Tuesday, December 16, 2014

Researching race, crime and policing: Controlling variables into meaninglessness

By Isabel Manuela Estrada Portales, Ph.D., M.S.

Science and research require methodology. A pillar of science is the ability to control for extraneous and confounding factors that will render the results of an experiment questionable, or, at least, will significantly impair the ability to make sound pronouncements on the subject at hand.
Social sciences are no different. Hypotheses are tested, peripheral factors are isolated, that is, “controlled for,” and we arrived at conclusions. Obviously, the problem with social and behavioral sciences is that the pesky “object” of study happens to be a “subject.” Humans always complicate everything.
But the other humans in the scientific process are the scientists themselves. They are, despite the stereotypes, just as subjected to cultural conditioning as the lesser amongst us. Michael Foucault did plenty already to put through the ringer the presumptions of taintless objectivity of the research processes even in the hard sciences.
What can we expect when the subject of research is something as controversial, as downright divisive as race and discrimination, particularly in the United States? What can we expect if we add the ingredients of crime, discriminatory policing, biased law enforcement?
It requires significant intellectual and scientific honesty to realize our own limitations and acknowledge the very high degree by which our research is colored – pardon the pun – by our historical and social lenses. It also demands a lot more work and effort to build a parallel bounty of resources and a theoretical arsenal to ask new questions, to re-hypothesize, to question the origins of not merely our conceptions and frameworks, but even the wording and, certainly, the “objective data” – crime reports and statistics, for instance - that underpins our sociological and cultural research.
In the case of racial discrimination in the various realms of society, the variables, particularly those variables researchers “control for,” actually tell the story. Controlling variables supposedly makes your research more precise, but, as Ezra Klein discussed, “you can control for too much. Sometimes you end up controlling for the thing you’re trying to measure.”
It may sound contradictory, but it makes perfect sense, because, for instance, in the case of racial exclusion and the economic plight of blacks in America, when you try to control for the “lack of father in the household,” “the lower educational status,” and the like, you are forgetting the key piece: How come those are problems in this particular community? What are the historical bases for it?
Let’s examine the issue of crime statistics. There is nothing more “precise” and black and white – yes, another vulgar but irresistible pun – than numbers of crimes committed. Or is there? Because what we have learned by asking other questions is:

  • The crime numbers – not the actual crimes – go up in more heavily policed areas. Therefore, if there is a policeman with a quota to fill and a training that sends him to “prevent” crime in a black neighborhood, one can bet he will catch a lot more crime. The question is: would he have caught the same number of crimes if he were policing a white middle class area after school or a college on a weekend?
  • If we prosecute more aggressively minor crimes committed by black offenders, does that mean they are more prone to criminality?
  •  How come the use of certain drugs is more prevalent in affluent white communities, but the people caught for them are predominantly blacks?
  • Not to belabor this point, but then, if I am studying black criminality based on the exacting figures of the crime reports or the prosecutions, how good is the sample I’m throwing in that glass tube to analyze? Is there even a moral question in terms of studying “black criminality” without any examination of the constraints and conditioning of the terms?
As Klein puts it:
The problem with controls is that it's often hard to tell the difference between a variable that's obscuring the thing you're studying and a variable that is the thing you're studying.
There is plenty of conflicting data and evidence regarding discrimination in law enforcement, and racially related criminality.
In his nicely summarized review of studies on the subject, Scott Alexander states:
There are of course many other forms of police stop. These tend to follow the same pattern as traffic stops – strong data that police more often stop black people, police making the claim that black people do more things that trigger their suspicion instinct (including live in higher-crime neighborhoods), and difficulty figuring out whether this is true or false. (Emphasis is mine.)
And this begs the question about the questions we ask when researching. Do we really think black people choose to live in higher-crime neighborhoods? That is, we believe they all have the chance to live on the right side of the tracks and send their kids to the good schools, but they choose to go to the other side and dodge bullets. No economic causation. No history. They just live in higher-crime neighborhoods. I just wonder how you even control for that in your research without rendering it not merely meaningless but laughable, inching towards insulting.
Yet again, though, “high-crime neighborhood” is another verbal and social construct. This seems like the serpent that bites its tail, but therein lays the problem.
Yes, it may seem quite simple, quite clear to see what a “high-crime neighborhood” is. Just look at the numbers! Oh, yes, sure. Let me look at the numbers. And I see a lot of policing of minor offenses transforming kids into criminals for one bad decision. I also see a different set of numbers. Those of my neighborhood. Those of the kids who do stupid things along with my own kids. Those kids with good family lawyers and with good police officers who have no intention of messing up the lives of “promising” kids for dealing and consuming drugs, having drinking parties, skipping school, driving while drunk without injuries beyond property, and a very long etcetera. I also know those kids are definitely not poor, and MOST DEFINITELY not black. I know the parents: doctors, lawyers, think-tankers, researchers, government workers. We have commiserated about our children’s stupidity. We have even remembered our own youthful indiscretions – “oh, you know, things were so different then!”
As I have learned while navigating high school, I live in a VERY high-crime neighborhood…but you would never know it, because the numbers that are not recorded don’t tell a story. It is the safest neighborhood ever. It has the best schools and plenty of money. It also has kids that are too bored with all the niceties of life to not engage in the riskiest and most criminal of behaviors. It has police officers that come before you finish dialing their number…but are never there over policing or “preventing,” hence, they are never there “recording crimes” that we have not deemed appropriate.
Our kids can make mistakes. Black kids (particularly poor black kids) cannot.

MUST – not merely Good – Reads
How racial discrimination in law enforcement actually works

Race and Justice: Much more than you wanted to know

Post a Comment