In recent weeks, I've offered operational definitions for some of the terms used in the discussion of race. The first was discrimination, which can be broadly defined as the act of choice. When one selects one activity, good or person, of necessity he must discriminate against an alternative activity, good or person.
The second term was prejudice, which can be seen as people making decisions on the basis of incomplete information. We could call it pre-judging or stereotyping. Information is costly. To gain more information requires the sacrifice of resources, be they time, money and perhaps one's life, so people seek information shortcuts.
Imagine an employer plans to hire 20 strong people to manually unload a ship. Fifty people show up for the job, and they all appear equal, except by sex. The employer has zero information about any other attribute, and he would like to hire the physically strongest people in the group. How might he select employees?
I'm guessing the average employer's first approximation would be to select the men in the group. He does so because he surmises that sex is highly correlated with physical strength. Of course, some of the women in the group could be just as strong, or stronger, than the men, but the employer is playing the odds.
This example produces an important observation. By observing the employer's behavior, can one conclude that he doesn't like women? The fact of the matter is that by observing his hiring choices, there is nothing unambiguous one can say about his preferences. To identify a behavior as preference indulgence when it's really an effort to economize on information costs is to misunderstand the behavior.
Some might argue that the employer should seek additional information before making a choice; however, expending the additional resources might not be worth it to him. That's similar to decisions one makes when shopping. One doesn't acquire all the price information possible when he's shopping for, say, shoes. At some point, he concludes that further searching isn't worth the additional resources of time and money, even though he guesses that somewhere there might be a lower price.
What's a woman to do who has the physical strength to perform just as well unloading the ship as a man? One strategy is to provide what the employer views as reliable information about her strength. Another alternative is to offer her services at a lower wage. There's no better way to get people to experiment, and perhaps revise their expectations, than by lowering prices. That's why a new, previously unknown, supermarket, restaurant or other establishment might use sales to attract customers.
What would be some impediments to getting an employer to experiment and take risks that might ultimately revise his expectations? One would be legislation requiring the employer to pay everyone the same wage. Another would be legislation making it costly to fire a lousy worker. After all, if the employer's hunch didn't work out, he would have to bear the costs of discrimination suits, and possibly costly settlements, to get rid of the employee.
A few readers, in response to my discrimination discussion, said that my argument justifies the racial segregation of the past. To explain phenomena is not the same as justifying phenomena. You could fall off the roof of a tall building. I explain that your death is a result of the forces of gravity that caused you to accelerate at 32 feet per second and the sudden deceleration when you hit the ground. Wouldn't it be silly if someone accused me of trying to justify gravity and your death?