There is no evidence for this notion -- and tons of evidence against it, from countries around the world.

 American men are struck by lightning six times as often as American women. Who is discriminating? Men are just 54 percent of the labor force but they suffer more than 90 percent of all deaths on the job. Discrimination?

 Is it discrimination against whites when Asian Americans have their applications for mortgage loans approved a higher percentage of the time than whites do, just as whites are approved a higher percentage of the time than blacks are?

 Discrimination has joined a long list of charges, including sexual harassment and child molestation, in which those who are accused are expected to try to prove their innocence. When it is impossible to prove a negative, the accused loses -- or else settles out of court, in effect paying legalized extortion, to avoid dragging out the bad publicity.

 Recently Gerald Amirault was released from prison in Massachusetts after spending 18 years behind bars on child molestation charges that today virtually no one believes. Those who have examined the evidence -- whether lawyers, laymen or judges -- have expressed amazement that such stuff had been taken seriously as evidence in a court of law.

 That is what happens when people start with a preconception and seize upon anything that looks consistent with it. Statistical disparities have been a major source of such fallacies.

 Back in the 19th century, Dr. Marcus Whitman, for whom Whitman College is named, worked on the western frontier and treated both whites and American Indians who had been stricken with measles. The whites recovered and the Indians died.

 This statistical disparity caught the attention of local Indians, who massacred Dr. Whitman and other whites. But the reason for the disparity was quite simple: Measles had never existed in the Western Hemisphere before Europeans brought it here, so the indigenous people had no biological resistance.

 One of the reasons given by Earl Warren for supporting the internment of Japanese Americans during World War II was that they lived clustered around military bases to an extent that greatly exceeded what could be accounted for by random chance.

 If you start with the preconception that Japanese Americans were likely to try to sabotage the American war effort against Japan and then add a statistical anomaly, you are following the same procedure that leads in many other situations to the grand fallacy that preconception plus numbers equals proof.

 In reality, the Japanese Americans, who were largely farmers in those days, lived where they did for the same reason that the military built bases there: The land was cheap. In fact, the Japanese Americans were there first and the military then came in and built bases in their midst.

 It is so easy to go so wrong when numbers are added to preconceptions.

 The fundamental problem is that our legal system allows one side to impose huge costs on the other, at little cost to themselves. When a false charge of discrimination can force the accused to mobilize teams of high-priced lawyers, but making that false charge brings no penalty to the accuser, this is virtually a guarantee of a flourishing industry of legalized extortion.

 If I can spend $10,000 and impose a million dollars worth of costs on you, then the law is in effect enabling me to extort hundreds of thousands of dollars from you to go away.

 Wal-Mart has in the past resisted such lawsuits. Given the huge size of this one and the huge public relations hit that Wal-Mart could take if they go to trial, they may be advised to settle out of court. Let's hope they resist any such advice, so as not to encourage more legalized extortion throughout American society.