Similarly, when the president calls for an increase in the child care credit, that’s also treated as a tax cut rather than the rise in spending that it actually is.

According to calculations of the Treasury Department that are hidden deep in the government’s annual budget, there are hundreds of billions of dollars of spending every year that are recorded as tax reductions. The biggest of these “tax expenditures,” as they are called, is the exclusion of employer health insurance premiums from the taxable income of employees. That exclusion resulted in a tax reduction of $160 billion in 2010 and is projected to be $1.4 trillion between 2010 and 2016. That’s a $1.4 trillion subsidy to health insurance that is disguised as a tax reduction.

Limiting total tax expenditures could produce enough revenue to achieve very substantial cuts in future budget deficits while also lowering personal tax rates. Indeed, if tax expenditures are not reduced, Congress will eventually be forced to raise tax rates or to introduce new taxes. That’s because there is no way to achieve the needed reductions in future deficits just by cutting traditional government outlays, even if there are further cuts in defense spending. And slowing the growth of Social Security and Medicare is needed just to avoid an explosion of future spending on those programs.

Limiting tax expenditures should have bipartisan appeal. Republicans should welcome limits on tax expenditures as a way to cut hidden government spending. Democrats should accept it as a way to raise the revenue that they insist must be part of any deficit reduction plan. And as I will explain below, it is also a natural way to achieve an automatic “failsafe” mechanism to make sure that deficits decline as promised.

Although limiting the use of tax expenditures would produce additional tax revenue, it is very different from other possible revenue increases. It doesn’t raise marginal tax rates, doesn’t discourage work or entrepreneurship, and doesn’t tax saving and risk taking. It is really a reduction in government spending, not a tax increase. And deep enough cuts in tax expenditures would actually allow reductions in personal tax rates as well as in budget deficits.

I am surprised that some conservatives who favor cutting government nondefense spending oppose limiting tax expenditures because they regard the resulting increase in tax revenue as a tax increase. That fails to distinguish between the accounting effect of cutting tax expenditures and the economic effect. Although government accounting rules treat the end of a tax credit or the limit of a tax deduction as a revenue increase, the economic effect is the same as a cut in spending. Anyone who favors less government spending should also favor cutting tax expenditures.

At a personal level, reductions in tax expenditures would be unpopular with those who see their own benefits reduced. That’s true of any form of spending cut, whether through the tax code or the outlay side of the budget. But voters who recognize the seriousness of the deficit situation may nevertheless support a limit to tax expenditures if they believe that all taxpayers are being asked to accept a sacrifice.

That’s why I think that putting a cap on the total value of tax reductions that each individual can achieve through tax expenditures is better than those plans that would limit only a few of the tax deductions. More specifically, I favor limiting the tax reduction that individuals can achieve through itemized deductions and the exclusion of employer health insurance payments to 2 percent of each taxpayer’s adjusted gross income. Such a cap would allow each taxpayer to benefit from all of the current tax rules but would limit the total resulting tax reduction to no more than 2 percent of that taxpayer’s total income.

Note that the limit applies to the resulting tax reductions and not to the deductions themselves. A taxpayer with a 15 percent marginal tax rate who pays mortgage interest of $3,000 a year would have to count 15 percent of that $3,000, or $450, toward his allowable cap.

I have analyzed the revenue effects of such a plan in a study with Daniel Feenberg of the National Bureau of Economic Research and Maya MacGuineas of the Committee for a Responsible Federal Budget (our paper, “Capping Individual Tax Expenditure Benefits,” is available online). We found that if a 2 percent cap were in effect in 2011 it would result in additional tax revenue of $278 billion. That’s the extra revenue for just a single year. In a decade it could produce more than $3 trillion of additional revenue, enough to achieve a substantial reduction in the national debt while also permitting cuts in personal tax rates.

The tax expenditure cap could also play a key role in a failsafe plan. The tax expenditure cap could initially be set at a larger percentage of adjusted gross income, thus producing less revenue, but could be scheduled to become a tighter cap if additional revenue is needed. An initial cap of 5 percent of adjusted gross income would produce revenue of $110 billion in 2011. If the other outlay cuts and the rise in revenue resulting from economic growth brought the deficit down along an agreed path over the coming decade, the 5 percent cap could remain. But if deficits remained unacceptably high, the cap could be reduced to 3 percent or 2 percent.

President Obama has said that he would include reductions in tax expenditures as part of a failsafe plan to achieve reductions in future deficits. That is a welcome sign, but it is important how those reductions are achieved. To be seen as fair and to be capable of raising substantial revenue, they should apply to all deductions and to the health insurance exclusion. The limit should apply to all taxpayers and not just to high-income taxpayers, as the president suggested when he proposed limiting the tax rate used for tax deductions to 28 percent. But a broad cap applied to all taxpayers should be a central part of the deficit reduction plan and could be used as an automatically triggered failsafe option.

Martin Feldstein is a professor of economics at Harvard. He was chairman of the Council of Economic Advisers from 1982-84. .