By Michael M. Thomas

February 18, 2001 | 7:00 p.m

Not long ago, three of us-men of a certain age and

background-found ourselves talking about the estate tax, which the Bush administration proposes to phase out. We agreed that elimination of the death tax would benefit us, or people close to us, to a significant degree. We also agreed to a man, with conviction and feeling, that eliminating the estate tax is a bad idea, one to which we were completely opposed. Why? Because the estate tax is probably more important for what it says than for what it collects. It is, if I may once again drag out the tired axiom, the principle of the thing that matters. This is probably true of taxation in general, and as taxation is what the administration has made the centerpiece of its domestic politics-a tax cut that is, as opposed to tax reform -I thought it might be amusing to brainstorm, in the hope (if not expectation) of stirring up intelligent readerly reaction, what one might come up with if asked to design a brand-new, equitable Y2K+1 tax system for these United States. I was very impressed with the ideas articulated by former Treasury Secretary Robert Rubin in his New York Times Op-Ed piece this past Sunday, Feb. 11. Clearly, the tax plan the Bush people have placed on the table isn't very well thought out. In fact, it's probably worse than that, as evidenced by the quality of the people schooling like piranhas to support the tax cut: lobbyists who wouldn't know the bottom bracket if it bit them in their fat behinds. The company one attracts is as infallible a benchmark of policy as it is of character. In conception, the Bush people are saying the right things-money in the hands of the people, etc.-but in execution, I fear we could be looking at the Marc Rich of tax legislation. In a well-conceived plan, the bulk of relief should go to the bottom half of the tax table. By relief, I mean a reduction in cash outflow coupled with a reduction in the degree to which one is screwed relative to those in higher brackets. The quid pro quo at the top is to tax as rich what really-by today's norms- is rich. In a perfect world, one could find a way to differentiate, taxwise, between the deserving rich and the undeserving. For an example of the latter, I would submit class-action lawyers who have carved out billion-dollar fees in settlements in which Uncle Sam, whose salary is paid by We the Taxpayers, has wielded the big stick and done the heavy lifting. And, of course, there's always Donald Trump. Obviously, the simpler, the better. The easier a tax is to calculate, the easier it is to collect. That means elimination of loopholes, probably even including a phase-out of mortgage interest. To take advantage of most tax loopholes requires capital, which is something most American families don't have, apart from their homes. If you've got the collateral, you can borrow your way into sheltering an eight-digit income. That isn't right-and right has to have something to do with what we're talking about. "Bracket creep" is the big swindle. Which would seem to suggest that an equitable graduated income tax should be levied incrementally: that is, income up to $X (including capital gains) would be taxed at the base rate, then income above $X but below $Y at the next bracket, then income between $Y and $Z at the next, and so on up to the top. Brackets could be determined by the census and the I.R.S., using the same kind of "multiplier" calculations that are used to point up the disparity injustice between factory-floor wages and chief executives' compensation. You've all heard those numbers: how today chief executives earn roughly 475 times what the assembly-line worker does, whereas 20 years ago the multiplier was a lot less. Why not base tax bracketing on the same kind of calculation, and update it every census year? Thus you could have a base rate of 10 percent on income up to $25,000 rising by, say, ten 3 percent increments, to 40 percent on marginal income in excess of $10 million. Or $5 million. Or $1 million. Or $25 million-or whatever formula works to achieve a fiscal balance. The beauty of a $3 trillion surplus (excluding Social Security overages) is that it buys whatever time it takes to get the arithmetic right. Given the advanced state of financial information technology and data retrieval, we should be able to do a better job than we do about coming up with numbers that bear some relation to observed reality. Frankly, the arithmetic doesn't bother me. Back a hundred years ago, when I was a partner of Lehman Brothers and (lacking computers) we had to feel our way into the numbers on a deal, I was always surprised by how neatly the arithmetic tended to work out if we got the basic idea right. Scientists say the same thing of their work. E = mc 2 isn't where Einstein started, it's where he ended; it's the mathematical confirmation of a hunch. One theorist on whom I tried out my idea of "multiplier-based" tax bracketing told me it amounts to "codification of class war." I just don't believe that. I think the same kind of approach could apply to estate taxation: 10 percent on estates up to $1 million, rising thereafter to 40 percent on incremental wealth in excess of $100 million. Or $500 million. As I see it, there are only two strong arguments that can be made for the elimination of estate taxes. One-the more persuasive-is that it forces the untimely sale of closely held family businesses. Lower brackets will help mitigate this problem, as will permitting estate-tax payment to reflect the liquidity of an asset, denominated in kind or as a slow note. The anti-death-tax casuists also argue that since the rich don't pay the death tax anyway, thanks to shelters and legal strategies, why bother to have them? Frankly, I find this to be as unacceptable morally and ethically as it may be unarguable arithmetically. To repeat myself: The estate tax is more important for what it says about the way many of us think about this country than for what it collects in terms of cash money. The point man in the administration's selling effort is going to be Paul O'Neill, the new Treasury Secretary. I find him puzzling. He's not Wall Street's man. In a way, that's good, although we should factor in the consideration that buying and selling securities is our main industry now: It is to 21st-century America what farming was to the 19th and heavy manufacture to the 20th. Mr. O'Neill says: no more bailouts. Sounds good-but we'll see. Bailouts may be necessary, but those responsible for the mess should be held accountable. Personally, I'd advocate mandatory federal prison terms for directors of federally bailed-out enterprises, say one year for every $100 million of bailout. Mencken thought such people should be hung, but I still haven't made up my mind on capital punishment, so incarceration will have to do. All in all, Mr. O'Neill sounds very prudent and practical. Even though he has been quoted to the effect that Fannie Mae and Freddie Mac aren't subsidized, which is either ignorant or disingenuous, neither of which is acceptable in a person with his new job. This sure as hell isn't what the market thinks when it buys Fannie and Freddie's debt at preferentially low interest rates (get hold of Grant's Interest Rate Observer of Feb. 2 and bone up on the "government sponsored" F.M. twins and how they do that thing they do). Let us not forget that a recent Fannie Mae chief executive retired with a severance package of $27 million, which is enough to pay the salaries of the entire Senate and 20 percent of the House. Mr. O'Neill accomplished great, creative things at Alcoa, one of the rustiest American companies, which is cause for confidence, unless it turns out he used the Lucent accounting model. On balance, he projects common sense, greatest of all American qualities in a pinch, but I also detect more than a whiff of C.E.O. "I know best." He will be interesting to watch. The market has given back whatever it gained since the first of the year. That's the tape's way of telling us it doesn't like what it feels. I've been prepared-strictly because it's early days-to give Mr. Bush and his colleagues the benefit of patience. But this tax business is their baby. And we do have to face a troubling fact: One thing we need at the top is some rational perspective on the subject of money. We have gone from an administration of people who probably had too little wealth in their lives to an administration of people who have probably had too much (and, in most cases, to whom the wealth has come quite late, and probably quite traumatically). Just as the first led to greed in the Clinton regime, the other may lead to avarice under Mr. Bush-which is equally to be guarded against. Greed and avarice, Ivan Boesky and Scrooge McDuck to the contrary, are never good-least of all as policy.