A recent social media post by Will Solfiac has ignited discussion regarding the statistical methodology employed in reports analyzing racial disparities in deaths in custody. Solfiac criticized a report that presented figures of "86 whites, 23 blacks, 5 asians" for deaths in custody, arguing that comparing these numbers to overall population proportions is a fundamental misuse of statistics.
According to Solfiac, such a comparison leads to an erroneous conclusion that "the police are somehow killing people in custody because they're black." He emphasized the necessity of controlling for the actual proportions of each racial group "actually being IN custody" to draw accurate conclusions. Solfiac posited that the reason such a controlled study might not have been conducted is because it "doesn't show anything" supporting the initial, flawed conclusion.
Experts and official reports often highlight the complexities of analyzing racial disparities within the justice system. The National Institute of Justice (NIJ), in a comprehensive literature review on deaths in custody, notes that males and Black individuals are indeed overrepresented in custody deaths relative to their share of the U.S. residential population. Crucially, the NIJ report clarifies that "this overrepresentation reflects, to a large extent, the fact that males and Black individuals are disproportionately likely to be arrested, jailed, and imprisoned." This finding supports the methodological critique raised by Solfiac, underscoring the importance of using appropriate baselines for comparison.
However, the issue of racial disparity in custody is multi-faceted, extending beyond simple population comparisons. For instance, research from the Institute of Race Relations (IRR) in the UK indicates that Black individuals are seven times more likely than white individuals to die following restraint by police, suggesting that even with more refined data, significant disparities can emerge based on the circumstances of death. This highlights that while statistical methodology is critical, the underlying issues contributing to disparities can be complex and require varied analytical approaches.
The debate underscores the critical importance of rigorous and appropriate statistical analysis in understanding and addressing sensitive issues like racial disparities in the justice system. Ensuring that data are contextualized correctly, by comparing groups within relevant populations (e.g., those in custody versus the general population), is essential for informing public discourse and developing effective, evidence-based policies.
