The Ricci riddle and the law's limits

What you thought about the New Haven firefighters case was probably shaped less by logic or law than by your attitudes about the world.

Opinion

Can you break the law to avoid breaking the law? It sounds like a riddle, or a logical paradox, but in fact it is the question underlying yesterday's Supreme Court decision in Ricci v. DeStefano. And like other riddles, Ricci may have something deeper to teach us.

In 2003, the city of New Haven administered a test to select firefighters for officer positions. One hundred and eighteen firefighters, including 68 whites, 27 blacks, and 23 Hispanics took the test. The top 19 would be eligible for an immediate promotion. When the tests were scored, those 19 turned out to be 17 whites and 2 Hispanics. Overall, the pass rate for minorities was about one-half that of white candidates.

The racial difference in scores concerned city officials. It was enough of what's called a "disparate impact" that it looked as if the city might be vulnerable to a federal antidiscrimination lawsuit. Federal law prohibits employers from treating employees differently on the basis of race, and it also prohibits the use of apparently neutral employment tests on which scores vary significantly by race, unless they can be shown to be based on business necessity. City officials heard various arguments as to why the test they'd used might be unintentionally biased, and they heard their lawyers explain how the results could support a lawsuit for unlawful discrimination. Eventually, the city took a radical step. It threw out the results of the test entirely.

Unsurprisingly, the white firefighters who'd scored well sued. Now, they claimed, the city was discriminating against them. Two lower federal courts ruled in favor of the city, but yesterday the Supreme Court reversed those decisions. New Haven had broken the law by throwing out the test scores, the court said.

The vote in the Supreme Court was 5 to 4, which is one indication that the case was a hard one. Another is that even the majority admitted that it had to create a new standard to decide the case. What should happen when an employer discriminates in one way to avoid discriminating in another is a question that existing law simply did not answer.

If law runs out, perhaps logic can help? New Haven had a choice. It could accept the results of a test that had a disparate impact on minority applicants, or it could reject them. The two decisions are mirror images in terms of their consequences, one harming minorities and one harming whites. If one is unlawful, it would make sense that the other should be, too. But if both options violated the law, it would be an unendurable paradox. And so, one might think, the answer must be that both are lawful and New Haven could choose either.

But as Justice Oliver Wendell Holmes wrote, "The life of the law has not been logic; it has been experience." There is a case to be made that the two choices are equivalent, but few people would accept it. Instead, most find one much worse than the other. And which one they find worse tends to depend on what they think about the original racial disparity in scores. Those who think that the disparity was mostly due to differences in merit tend to think throwing out the scores was terribly unfair, while those who think the original test was biased think discarding the results was the right thing to do.

Strip away some of the legal window dressing, and these different views of the original test seem to be the main division between the majority and the dissent. It would be nice if we could bridge the divide, but it is probably impossible. There is evidence pointing in both directions. New Haven hired an independent company to design a fair test, the majority notes. But test preparation materials were expensive and may have been more available to white candidates, the dissent counters. What people make of the conflicting claims will depend largely on the perspective that informs their judgment.

That means that the way people think about Ricci – and this includes the justices – is in large part shaped not by logic or law but by their attitudes about the world. In particular, it depends on whether they think it is more likely that minority candidates were simply not as good as the whites, or more likely that there was some unintended bias skewing the results. What drives these attitudes, as Holmes knew, is experience. The facts of Ricci are an inkblot in which we all see the pictures life has drawn for us.

Ricci is not the only case in which law runs out, and in which judges will be guided by experience. When we ask whether empathy or life history are appropriate qualifications for judges, we should be thinking about these cases. We like to believe we write laws that judges can apply as if they were machines. We like to believe that about the Constitution, too. But in cases such as Ricci, the human element enters the law. Then we learn a little bit not just about how our judges think, but about who they are.

Kermit Roosevelt, who clerked for Supreme Court Justice David Souter, teaches law at the University of Pennsylvania's law school. He's the author of "The Myth of Judicial Activism."