The weird gyrations of the Senate Finance Committee, where that chamber's version of "tax reform" has been incubating, show that it's easier to be principled in private than in public.

Until about two weeks ago, public sessions, under the scrutiny of the lobbies, were turning the Senate tax-reform bill into the ultimate Christmas tree, laden with tax preferences. The prospects for terminal self-parody seemed excellent.

Then a funny thing happened. Chairman Bob Packwood of Oregon wearied of reading that he was deliberately sabotaging tax reform. He called off the open votes, shut the doors, huddled with his staff and came up with a pure bill offering simplicity and flatness in nearly pristine form.

Packwood II proposes only two tax brackets, a top of 27 percent, and 15 percent for about 80 percent of American taxpayers. Preferences and deductions for business are slaughtered. With this prototype, Packwood has managed, in private, to move a "streamlined" bill along. It has been brilliant politics, though the fate of the bill on the Senate floor and in conference is uncertain.

But the manner of this change -- and the committee's 166-degree swing -- raise many questions.

If the Senate's key taxation committee can box the tax-reform compass in a couple of weeks, what does that say about its basic convictions? Can a majority really have believed in March that complex taxation is good taxation, but believe in May that a good tax code is radically simple?

And what about those who base business and personal investment plans on their tax liabilities and incentives? What degree of consistency might they reasonably expect? Indeed, what coherent political standards, if any, does so mercurial a process meet?

Tax reformers dogmatically equate simplicity with equity. It is the unexamined premise of the whole exercise. But even if Packwood has been converted to that view, it ain't necessarily so.

Let us agree for the sake of argument that 14 income-tax brackets are too many, and a 50 percent top rate is excessive. Does it follow that two tax rates are enough, or that halving the maximum rate is a good idea? Out of what Platonic vision of the good society was that dogma plucked?

And what about policy? Packwood II would, for example, abolish the hard-fought change of some years ago that scaled back tax rates on capital gains. That change, also celebrated in its day as "reform," has stimulated investment and capital formation. Packwood II's proposed abolition of the capital gains preference, especially when coupled with the denial of IRA deductions to pensioned taxpayers, will decrease the tax code's hospitality to savings and investment. This one-two punch could significantly damage the U.S. competitive position in international markets. It is simple, all right. Is it good policy?

Businesses have no natural right to this or that tax preference, even if it is on the books. But tax decisions do affect real people, more workers than managers, and Congress owes corporations and individuals some reasonable predictability over time. Having induced the herd to run one way five years ago, Packwood II would start a stampede in the other direction. There will be dislocations; jobs will be lost.

When you begin to ask such questions, it becomes obvious why the Finance Committee's performance is a laboratory demonstration of the flaws of the tax-reform process.

Not so long ago, strong chairmen such as Russell Long and Wilbur Mills could pretty much run tax policy out of their vest pockets. That was excessively autocratic, but policy conformed to some coherent scheme even if it was wrong. Now, with the atomization of Congress, tax writing suffers from an excess of democracy -- at least when committees do it in the street and not in the closet.

Packwood I showed that a tax policy that is everybody's baby is nobody's baby, a foundling of astonishing incoherence. Significantly, Packwood II achieved coherence only in secret, behind closed doors. And coherence predictably began to be lost when accountability returned with open votes.

The lesson, plainly, is this: Congress, to reform taxes usefully and clearly, must first reform itself.