There's been a lot of chatter lately about the need for "honest" public discussion of race and law enforcement. Unfortunately, we've heard a lot more clamor for "honest" discourse than we've heard honest discourse. Perhaps a frank discussion of the complex relationship between rational judgment and racial prejudice can serve as a step in that direction.
Suppose the police receive a 911 call reporting that two men are attempting to force entry into a home. There are two obvious possibilities. First, the men may be lawfully gaining entry through a stuck door. Second, the men may be unlawfully breaking and entering. In approaching a home where such a burglary might be in progress, should it matter at all to the police officer whether the caller identified the men as black or white? Ideally, of course, the answer is "no." But is that rational?
In any such incident, the police officer must make a risk assessment. Most importantly, he must decide how likely it is that he is about to confront an on-going burglary. The greater the likelihood that a burglary is afoot, the more careful he must be. If he thinks a burglary is probable, he will be more inclined to call for back-up, prepare for possible violence, and have his adrenaline pumping than if he thinks a burglary is unlikely. For example, if the neighborhood has recently experienced a rash of burglaries, the police officer will take more serious precautions than if burglaries in the area have been scarce. This is rational behavior.
How does this relate to race? Blacks comprise 13% of the U.S. population but commit 30% of all burglaries. Non-blacks comprise 87% of the population and commit 70% of all burglaries. Thus, although 70% of all burglaries are committed by non-blacks, a black is statistically three times more likely to commit a burglary than a non-black.
It is important to note this says almost nothing about race, as such, but a lot about class. Both crime and race correlate with such factors as income and education. 33% of black families have annual incomes under $25,000, compared to only 15% of white families. Black children are twice as likely to drop-out of school as white children. Much of the disparity between black and white crime rates is explained by differences in family income and education, which in turn are largely the product of systemic racial discrimination, both past and present. This poses serious questions about our nation's social policies, but isn't of much help to the police officer who is approaching the house. He must deal with the situation at hand, regardless of the causes of that situation.
Let us return, then, to our police officer. If two men are seen attempting to force their way into a home, should a rational police officer take greater precautions if the men are black than if they are white? Sadly, the answer is "yes." This may seem shockingly racist, but it makes sense, given that we know that blacks are statistically three times more likely to commit a burglary than whites. Because race correlates in this way with burglary, the same ambiguous behavior engaged in by blacks is more suspicious than if it is engaged in by whites.
Assume, for example, that when the police receive a 911 call reporting that two men are trying to force their way into a home, there is a 50% chance that it is a burglary and a 50% chance that there is an innocent explanation. If the call reports two white men trying to force the door open, a statistician will tell us that there is a 44% chance that the men are committing a burglary, but if the call reports two black men trying to force entry there is a 70% chance that a burglary is in progress. This is a very significant difference in probabilities, and armed with this knowledge a rational police officer would exercise considerably more care if the suspects are black than if they are white.
To take this out of the race context for a moment, suppose the 911 caller identifies the two individuals by name and the police records show that the two men both have convictions for burglary. The police officer will be much more cautious in approaching the home in that situation than if the caller had identified the two men as Harvard professors. This is rational decision making.
So, when the police officer approaches the home where a burglary may be in progress, it is both rational and racist for him to think differently about the situation depending on whether the suspects are black or white. But how can the same response be both rational and racist? We tend to think of prejudice as irrational. But prejudice is not necessarily irrational. Often, groups about whom we hold beliefs we've deemed prejudiced do in fact have the characteristics on which those prejudiced beliefs are based. Blacks in our society are more likely to commit crimes than whites, women tend to be less strong than men, gays are more likely to have HIV than straights, persons of Irish descent are more likely to drink than Jews, individuals over 65 are less likely to be energetic workers than people under 40, and so on.
When we declare certain beliefs to be racist or prejudiced or sexist, we do not mean that they are necessarily irrational. We mean, rather, that it is unfair to act upon them. It is unfair for at least three reasons. First, we know from experience that in some circumstances we tend to exaggerate the importance of real differences. Rather than think carefully about the actual magnitude and significance of the differences, we blow them out of all proportion. A belief about people is racist or prejudiced or sexist when we give too much weight to the belief and therefore act irrationally on the basis of even real difference. Second, we usually condemn as prejudiced those beliefs that concern characteristics that are either largely immutable or that we believe individuals should not have to change in order to avoid discrimination, such as their religion. Third, we know that if we legitimate treating people differently on the basis of such characteristics, we will encourage thinking all these lines, thereby increasing the irrational use of these differences and fostering destructive divisions within society.
For these reasons, then, we generally prohibit employers from refusing to hire people because of their race, religion, gender, age, national origin, etc. We do this not because it is irrational for employers to consider these factors, but because we know that employers are likely to exaggerate the importance of these factors in order to make decision making easier, because we think it is unfair to deny an individual employment because of his race, religion, gender, age, or national origin, rather than because he himself has done something to justify not being hired, and because we know that such decisions, even if rational, will generate divisive and destabilizing feelings of unfairness and antagonism throughout society.
This brings me back, finally, to our police officer. Even if it is rational for the police officer to be more suspicious of two black men than two white men, do we want our law enforcement personnel actually to treat people differently because of their race? Should the police officer arrest the two black men (but not the two white men) in my earlier hypothetical because the probability of their guilt (70%) statistically satisfies the Fourth Amendment's "probable cause" standard for arrest? Should the police officer call for back-up or use handcuffs or draw his gun more readily when the suspects are black than when they are white because it is rational to do so? Is this any different from the employer who wants to disqualify all applicants of Irish descent because they are more inclined to drink than other applicants? What are the relevant similarities and differences between the police officer's situation and that of the employer?
I leave this as a question, without intending to suggest a "right" answer, because it is, in fact, a hard question. Perhaps it can provoke some "honest" discussion.
