Racial Bias Can Exist Without Explicit Racial Animus

The prosecutor in Arlington, VA has announced an initiative to quantify and (potentially) reduce racial disparities in the criminal justice system. According to what they already know, although only 10% of the county’s population is black, 50% of those arrested and 48% of those charged are black.

Arlington’s previous prosecutor strongly criticized this “race-aware” initiative:

Theo Stamos, who served as Arlington’s chief prosecutor until losing a primary to Dehghani-Tafti in 2019, said she never looked at race in her decisions. Doing so, she said, “is an odious practice that makes a mockery of blind justice and corrodes confidence in the criminal justice system.”

The idea that racial disparity are necessarily conclusive proof of racism is a baseless truism made popular by the indefatigable efforts of Ibram X. Kendi and the like. It’s certainly possible that disparities in arrests and convictions among groups could simply be the result of differences in criminal propensity. After all, I am aware of virtually no one claiming that the reason males constitute ~90% of the incarcerated population is because of institutional misandry in the criminal justice system.

So I think it’s perfectly fair to avoid using disparity=racism as a heuristic, because it’s a naive and overly simplistic mental model. But at the same time, knowing nothing about how Arlington county’s criminal justice system operates, I'd like to try and steelman the prosecutor’s initiative.

Prosecutors cannot file any charges unless the police notify them of a violation, and the police can't know about this unless they are summoned or by civilians, or they happen to observe it.

At any one of those steps, there could be a systemic bias that could tilt enforcement in one direction over the other. It's possible that civilians are more likely to call the police on particular races and demographics. It's possible that some areas are much more likely to be patrolled by police. It's possible that cops are much less likely to let someone go with a warning based on their race. All those combined could potentially lead to a shift in terms of what gets charged.

I'm broadly familiar with how charging decisions are made, and I'm inclined to also echo the comment about prosecutors never looking at race in their decisions. With the prosecutor's office I deal with, "charging" is a task that attorneys get rotated through, and it's generally seen as grunt work. They sit by themselves just churning through dozens of police reports submitted, and they make a determination based on whether a crime has been committed, whether it makes sense to charge it, and whether there is enough evidence to prove it. Sometimes they decline charges because of blatantly illegal conduct by the police (such as no probable cause for a stop) and they send notes to the cops to advise them to avoid that mistake in the future.

Because of the sheer volume of these cases, the prosecutors basically never view video or anything else, they go off only the report. But even with such a relatively “blind” approach, you can still surmise the race of the individuals involved. Names are an obvious tell, but even something like which particular drug substance (guess which race loves using meth?) can tilt things in one direction over the other.

Not all of this has to be borne out of invisible and deep-seated bigotries. When I said “systemic bias”, this could be purely based on biases which are not at all racially motivated. An example situation is to consider marijuana arrests. Some studies have found that marijuana use is identical between white and black populations (I'm not getting into the methodology of this, because it's not relevant to my example), but black people are far more likely to be arrested/charged/convicted of marijuana possession. A potential explanation for this is that ceteris paribus white people are richer, more likely to own their home, and their home is larger, so there are more 'private' places to smoke the devil's lettuce. White people would also be more likely to live in suburban areas, which won't be patrolled as heavily by squad cars. All those can cause a discrepancy in drug possession prosecution, but none of which is explicitly motivated by bias.

I'm aware of the arguments that try and assert there is a genetic or cultural component with regards to proclivity towards crime. I'm not addressing those. I just want to remind everyone that we don't live with a criminal justice system whose machineries are completely detached from personal bias and animosity, or otherwise operated by an AI examining a blank canvas. Bias can exist, and sometimes from completely unbiased reasons.