(FORTUNE Magazine) – Like everybody else in the media, we are loath to defend Leona Helmsley, as it is considered naughty rather than nice to take business deductions to which you are not entitled, especially for girdles and Spanish marble dance floors. We are somewhat less loath to talk back to the media critics who have been making this point and have now officially had it with their sermonettes about taxes and the little people. Nexis recently produced on command 64 articles in which ''Leona'' appears within 30 words of ''little people.'' As you might expect, the articles gravitate rapidly to the evil epigram attributed to her at the trial by an erstwhile housekeeper, to wit: ''We don't pay taxes -- only the little people pay taxes.'' It is fairly obvious why that sentence has reverberated so loudly: A lot of people with bylines believe it to be true and have, accordingly, worked hard at soliciting quotes from certified midgets about the injustice of it all. Typical was the Newsweek cover story on Leona (sophomoric headline, next to her portrait: RHYMES WITH RICH), which quoted a spunky little widow as avowing that she did so pay taxes. Rummaging through the Nexis file, we note that the underlying validity of Leona's mot was attested to by Michael Waldman, a Naderite who runs Public Citizen's Congress Watch. Also by Populist Mark Shields, the Washington Post columnist and MacNeil-Lehrer commentator, who used Leona's maxim as his lead and then belted home the message: ''Middle- class Americans painfully know they are the ones who pay the taxes . . . Above them in wealth and clout are the powerful insiders . . . who play only by rules of their own making.'' Carried away more than anybody was James S. Henry, identified as an economist and lawyer, who lengthily informed readers of the Washington Post that ''Leona turns out to be absolutely right . . . The rich command a disproportionate share of tax evasion.'' Henry's evidence seemed risible, however. It consisted of IRS audit data showing that the unreported income sniffed out by the bloodhounds belonged mostly to high-income characters. (''Individuals among just the top 12% of taxpayers account for a striking 40% or more of individual tax underpayments.'') What it all comes down to is a rather unstunning proposition: that chiseling by the rich involves more money than chiseling by the poor. It does not, however, address the point at issue -- Leona's point -- which is whether the rich do more tax evading than other folks. What does anybody know about this matter? Searching for an answer, we turned up a number of recent studies evidencing awesome advances in methodological sophistication. An ''agenda for research'' recently prepared for the U.S. Treasury by scholars associated with the National Research Council elaborates lengthily on recent econometric models of taxpayer compliance. The median taxpayer, we gather, is perceived to be a kind of calculating machine with a built-in equation that balances probabilities of detection, severity of sanctions, marginal utility of money saved by cheating, and personal ethical attitudes toward the permissibility of cheating. But hard data on the prevalence of cheating are still relatively scarce. The best we found were figures worked up by Kent Smith, a sociologist based at the American Bar Foundation (a research group affiliated with the American Bar Association). Smith has been analyzing data from a detailed 1987 survey commissioned by the IRS. Major finding: One powerful variable in the cheating statistics is age. No matter how you break down the taxpayers -- by amount of income, type of income, education, whatever -- you keep finding that the rate of compliance rises with age. Income itself is not a powerful variable. It seems that there is no general relationship between level of income and propensity to cheat. But we must add a footnote to that statement. One of Smith's major accomplishments was the creation of an ''opportunities index'' measuring the odds of the IRS's detecting various kinds of underreported income. And when you focus only on those with the most opportunity to cheat -- e.g., people with significant cash or barter income -- income suddenly does become a significant variable. Within that universe, however, it is the low-income taxpayers who do the most underreporting. In short, Leona had the facts wrong. To be sure, she was not alone.