Data on Racial Inequities in the United States
The recent killings of African Americans George Floyd, Ahmaud Arbery and Breonna Taylor among countless others have sparked pain, fear, and outrage among many communities throughout the United States. As one Kellogg family, with hallmarks centered around teamwork, collaboration, and inclusivity, we support and stand with our African American brothers and sisters and by extension other communities of color against hate, racial violence and institutionalized racism in the United States. These senseless acts of racial violence tarnish the pillars of diversity, freedom, opportunity, and democracy which have served and continue to make this country strong.
It is our hope that in our own small way, we as a student community are able to support one another in any way possible and also contribute to the necessary change required to transform our society as we have come to see it today. We plan to take a series of actions in the coming year to engrain equity into the core of how we operate as a club and community and we urge everyone to take steps to be intentional and lend support to the fight against racial injustice — both as individuals and as a collective student body. As a start, to help promote and encourage a dialogue around race in America, the Kellogg Data Analytics Club has helped compile an initial summary of data-driven research to aid communal education on racial disparities in the United States.
Section 1 — The issues
A. Racial Violence and Bias in Policing
From research, African Americans have been shown to suffer from adverse targeting and violence from police in the US. A study from 2015 looked at data across the United States on police shootings that includes a broader set of instances outside of only self-reported FBI incidents and found that the probability of being shot by police as an unarmed African American is about 3.5 times as high as the probability of being shot by police as an unarmed white American. This risk is particularly high in larger metropolitan areas with low incomes, high inequality and large African American populations. The statistic also controls for local levels of crime among different populations. Another 2018 study found that African American men have a 1 in 1000 chance of being killed by police in their lifetime, over twice that of white men.
An analysis from the New York Times recently studied police use of force in Minneapolis where George Floyd was killed. It found that Minneapolis police used force against African Americans at 7 times the rate it used force against whites; 60% of displays of police force in Minneapolis were against African Americans despite African Americans comprising only 20% of the population.
This type of violence and the discriminatory outcomes that result are not confined to shootings. Analysis from the Brookings Institution shows that in Minneapolis, black residents are stopped by police at a disproportionate rate relative to their population share — 5 times more than the average white person.
Discrimination in policing is also particularly present close to home, in Chicago, right in Northwestern’s backyard. A 2016 report from the Police Accountability Task Force for the city of Chicago, chaired by now current Mayor Lori Lightfoot, found that African Americans were stopped at a much higher rate by police relative to their population share in the city and that police actually found contraband on African American individuals at a far lower rate than Whites or Hispanics, indicating significant bias in stoppage.
B. Economic Disparities
Given a history of discrimination and unequal access to everything from education to hiring, various research has shown unequal economic outcomes among African Americans relative to whites and other minority groups. For one, the African American unemployment rate is consistently higher than any other ethnic group. This is true even when controlling for level of education.
Analysis from the Minneapolis Fed for example shows that African Americans with a BA or higher have similar levels of unemployment to Whites with only some college education (i.e. no degree) and that average annual wages of African American workers with a college degree is lower than their white counterparts. Analysis from the Economic Policy Institute shows this is true at various levels of education.
In a widely covered analysis using extremely granular, anonymized census and tax data, Raj Chetty, Nathaniel Hendren, Maggie Jones and Sonya Porter, in association with the Opportunity Insights project based out of Harvard University, studied differential financial outcomes of individuals from different racial and socioeconomic backgrounds.
The study found that black boys raised in wealthy families with similar family structures and levels of education ended up with worse economic outcomes later in life than their white counterparts in almost every county in America. The difference could not be explained by any heterogeneity in test scores or cognitive ability. The finding was true in reverse as well; African American boys who grew up poor were far less likely to climb up the income ladder than their white counterparts were, again controlling for cognitive ability, family structure, and other factors. In other words, the story of upward mobility and the American dream is a much more arduous one for African American men. These disparities were found for Hispanics and Native Americans as well but were far wider for Black boys.
Even though these same effects were not statistically observable among African American girls and women, the extreme levels of discrimination and inequality faced overall by African American women also calls for attention. One study found that of the 36 cent wage gap faced by African American women relative to white men, more than half is unexplained by factors related to age, geography, education, family, industry or occupation; most attribute this unexplained residual to racial discrimination. The study finds about a third of the wage gap is unexplained when comparing African American women to white women as well (African American women are paid about 81% of what a white woman would be paid).
C. Health Outcomes
In addition to adverse economic outcomes, African Americans suffer from adverse health outcomes in part due to limited access to health care. African Americans have lower life expectancy than whites and their discriminatory access to treatment for a variety of reasons has been well documented.
In the United States, a study conducted about the alarming racial differences in maternal mortality found that black women are 2 to 6 times more likely to die from complications of pregnancy than white women. Even though the authors are unable to explain the racial differences in maternal mortality rates, they hypothesize that the “quality of prenatal delivery and postpartum care, as well as interaction between health-seeking behaviors and satisfaction with care may explain part of this difference.”
Another study found that an algorithm used across several US hospitals designed to refer patients to receive appropriate care was less likely to refer African Americans who had the same chronic conditions as whites. Why? The algorithm in part correlated the predicted cost of health care spending with the severity of sickness. However in the data that was used to train the model, African Americans had to be sicker on average in order to get the same level of treatment as whites for the same illness, meaning they were ‘cheaper’ to treat overall.
More recently, as the Covid-19 pandemic has ravaged the US, African Americans have suffered disproportionately. The CDC recently released a report showing that hospitalization rates among African Americans was much higher than their relative population share. Clyde Yancy, a faculty member with the Northwestern University Feinberg School of Medicine, recently wrote about the disparities in health outcomes by race stemming from Covid-19. He discusses how the infection rate in predominantly black counties is 3 times as high as in predominantly white counties and the death rate is 6 times as high. Yancy writes that these differences are likely driven in part by social determinants of health such as socioeconomic status and access to care, in addition to co-morbidities.
Section 2 — Using data to find solutions
In evaluating and comparing various potential solutions to address these racial discrepancies, some public data sources help provide clarity to the policy debate.
Campaign Zero has compiled various 3rd party research reports, data, and analysis across ten broader policy categories aimed at decreasing violence in policing. Some related research on limiting use of force highlights eight policies utilized by police departments with fewer police killings per capita, even when controlling for factors such as number of arrests, average income and inequality, and share of minority population. Anecdotally, several cities have reported improved outcomes from community policing policies, and research indicates that positive contact with police around non-enforcement actions significantly improved attitude towards police especially among non-whites.
For further research, the FBI has launched a new database on police-involved shootings and use-of-force Incidents. While expected to be ready by this summer, recent news notes that fewer than half of law enforcement officers nationwide are submitting information to a database. There has also been renewed interest in Obama-era recommendations for police reform from the President’s 2015 Task Force on 21st Century Policing.
On economic policies, analysis from the Urban Institute outlines various initiatives over the last decade to increase economic opportunity for young men of color. The research identifies seven approaches and organizations that are particularly promising given initial data.
On health, the National Academies of Sciences, Engineering and Medicine performed an assessment on the drivers of differences in health care received by U.S. racial and ethnic minorities, providing a variety of potential interventions with a summary of the interventions here.
While no one solution will be enough, we encourage our fellow students to continue to use the knowledge gained within the four walls of Kellogg as well as our privilege and platform to drive meaningful (data-driven) conversations while being introspective and intentional about learning, engaging and supporting African Americans and communities of color.
