(FORTUNE Magazine) – THE JOKE FROM HELL

A wisenheimer whose name we cannot recall once observed of sexual harassment that it was a crime so horrible even innocence was no defense. Coming off the recent experience of O.J. mouthpiece Bob Shapiro, one wonders if the sentiment might also be applicable to ethnic jokes.

The Simpson team, you will recall, had some good days in court a while back at the expense of an overmatched prosecution witness named Dennis Fung. Afterward, Bob gave some reporters fortune cookies, laughingly adding that they had come from the (nonexistent) Hang Fung Restaurant. This flight of fancy resulted in protests from the Organization of Chinese Americans and the Southern California Chinese Lawyers Association, before both of which Shapiro has now extensively groveled. Meanwhile, he has also been hit with a blizzard of news stories and/or editorial denunciations in the New York Times, the Chicago Tribune, the Rocky Mountain News, the Louisville Courier-Journal, and lots of other places, all assuming a mile a minute that something godawful had happened. If anybody in the whole country stood up and said "What's the problem?" we haven't heard about it.

What exactly is the problem? Shapiro did not trot out any ethnic stereotypes, or verbal slurs, or in any way manifest ill will to Asian Americans. He just made a silly little joke. And showed that the culture is still testing the outer limits of political correctness.

ASK MR. STATISTICS

Dear Odds Fellow: As the longtime owner and personal companion of a pit bull who is extremely interested in letter carriers, the undersigned cannot help but notice that 12 fortnights have elapsed since you last lashed out at the U.S. Postal Service, and he wonders if you are going soft on those guys or have possibly been turned around by recent claims of performance breakthroughs, as adumbrated in the June 13 press release issued by Marvin Runyon under the strangely commaless headline service breaks all records announces postmaster general. Demonstrating anew that he is not in Damon's class as a sports commentator, Marvin proceeds in this broadside to breathe heavily over stats recording recent overnight delivery of first-class mail at an 87% rate nationwide, and along the way exults: "We knew going into this [March-May] quarter that we had the resources and systems to hit a grand slam. Our employees hit the ball out of the park. I am proud of what they've achieved." Do I have to worry that prose like that might lead you to buy the postalite party line at a time when the canine companion and I get hardly any letters at all for some reason?

AGAINST SNAILMAILERY

Dear Pitperson: You and your obviously only friend will be thrilled to hear that Marvin's emetic recitation has triggered a high-powered statistical analysis of slugging percentages at one's own post office here on the Upper East Side of Manhattan. Our database is a minispreadsheet showing the dates and times of day on which Mr. Statistics mailed a first-class letter to himself, always at a letterbox within a block of his own residential tower and within three blocks of the Lenox Hill Post Office. All the pickups should have taken place no later than 3 p.m., and all should have been treated as NDD (next-day delivery) missives. The press release hailing the new 87% NDD record nationwide also indicated that things were a bit less roseate here in the Big Apple, with an 83% record. But in our own experiment, the local team failed to get anyplace near even that figure. Of the letters mailed in the ten postal days (i.e., excluding Sundays) between the press release and the deadline for this item, only four were delivered to Mr. Statistics' apartment overnight. Our spreadsheet and the letters are available for inspection by Price Waterhouse, the source of the USPS grand slam figures.

Possibly you are wondering if the science of statistics can do anything at all with so small a sample size--only ten letters. The answer is, yes, it can. We begin with the claim that here in Gotham, 83% of within-city mail is delivered overnight. The figure for within-zip-code mail like ours should be higher, but let us settle for 83%, i.e., for the statement that 17% of mail that is supposed to get delivered overnight takes longer. So the question is: What is the likelihood that an event generally occurring 17% of the time will, by chance, occur as often as six times in an experiment featuring ten trials?

Here we pause to rhapsodize about @BINOMIAL, a statistical function that comes with one's newly acquired Lotus 1-2-3 Release 5 for Windows. (Comparable functions are available in recent versions of Excel and other spreadsheets.) Wanna know the probability of exactly 5,000 heads in 10,000 tosses of a coin? You just type @binomial (10000,5000,.5) and instantly see the answer displayed: 0.79%. The answer for at least 5,000 heads: 50.40%.

Answer supplied by @BINOMIAL to the postal question posed above: 0.27%. That figure can be taken as the likelihood that one has been extremely unlucky in the postal delivery dimension of life during the past ten days, the 99.73% alternative being that one's own service is basically no good no matter what it says in those press releases.

Dear Statmeister: Just to annoy my parents, who recently spent $52,000 putting me through business school, I have decided to live a life of genteel poverty. I was patriotically assuming until June 25 that I would do it here in the U.S.A., but I see by an article in that day's Washington Post that other countries are nicer to poor people. Headlined america's tide: lifting the yachts, swamping the rowboats, the article posits that antiegalitarian practices here in the homeland leave folks in the tenth percentile of the income distribution worse off than they would be abroad. Specifically, we Yanks at that level have less after-tax income than do denizens of all except three countries (Italy, Britain, and the Netherlands) among 13 Western countries surveyed. At the tenth percentile in Switzerland, the goatherds or whoever they are actually make 64% more than do their Americanski percentile counterparts. Is expatriation the answer?

NAME WITHHELD, MBA

Dear Annoyance: The article causing you angst may be thought of as the third round in a debate that began with a recent New York Times front-pager worriedly describing the U.S. as "the most economically stratified of industrial nations." In Round 2, George Will and other commentators responded by arguing that stratification is not so terrible, that inequality is associated with higher growth rates, and that the more egalitarian societies pay a heavy price in the form of diminished incentives and lower incomes overall. The Post article, by economists Gary Burtless and Timothy Smeeding, responds by saying, yeah, that's all very well if you're affluent like Will, but if you are way down there in the tenth-percentile dumps, you are still better off in Oslo than Oshkosh. Accompanying the article are index numbers representing each of the 13 countries' after-tax family incomes at the tenth percentile, the median level, and the 90th percentile, with a fourth column that gauges inequality in each country by dividing tenth-percentile income into 90th-percentile income. In the U.S., for example, income at the tenth percentile is 100, and at the 90th percentile it is 594, so the basic measure of inequality (highest in the group) is 5.94. The comparable "income ratio" for Finland (lowest in the group) is 2.59.

This brings us to Round 4, wherein Mr. Statistics re-analyzes the four columns and testifies as follows: (1) It is true, as noted in the Post, that all other things being equal, poor people are better off in egalitarian societies. The correlation between tenth-percentile incomes and those "income ratios" is distinctly negative at -0.34, meaning that incomes tend to decline as the measure of inequality rises. (2) It is also true, as Will argued, that all things do not remain equal as you move up or down the egalitarianism scale. The stratified U.S. has the highest median income of any country on the list, and the correlation between median incomes and the "income ratio" is distinctly positive at 0.45-meaning that incomes tend to rise as countries become more unequal. (3) In the 13 countries as a whole, rising tides lift rowboats. Correlation between median and tenth-percentile incomes: a powerful 0.67. The U.S. exception to the rowboat rule appears not to reflect antiegalitarian policies but the fact that poor American families are far more likely than those in the other countries to be headed by females on public assistance and living outside the market economy. One has no data on MBA poverty.

THE SCRUTINIZERS

In Adarand Constructors v. Pena, the Supreme Court's latest failed effort to say something clear about affirmative action, the word "scrutiny" appears 89 times. The burden of the decision is that race-based set-asides for government contractors must undergo "strict" scrutiny, and not just "intermediate" scrutiny, and certainly not "relaxed" or "minimum" scrutiny. The decision will clearly make it harder to defend federal race-based preference programs. It will not, however, abolish them; worst of all, it leaves us still clawing at the air trying to figure out what's legitimate and what isn't.

A program subject to strict scrutiny is one that cannot pass muster under the Constitution's "equal protection" mandate unless there is a "compelling government interest" in its objectives. In contrast, intermediate scrutiny requires only that there be a "substantial" interest in the objectives, and minimum scrutiny requires only a "reasonable" relationship to a "legitimate" government interest. It sounds tidy enough at first encounter, but what one learns from the court's recurrent visits to the subject is that those terms are wonderfully elastic. In the judicial metaphor, there is no "bright line" separating compelling and substantial interests. Some justices (but not a majority) have argued in recent years that the federal government has a compelling interest in preference programs designed to increase role models for minority kids. In the 1978 Bakke decision, on preferences in medical school admissions, Justice Powell (but nobody else) said the future diversity of the student body was a compelling interest. In deciding that set-asides harmful to white-owned contractors require strict scrutiny, the Supremes in Adarand were reversing at least one prior decision of their own.

What is most maddening about Adarand is the missed opportunity to create a bright-line rule. Brother Scalia, who endorsed the majority decision while manifesting unhappiness with its logic, said he would have preferred a simple distinction: The Constitution cannot conceivably view government as having a compelling interest in laws specifying race-based group preferences. It allows only recompense to identifiable individuals wronged by past discrimination.

The majority opinion, written by Sister O'Connor, went on for pages noting distinctions between her views and those of the dissenters. But it warily avoided saying whether she agreed with Scalia, and if not, why not. Unless Congress does something decisive to end affirmative action--which is, to be sure, looking possible--you can count on decades more of hairsplitting about what's compelling, what's substantial, and what's reasonable. It ain't reasonable.