(FORTUNE Magazine) – Is Bill Clinton a centrist and moderate, as loudly advertised at the famous clambake in Gotham? Or is he, like mate Hillary, a flaming liberal at heart? Carloads of printer's ink have already been squandered in addressing this question, so in sidling up to it once more, your servant needs a fresh angle. Luckily he has one. His analysis suggests but does not conclusively prove that deep down inside, Bill too is a flamer. We take it that Hillary's left-liberal credentials are not in dispute. In her days as chairman of the Legal Services Corp. (she was appointed by Jimmy Carter), the theoretically nonpolitical LSC scaled new heights of political activism. When she headed the New World Foundation, it funded numerous far- left organizations, including Cispes (Committee in Support of the People of El Salvador), the National Lawyers Guild, the avowedly radical Institute for Policy Studies, and the Christic Institute, which specializes in demonizing the CIA. For a good summary of her agenda, see ''The Lady Macbeth of Little Rock,'' by Daniel Wattenberg in the August American Spectator.

What does all this tell us about Bill? Nothing conclusive, of course, and yet -- something. We begin with certain fundamentals about husbands and wives. It is well established that, on many different measures, American spouses are generally similar to one another. It is sometimes argued that ''opposites attract,'' but the available data show that people generally end up with mates like themselves. The correlations are positive, not negative. On a scale in which positive correlations range between 0 and 1, spouses correlate about .15 on scales that measure personality. They correlate about .30 for height and about .45 for IQ. On measures of physical attractiveness (based partly on self-reported ratings, partly on assessments of photographs by disinterested observers), the correlation coefficient goes up to .50 or thereabouts. This brings us to a fascinating detail. Of all the measurable traits for which there are data on spousal similarity, the one with the highest correlation is (gasp!) political attitudes. Professor David T. Lykken of the University of Minnesota psychology department, who helpfully held our hand as this item was being prepared, cites a number of estimates for such correlations in the zone between .40 and .70; Lykken himself believes that the best estimate is around .60. Political attitudes can be gauged in different ways, but the data cited here reflect responses to several subtests in the Multidimensional Personality Questionnaire. The MPQ was developed at Minnesota and has been used extensively in the testing of identical twins. A major, astonishing finding of the Minnesota Study of Twins Reared Apart is that political attitudes have a genetic basis. How can anybody know this? Imagine a spectrum of political attitudes. At one end are people eager for change and experimentation, ever ready to kick over the traces; at the other end are people committed to traditional values and established authority. It turns out that the separated identical twins generally end up at about the same point on this spectrum. Having been separated soon after birth and put up for adoption, the twins have dissimilar environments. Clearly, their similar political attitudes are being driven by their identical genes. Spouses also tend to end up at the same point on the political spectrum, and the correlation for their attitudes is as strong as that of the twins. The mechanism by which spouses get to that correlation of .60 is not entirely clear. But it is not because years of living together have made them politically alike: The Minnesota studies show that the political similarities of husbands and wives are basically in place from the beginning of the marriage. The studies have a stunning implication, possibly of even greater significance than Bill Clinton's correlation with Hillary. If spouses were indifferent to one another's politics, then their offspring would have no discernible political tendency and as a group would gravitate to the middle of the political spectrum. But the fact that spouses do resemble one another politically, combined with the fact that political attitudes are affected by genes, means that the extremes of left and right are being endlessly renewed in each generation. Lykken says he does not know what evolutionary purpose is served by this process, and we cannot figure out whether it mainly helps Bush or Clinton.