Our tax system is in serious disrepair, but the double tax on dividends is not the most pressing problem.

Ending taxes on dividends might temporarily generate more commissions for Charles Schwab, the brokerage executive who promoted the idea to President Bush at a conference last August. But economists at the Congressional Research Service recently concluded that a dividend exclusion is unlikely to stimulate the economy.

As tax experts, we recognize that there are legitimate reasons to eliminate the double tax on dividends. However, we believe that these reasons are outweighed by the distortions that Bush's proposal would create, the complexity that it would add to our already mammoth Internal Revenue Code and the cottage industry of tax shelters that will develop if the proposal is enacted. In the past two years, Japan, England and Germany have repealed or scaled back their own schemes to avoid double taxation of dividends. Congress should not repeat their initial mistake by enacting one.

The legitimate reason to exempt dividends from tax is this: Under our current tax system, corporate earnings are effectively taxed twice -- first, in the hands of corporations, which are taxed on the profits they make, and then again when the payouts reach the hands of a taxable investor. This double hit, some experts argue, discourages businesses from organizing as corporations, penalizes firms that pay dividends and encourages corporations to borrow money and pay tax-deductible interest rather than nondeductible dividends.

So goes the theory. In reality, it doesn't always work that way.

Virtually all companies (except some partnerships) whose shares are traded on public stock exchanges are taxed as corporations, and the benefits almost always outweigh the costs of a corporate tax. Thus, in 1999, when Goldman Sachs, formerly a partnership, went public, it voluntarily subjected itself to a corporate income tax in order to raise capital in public equity markets (and enrich its individual general partners). During the dot-com craze of recent years, firms rushed to do initial public offerings even though that subjected them to corporate taxes. So the corporate tax has not had any discernible effect on choices about corporate form.

One reason may be that U.S. corporations pay less tax each year. Although the top corporate tax rate is 35 percent, corporations on average pay only about 15 percent. And corporate tax collections, which were about 5 percent of GDP in the 1950s, have steadily declined to just over 2 percent today.

Despite the fact that dividends are taxable for individuals, they are already tax-free for the nation's largest shareholders: corporations themselves (which are entitled to a deduction for dividends received) and tax-exempt pension plans (which are not subject to tax on dividends). Only a third of all dividends are actually subject to tax. And there is no evidence that the taxation of dividend income has any significant effect on the dividend policies of publicly traded companies.

Although our tax laws encourage companies to borrow money and pay interest rather than dividends, exempting dividends from tax would be unlikely to cure this problem. Instead, it is equally likely that if dividends were exempt from tax, corporations would simply increase their borrowing rate to pay more dividends.

So the portrayal of the double tax on dividends as evil is more fiction than fact. Dividend exclusion is unlikely to have the desired effect of lifting the stock market, and the administration's proposal won't really cure the woes of our corporate tax system.

Instead, it raises serious issues. First, if the measure does encourage individuals to shift investments from taxable bonds into more risky tax-free dividend-paying stocks, the investing public will be less diversified and less cushioned for the next stock market downturn. And if the proposal succeeds in increasing dividend rates, corporations might borrow more to pay higher yields, making balance sheets all the more shaky.

Second, states and cities, already hard-pressed for revenues, will be squeezed further. Tax-free municipal bonds will become less alluring if dividends are even partially tax-exempt, and states and cities will have a much harder time raising funds for necessary infrastructure improvement. States whose income taxes piggyback on federal payments also will lose revenue.

Third, under most tax treaties, the United States would be obligated to provide any dividend tax-exemption to the residents of its treaty partners who buy U.S. stocks. So the United States would lose the tax it collects from these foreign investors.

And fourth, the exclusion will benefit the wealthy. The Urban-Brookings Tax Policy Center estimates that the wealthiest 1 percent of all taxpayers could capture as much as 42 percent of the benefits of a dividend exclusion. This "tax cut" won't put any more money in the hands of working Americans.

Excluding dividends from taxes would certainly add to the already tremendous complexity of the Internal Revenue Code. That's because the administration plan isn't a blanket exemption for the full amount of dividends. It exempts only the portion of dividends on which companies have already paid taxes. That would be a different amount for every company. Companies could bank and allocate credits as they see fit. There is no mechanism for doing this now.

Naturally, administration economists insist their plan will benefit individual taxpayers and stimulate the economy without causing irresponsible borrowing or hindering the states' ability to raise funds. The macro-economic effects are unknowable in advance.

But we are certain of one thing: Excluding dividends from taxation will create opportunities for new tax shelters. There will be taxshelters that permit corporations to artificially manufacture dividends for shareholders and shelters that permit shareholders to receive tax-exempt dividends without economically "owning" any of the underlying stock. Each of these shelters is easy to develop under current law.

If history is any guide, Congress will respond by enacting new rules to stop taxpayers from abusing the exclusion. Given the IRS's recent track record, these anti-abuse rules will run to hundreds of pages, few cheats will ever be caught, those cases will take years to litigate and ultimately the anti-abuse rules will prove ineffective.

It's no wonder that Japan, England and Germany have moved closer to the existing U.S. system. France and Italy are expected to shortly follow suit.

Greater priority should be given to correcting our tax system's deeper flaws. One candidate: The alternative minimum tax (AMT), which will hit some 35.5 million individual taxpayers by the year 2010, should be repealed, and the inappropriate deductions and tax credits should be eliminated.

Abandon the dividend-exclusion proposal. Instead, repeal the individual AMT. Focus on simplifying the Internal Revenue Code and protecting it against abuse. Reuven S. Avi-Yonah is the Irwin I. Cohn Professor of Law at the University of Michigan. David S. Miller is a tax lawyer in Manhattan.