Institutional Racism and Data

The week preceding Easter is called Holy Week.  But for many in Detroit, New York City, and New Orleans, it was a grim week as the death toll from COVID-19 set records for consecutive days.  On Maundy Thursday, the commemoration of the Last Supper, the number of deaths in New York State hit a record 799 in one day.  On Good Friday, recognized as the day of Jesus’ death by crucifixion, the number of confirmed COVID-19 cases in New York State, 174,500, exceeded that of any foreign country in the world, and that number includes only those tested.

As the number of deaths from COVID-19 increases, reports are coming in about the relatively high impact on menBlacks and Latinos, and Native Americans.  Why COVID-19 disproportionately impacts those communities is multifaceted but likely related to high rates of service sector employment and low wage jobs, higher rates of diabetes and hypertension, and less access to healthcare.  A common thread for these contributing factors is a long history of institutional racism in America.

Last August, I wrote, “Institutionalized exclusion is not just entrenched in the equitable access to quality education, employment, healthcare, housing, and a host of other goods and services, but is also rooted in how data are collected and reported.”  When the headlines read “In N.Y.C., the Coronavirus Is Killing Men at Twice the Rate of Women,” who are these men? Are they Asian? Black? Hispanic?  If you wonder why it matters, read this week’s WISER-Op, “Show Me The Data, But Disaggregate It First.”

Stay safe and well,
Rhonda V. Sharpe
President

Show Me the Data, But Disaggregate It First

By Rhonda Vonshay Sharpe

Last week, Rep. Ayanna Pressley and Sen. Elizabeth Warren requested comprehensive demographic data on people tested and treated for COVID-19. The legislators cited racial and ethnic disparities in health outcomes as the rationale for the request. But their request does not go far enough.

While all federally funded research should report data by race, it also needs to include gender, class, and other characteristics.  It’s simply not enough to look at the numbers based on age or race or gender, we must look at the interactions of all of those factors at once.

For example, the state of Michigan reports COVID-19 data separately for race, ethnicity, gender, and age.  We know how many Black people have been diagnosed or have died from COVID-19. We know how many people over 65 have been diagnosed or have died from COVID-19. But we don’t know how many Black people over 65 have been diagnosed or have died from COVID-19. Their data do not show how many of the deaths have been Black men, or how many of those 60 years of age or older were Black women and poor. Data aggregated by race/ethnicity and gender masks disparities in health outcomes and obscures insight into causative factors.

An intersectional approach takes into consideration the fact that people often fit into more than one of these categories. When data are disaggregated this way, the results can inform policies to increase health and economic outcomes of more vulnerable groups.

Disaggregated COVID-19 data could reveal nuances in health outcomes that may be the difference between life and death.  Disaggregating data saves lives because it provides health care professionals and policymakers with detailed information to tailor both treatment and prevention strategies.

We know that CDC reports state that people age 65 years and older are at a higher risk for severe complications after contracting the coronavirus. Other vulnerable groups include those with chronic lung disease, asthma, heart conditions, a compromised immune system, diabetes, severe obesity, and liver disease. However, when typically presenting their data, the CDC doesn’t thin-slice its reports the number of women with asthma or how many Asians have diabetes. It’s the same for hypertension and heart disease, although their reports on COPD and obesity are broken down this way.

Disaggregating data moves us away from gender-bias (where men are the norm) and racial-bias (where Whites are the norm) in how we report data. Studies that use an intersectional approach acknowledge the complex diversity of lived experiences of the U.S. population. The ability to do that is crucial when you’re trying to provide critical care in a crisis such as the one we’re in.

Health care professionals, policymakers, public health experts, and others need to see the information broken down. Time and time again, doing so has led to essential medial discoveries that save lives.

How did we learn that Black women have a lower incidence of breast cancer but a higher mortality rate? Or that Black men are more likely to get prostate cancer and twice as likely to die from it? We disaggregated data by gender and race.

How did we learn that Black men have the highest rate of new cancer diagnosis of any group? The data were disaggregated by race, ethnicity, and gender.

Rebecca Dixon, Executive Director of the National Employment Law Project, says that economic strain has historically provided an opportunity for bold policy changes.  She suggests this pandemic might be one such chance.

I agree.

This pandemic gives Congress the impetus to mandate a change from the standard of surface-level reporting of race/ethnicity and gender data to highly-detailed disaggregated reports that capture the nuanced differences in data outcomes.

In the long-run, disaggregating data can inform policies to improve the health and economic outcomes of the most vulnerable Americans.  In the short run, disaggregating COVID-19 data on testing, treatment, and treatment outcomes will save hundreds – if not thousands – of American lives.