For instance, did it ever occur to anyone that the abolition of the estate tax can't automatically be assumed to benefit the very rich testator? Guess why? Because he's dead! He can't take with him even a dispensation from the IRS.

The above may tempt you to scoff, but the figures are there. You become statistically very rich (top 1 percent) when, to round off the figure, you accumulate $300,000. When you or your wife writes a will, in the overwhelming number of cases you will name more than a single legatee. The fissiparous process begins, and of course everybody knows that the descendants of John D. Rockefeller are in the thousands.

But earthbound analysis is also worth attention, and it pays to force yourself to recall that families with two workers earn more than those with no workers. In 1999, median income among family heads who worked full time all year was $58,502. Among families with two earners, median income was $74,711. Now this development, which would appear innocuous in economic terms, has the effect of putting one-half of all two-earner families in the highest fifth of income distribution. When Mr. Gore goes about the country blasting the rich, he can be seen to be blasting one-half of two-earner families.

Much emphasis has recently been put, by critics of economic momentum, on the apparent stability of the lowest earners, the bottom one-fifth. It is said, as Mr. Reynolds points out, that the gaps between two-earner and zero-earner families have grown wider over time. But this is so because real wages climb. There are 2.2 workers per family among the top 20 percent earners. In the bottom fifth, you have a consolidated figure of 0.8. That means less than one earner working full time. If real wages rise, the two-earner families in the top fifth are mathematically certain to have faster income rises than families in the bottom fifth, where full-time workers are rare.

Now train your eyes on the percentage of families earning more than $75,000. In constant dollars, these have risen from 12 percent in 1980 to 18.4 percent in 1997. That means, correlatively, that to be counted in the top 5 percent now requires $l45,000 per year, where in 1988, $92,000 would elevate you there. The impression will be that the earnings are rising precipitately. But that is a statistical illusion; because any top-income group has no ceiling, a few huge incomes can also grossly exaggerate the average at the top.

What we suffer from is the absence of truly reliable practical information. A learned tax specialist said it all, ruminating on a conversation with an old friend, a classmate. "He lives in the Midwest and does cataract removals. Three hundred every -- I forget, every week, every month, every year. He nets $1 million." How is it that the market permits/tolerates that? How much administrative cost is necessary to our system of public health? How would the application of fluctuating deductibles affect the burden of medical costs?

These are intelligent concerns, and not much light is shed on them by rich/poor talk, let alone talk that suggests that the rich get that way by burdening the non-rich. Deep veins of thought and inquiry run through the economic system, but not much light is coming through in the high holy days of democratic practice.

Caspar Weinberger was appointed "director of finance" for the state of California by Ronald Reagan. A year later he was swept up by Richard Nixon. After that, he went to war as head of the Department of Defense. He has much to tell us.