Walter E. Williams – June 17, 2017
Profiling is needlessly a misunderstood concept. What’s called profiling is part of the optimal stock of human behavior and something we all do. Let’s begin by describing behavior that might come under the heading of profiling.
Prior to making decisions, people seek to gain information. To obtain information is costly, requiring the expenditure of time and/or money. Therefore, people seek to find ways to economize on information costs. Let’s try simple examples.
You are a manager of a furniture moving company and seek to hire 10 people to load and unload furniture onto and off trucks. Twenty people show up for the job, and they all appear to be equal except by sex. Ten are men, and 10 are women. Whom would you hire? You might give them all tests to determine how much weight they could carry under various conditions, such as inclines and declines, and the speed at which they could carry. To conduct such tests might be costly. Such costs could be avoided through profiling — that is, using an easily observable physical attribute, such as a person’s sex, as a proxy for unobserved attributes, such as endurance and strength. Though sex is not a perfect predictor of strength and endurance, it’s pretty reliable.
Imagine that you’re a chief of police. There has been a rash of auto break-ins by which electronic equipment has been stolen. You’re trying to capture the culprits. Would you have your officers stake out and investigate residents of senior citizen homes? What about spending resources investigating men and women 50 years of age or older? I’m guessing there would be greater success capturing the culprits by focusing police resources on younger people — and particularly young men. The reason is that breaking in to autos is mostly a young man’s game. Should charges be brought against you because, as police chief, you used the physical attributes of age and sex as a crime tool? Would it be fair for people to accuse you of playing favorites by not using investigative resources on seniors and middle-aged adults of either sex even though there is a non-zero chance that they are among the culprits?
