There's gathering interest on the Democratic side in a new extra-top rate on very-high-income people. It's fine (they think) to whack the over-$200,000 crowd with a 39.6 percent rate. But something much more Draconian is required for the top 0.1 percent, the people who really throve during the Bush years.

For now, this talk is empty. Republicans will score big gains in November. Even if Republicans lack the clout in the next Congress to force renewal of all the Bush tax cuts, there’s no way will they tolerate the creation of a hefty new super tax bracket.

But the emergence of a political idea within one of the big parties is always significant, whether that party holds power or not. Sooner or later, today's "outs" become tomorrow's "ins." Ideas that sound outrageous now may edge their way into the debate if conditions get sufficiently desperate. Given the horrific deficits facing the U.S. Treasury, any idea that promises to raise revenue bears watching.

The super-rich have access to accountants and lawyers, and they know how to use them.

So it’s worth thinking about why an ultra-high top rate of income tax for ultra-big earners remains an ultra-terrible idea.

The short answer is: We've tried it before. In the 1970s, the United States counted almost a dozen tax brackets, reaching all the way to 50 percent. Faced with the loss of half their next dollar of income, high earners deployed amazing ingenuity to move money out of the taxman's reach. They invested in tax shelters to generate fictitious losses, concocted schemes to redesignate income as lower-taxed capital gains.

The most notorious of these games involved strategies called "straddles." Investors would create a chapter S corporation to trade currency or commodities. The trades would be carefully hedged to create equal gains and losses. The gains would be retained within the corporation, theoretically subject to capital gains tax at some future date. The losses would flow directly to the investors, reducing the net income on which they owed tax.

These tax-avoidance strategies saved some taxpayers some money. But on balance they rendered everybody worse off.

Capital flowed not to its best use, but into bogus transactions designed to dupe the taxman. Accountants and lawyers squandered their creativity on ideas that created no wealth. Investors accepted lower returns than they might otherwise have earned in order to keep a higher proportion of what they did earn.

What was "tax efficient" for the wealthy was not "economically efficient" for the nation.

The IRS took action against straddles in the 1980s.

But the most important action against tax avoidance occurred in the tax reform of 1986. The proliferation of tax rates was reduced to just two, 15 percent and 27 percent. The distinction between capital gains and ordinary income was abolished. Most credits and deductions were abolished. At those low rates - and with most loopholes abolished - most Americans decided it made more sense to forget about tax efficiency. Better to work and invest in whatever way made the most economic sense, pay a reasonable rate of tax, and keep the rest. The tax reform curbed tax avoidance by encouraging tax compliance.

The proposed billionaire's tax does just the opposite. A high new top rate for the most affluent and sophisticated will only invite the affluent and sophisticated to invest in devices to avoid that rate. Tax engineering will revive. Rich people will resume the hunt for strategies that reduce their obligation instead of maximizing investment returns.

Such a colossal waste! All to score a symbolic point that will not generate revenue for the Treasury or help society's poorest. Punishment is a bad guide to social policy. Must we again learn that lesson the hard way?