Such a measure is certainly worth considering as part of a fiscally responsible, balanced-budget package of tax reform. But as short-run macroeconomic stimulus, it is next to worthless, and in terms of its long-run impact on the deficit, it is downright irresponsible.

The Democrats' plan actually offers some stimulus. If it works as well as they hope, it would lead to robust growth by the summer of 2004, giving a big lift to Mr. Bush's re-election prospects. Hmmm. Perhaps the Democrats aren't quite as enthusiastic about their plan as they pretend to be.

One interpretation of the Bush proposal is that the White House believes that left to its own devices, the economy will recover by summer 2004 anyway, so the administration may as well carry on with its tax-cutting agenda now.

But leaving short-run politics and macroeconomic stimulus aside, what would be the long-run impact of tax-free dividends on financial markets? In thinking about this question, it is important to keep in mind two relatively recent developments.

First, we really have become a nation of shareholders. According to ''The Rise of the Equity Culture,'' a paper by the M.I.T. economist James Poterba, the number of individuals owning corporate stock has increased by nearly 60 percent in a decade, with about half of American households now owning stock, either directly or indirectly.

Second, much of this stock is held in tax-deferred accounts like 401(k) plans, IRA's, and Keoughs. This means that more than half of dividends paid already escape immediate taxation. But if the tax treatment of dividends is changed, portfolio allocations could also change, and a complete analysis has to examine the impact of such adjustments.

Let's briefly review the theory of portfolio choice in the presence of tax-deferred accounts. Suppose an investor has a tax-deferred account and a taxable account. He can invest in a risky asset (stocks) or a safe asset (corporate bonds).

Stocks tend to have most of their payout in the form of capital gains, which are taxed relatively lightly. Bonds pay out interest, which is highly taxed. Hence it makes sense to put the bonds in the tax-deferred account, holding the lightly taxed stocks in the taxable portfolio.

A neat theory, but, unfortunately, it doesn't hold up empirically. Professor Poterba, along with two other professors, John Shoven of Stanford University and Clemens Sialm of the University of Michigan, looked at the consequences of various investment strategies, using the actual returns on various mutual stock and bond funds over the last 36 years.

Contrary to the conventional wisdom, they found that allocating stocks to the tax-deferred account, and holding extra income in nontaxed municipal bonds, was far and away the best strategy.

There are two reasons for this discrepancy between theory and practice. First, most managed mutual funds are highly ''tax inefficient.'' They typically pay dividends and capital gains distributions quarterly or yearly, making investors pay taxes they would rather defer.

Second, municipal bonds typically have attractive yields compared with the after-tax yields on corporate bonds, particularly for investors in high tax brackets. Taken together, these two facts turn conventional wisdom on its head.

Still, the conventional wisdom holds true for some mutual funds. If you invest in an index fund, a highly tax-efficient vehicle, the economists show that it is best to hold it outside your tax-deferred portfolio. Since index funds, and other tax-optimized funds, have become more popular in recent years, the conventional wisdom is looking much wiser these days.

If taxes on dividends were eliminated, there would be an even greater incentive to hold stocks outside a tax-sheltered portfolio.

So we would expect to see investor portfolios shift more in the direction the theory predicts: taxable bonds in tax-deferred accounts, and stocks in taxable accounts to take advantage of lightly taxed capital gains and untaxed dividends.

What about municipal bonds? Odds are we would see their prices fall, since dividend-paying stocks would be pretty close substitutes under the Bush proposal. This means the cost of borrowing for state and local governments will be driven up -- particularly bad news given their precarious economic position.

One way to estimate the likelihood of the Bush plan's passing is to watch the prices of municipal bonds over the next few months: the more likely the plan is to pass, the lower those prices are likely to go.

Judging from the reaction of the municipal bond market so far, and the prices of dividend-paying stocks, the markets seem to think it unlikely that the Bush plan will actually pass.

So far we've focused on the demand for dividends; if the tax is eliminated, what would be the consequences on the supply of dividends?

Economic theory isn't very helpful here. Dividends have been so heavily taxed in the past that it's something of a mystery why companies pay them at all. The most compelling view is that investors don't trust managers to make sensible investment decisions with retained earnings, so they prefer the dividend bird in the hand to the capital gain pie in the sky.

If paying dividends became the norm, companies would have to subject more investment plans to market scrutiny, which, by and large, would be a good thing.

But a cynic might say little will change in this regard since the Bush plan allows a company to declare a dividend, but not pay it out to shareholders, instead keeping the money as retained earnings. This provision weakens a substantial benefit of changing the tax treatment of dividends.