaps might be more significant if black communities are underrepresented in testing overall.

The rate of COVID cases is disproportionately high for Black Americans, and the rate of COVID deaths is even higher. This chart excludes states that have published racial data only on cases or deaths, not both. [1][2]

However, not all racial data follows the same pattern of disproportionate impact. In the states of Washington, Massachusetts, and Virginia, for example, the percentage of COVID cases in black populations falls right in line with the demographic representation in the total state population. But this apparent race proportionality is not cause for celebration. In these states, race data reporting is incomplete; race was reported in only 42%, 60%, and 36% of cases, respectively.

It is imperative that states release racial data, but to release incomplete data is insufficient. This is our ask:

1. States like Washington, Massachusetts, and Virginia that have begun to release data related to race and ethnicity must do a better job of collecting and releasing data that is timely, accurate, and as complete as possible.
2. States like New Jersey, Ohio, and Nevada that have yet to release any COVID data separated by race and ethnicity must begin to collect and release such data as soon as possible, especially given their large minority populations and high case counts.

Completeness of states’ COVID case and death data by race. Scale is 0% — 100%, where 100% indicates that the patient’s race is known for 100% of cases and deaths within the state.

States where racial disparity seems to be non-existent (or even favor Black Americans) are generally those that have reported incomplete racial data on COVID. Disparity was measured by the difference between % population and % of cases.

Why this matters

As seen in the cases above, better data may reveal what we are afraid to admit — that while some have called COVID-19 a “great equalizer”, it follows a historical pattern of disparities in health outcomes, driven by social, economic, environmental, and structural inequities that cannot be fixed overnight [4][5][6][7][12]. Black Americans in the United States are perhaps the most disproportionately hit, but only one of many communities to experience unequal impact by COVID. To fail to act on account that the root cause is structural, however, is not an option.

Death rates by major diseases in the United States from 2017–18. Black Americans suffer higher death rates than whites for many major diseases. [9]

The fight against COVID is far from over. To continue this fight, and to be smart in this fight, we need data to better understand where to direct our resources. We must be smart about how to focus prevention efforts, where to place and promote testing, how to fill gaps around care, and how to support those affected by COVID — physically and otherwise — as the recovery will be just as painful. It is also our belief that interventions cannot be “one size fits all,” but that there is opportunity for similar communities to learn from one another. While we may not be able to solve decades old issues of structural inequality overnight, we must recognize these realities and tailor interventions accordingly.

Sources: Data accurate as of April 20, 2020

1. All population data is sourced from the 2019 U.S. Census.
2. All COVID case and death data sourced from the respective State Health Department websites.
3. Disparity was measured by the delta between % population and % of cases. See Measuring disparity for more perspectives on how to accomplish this.
4. Heart disease: racial/ethnic disparities.
5. Cancer: racial/ethnic disparities.
6. Stroke: racial/ethnic disparities.
7. Diabetes: racial/ethnic disparities.
8. Race vs. socioeconomic status for health outcomes.
9. Other major diseases: death rate disparity.
10. https://www.brookings.edu/blog/the-avenue/2020/04/16/mapping-racial-inequity-amid-the-spread-of-covid-19/.
11. https://www.brookings.edu/blog/fixgov/2020/04/10/how-to-reduce-the-racial-gap-in-covid-19-deaths/.
12. https://www.ncbi.nlm.nih.gov/books/NBK425845/.