Study finds Black mortality proportional to Google searches of the 'n-word'
In the midst of rioting in Baltimore and heated nature of conversations regarding race, it is relevant to note a University of Maryland study published last week in the journal PLOS ONE. Researchers examined “levels of racism,” quantified by the proportion of Google searches containing the “n-word,” for whether it was a predictor of Black mortality.
Previous research pinpoints racism as a contributing factor to the mortality gap between people of color and white people, but these studies have relied heavily on patient interviews, in other words, asking people whether they have experienced racial discrimination. Self-reporting aside, racist behavior is not always explicit or overt enough to be noticed by a patient. As David Chae, assistant professor of epidemiology and lead author put it “Contemporary forms of racism are more subtle, and people may not recognize that the social insults they experience are driven by discrimination or prejudice.”
The thinking behind using geographic internet-search data is that higher instances of searching for the “n-word” may reflect more well-concealed instances of racism. The study does not assume that all searches containing that word are motivated by racism, instead it assumes that higher relative concentrations of the searches likely occur in places with higher instances of racism.
The results were significant: Black mortality rates in 196 different media markets, as collected by the National Center of Health Statistics, are 8.2% higher in places with a one-standard-deviation increase in area level racism- roughly equivalent to 30,000 additional deaths annually. The researchers took into account demographic and socioeconomic characteristics such as education and poverty in their calculation and still determined a significant correlation between high levels of area racism and Black mortality. Part of an attempt to control for variables included an adjustment to express Black mortality as a relative value to White mortality indicate the persistence and breadth of a gap in values.
Quantifying the impact of abstract pressures like racism is difficult in academia. Correlation does not equal causation and while the researchers in this case controlled for as many variables as they could, studies like this do need to be read with a certain grain of salt. In this case the numbers are reasonably strong and certainly warrant further investigation. Creative research that takes advantage of the wealth of data at our disposal can definitely begin to guide policies and resolve lingering issues such as discernible mortality gaps between races.
Image is from the study which can be read here.