BETWEEN 1998 and 2000, the United States pulled off a rare and fleeting miracle: For the first time since 1969, it ran federal budget surpluses. This turnaround owed something to the tax increases passed in 1990 and 1993 and something to the spending discipline enforced by the Newt Gingrich Congress. But it also owed something to a mysterious revolution that occurred inside American businesses. The average American worker, whose output had been growing steadily at around 1.5 percent a year since the 1980s, suddenly clocked productivity gains of 2.7 percent per year between 1996 and 2000. That acceleration drove the economy's growth rate up to 4 percent, pushing tax receipts to an all-time high and creating budget surpluses that nobody had predicted.

If the Bush administration knew how to create the conditions for another growth spurt, critics might have to rethink their opposition to tax cuts. According to the Berkeley-Brookings projections we cited yesterday, the deficit is likely to register at around 3.5 percent of gross domestic product in 2014. But if the economy grows by just under 4 percent a year, rather than just under 3 percent as assumed in the projection, the deficit in 2014 would come to a far less alarming 0.5 percent of GDP.

What could President Bush do to boost growth? His officials argue that tax cuts will contribute, but this seems unlikely. Lower tax rates on wages do boost the labor supply; lower tax rates on investment may boost savings; more labor and more capital mean more economic output. But Mr. Bush's tax cuts also have an offsetting consequence. Because they have not been accompanied by spending cuts, government borrowing has gone up, nudging everybody's cost of borrowing higher than it would be otherwise. A range of econo- metric studies suggest that these opposing impacts -- more labor and capital on the one hand, higher interest rates on the other -- roughly cancel one another out. Therefore, to boost growth, the Bush administration will have to look elsewhere.

One option is trade policy. Fully liberalized global trade would create a boost to GDP of 2 percent, according to Harvard University's Jeffrey Frankel, who served on President Bill Clinton's Council of Economic Advisers. But this would be a one-time boost to the size of the economy, not a shift in the growth rate, so it would close only a small part of the deficit forecast for 2014, leaving untouched the monster deficits thereafter. Besides, completely free trade is a remote hope. Even success in the World Trade Organization's ongoing Doha Round of trade talks would fall short of that target.

Another option is to tackle the absurd tort system, which claims a far higher share of GDP than in any other advanced country. Reform, if it ever could pass Congress, would boost growth by reducing litigation costs, freeing money that might fund innovation and research, and -- by reducing companies' propensity to withhold products from the market -- eliminate the needs to order unnecessary safety tests and waste time on defensive strategies that are more about reducing legal exposure than about safety. But how much extra growth would this yield? Robert E. Litan of the Brookings Institution puts the answer at just 0.1 percent of GDP per year.

The same goes for regulatory reform, the policy of weeding out expensive regulations that buy little benefit in terms of health and safety, and replacing these with others that are less burdensome and more effective. A smart benefit-cost review of the regulatory state could reduce the drag on businesses without harming citizens. But though this yields some effect on the growth rate, the administration concedes that it's probably too small to measure.

What of boosting government investment in education and research spending? An extra year of education has been shown to boost a worker's lifetime wages by 5 to 10 percent, more than covering the cost of the tuition. Equally, government-funded research can generate growth-enhancing breakthroughs from biotechnology to the Internet. But the links are not always certain between extra education spending and productivity improvements in the workforce as a whole or between extra research spending and additional scientific breakthroughs. Although investing more in both areas would be a good idea, the payoff in terms of growth would be distant and impossible to measure.

The government's options are limited. The experience of the late 1990s shows that higher growth, unlike Social Security reform, has the potential to transform the fiscal outlook. But it's equally true that we don't know how to repeat the 1990s miracle, and the government's policy options -- a trade deal, tort reform, deregulation -- aren't powerful enough to do it. The productivity revolution inside American companies seems to have reflected technological and organizational changes that had been percolating inside corporations for at least a decade, none of which had much to do with government policy. It would be nice to believe that a second productivity revolution will fix the looming fiscal crisis. But it seems unwise to count on it.