Racism and Bad Statistics

I once spoke of my hope for a dystopian utopia, where all races and ethnicities bonded together against “the Man”. Sadly, today we’re possibly even more divided and contentious than before, with each side blaming the other for our societal issues. Even people who would normally be on the fence feel the need to take sides, often without (or even in the face of) critical thought and education.

One of the most frustrating behaviors is the presentation of statistics to prove our point… that “our” beliefs are more valid than “theirs”. Of course, arm chair experts spouting data is about as reliable and valid as someone using WebMD to diagnose their problems. (I know… my hypochondriac behind regularly does the latter). So, here I am to point out a couple of statistical claims and how they’re being used poorly to fit one’s personal perceptions or agenda.

A) More Whites are killed by Police than Blacks

Numerically, yes. According to a year-long study by the Washington Post, 990 people were killed by police officers in 2015. Of those, 494 were White and 258 were Black, confirming the above error… but only superficially. Most people with a knowledge of statistics in social sciences know you usually don’t look at raw population, but per capita numbers.

In America, the population is 63.7% White and 12.2% Black, so it would make sense mathematically for more Whites to be involved in a police-related incident than Blacks. Yet 50% of those killed were White while 26% were Black, which doesn’t match the population demographics. Statistically, more of the Black population were killed than White, showing a disparity between population and police-involved killings.

So yes, by raw numbers, more Whites are killed by police than Blacks. That makes sense considering our population consists of more White citizens than Blacks. It would be like going to Hawai`i and saying, “Man, I see more Asians and Polynesians here than elsewhere.” Yet, from a per capita perspective, Blacks are killed more often than Whites, and there’s no denying that.

B) More Police Officers are killed by Blacks than Whites

The latest now tries to flip the script on that per capita perspective. New arguments show that Blacks per capita kill a larger amount of police officers than their White counterparts. The 2004-2013 FBI statistics confirm this, with 51% of officer fatalities at the hands of Whites compared to 43% by Blacks. That seems to support the above allegations, but once more knowing numbers and applying them are two different things.

Most sociologists or psychologists know not to generalize, but instead to use the numbers as a guide to determine underlying factors. They wouldn’t use the data above to generalize that police are more likely to be killed by Blacks, but instead to ask why there’s a disparity between population and police fatalities involving Black individuals.

Could it be that minority communities are often relegated to poorer, more crime-ridden areas thanks to decades of segregationist practices, meaning interactions there are potentially more dangerous? Could it be that a millennia of oppression has lead to less trust and more miscommunication between minorities and law enforcement, creating the potential for violent interactions? Could it be the unethical and unjust practices of a poor justice system and for-profit, revolving door prisons that put known offenders back on the street and poison neighborhoods? These are answers that would require more complex studies than the simplistic and fallacious black-and-white thinking of most who simply spout numbers to fulfill their personal biases.

The point is, don’t spout numbers unless you know how to properly interpret and apply them.

When a scientist performs a statistical analysis, they often include conclusions that discuss the possible implications as well as other factors that should be explored. They try not to generalize or misinterpret, and often ask more questions upon discovering data than they do making final conclusions. Numbers are used to ask why, not to declare causal or other fallacious relationships.

This is doubly so when exploring race, ethnicity, and other demographics. It’s false use of statistics that has contributed to discrimination throughout history, from the pseudo-science of phrenology to studies that denigrated LGBT parents. When someone spouts those erroneous numbers, especially among like-minded individuals, they’re guilty of confirmation bias and being in an echo chamber. Worse, they often contribute to the same ignorant or racist behaviors that pervade society by perpetuating the underlying beliefs and perceptions.

As with all things involving critical thought, think before you speak. Numbers don’t always tell the whole picture, especially when you don’t fully comprehend them.