The Tax Reform Evidence from 1986

Congress's Joint Select Committee on Deficit Reduction is struggling to find $1.5 trillion in cuts over the next 10 years. This is a unique opportunity to use tax reform to reduce future budget deficits while lowering individual tax rates.

The Tax Reform Act of 1986, enacted 25 years ago last Friday, showed how a tax reform that includes lower rates can change incentives in a way that grows the tax base and produces extra revenue. The 1986 agreement between President Ronald Reagan and House Speaker Tip O'Neill reduced the top marginal tax rate to 28% from 50%. A conservative Republican and a liberal Democrat could agree to a dramatic reduction in top rates because the legislation also eliminated a wide variety of tax loopholes.

A traditional "static" analysis that ignores the response of taxpayers to lower tax rates indicated that those combined tax changes would leave total revenue unchanged at each income level. But the actual experience after 1986 showed an enormous rise in the taxes paid, particularly by those who experienced the greatest reductions in marginal tax rates.

To measure that response, I studied a sample of individual tax returns (stripped of all identifying information) for more than 4,000 taxpayers provided by the U.S. Treasury Department. Because the sample contained the tax return of each individual for the years 1985 through 1988, I could compare the taxable income of individuals in 1985 with their taxable incomes in 1988, two years after their rates were lowered.

Taxpayers who faced a marginal tax rate of 50% in 1985 had a marginal tax rate of just 28% after 1986, implying that their marginal net-of-tax share rose to 72% from 50%, an increase of 44%. For this group, the average taxable income rose between 1985 and 1988 by 45%, suggesting that each 1% rise in the marginal net-of-tax rate led to about a 1% rise in taxable income.

This dramatic increase in taxable income reflected three favorable effects of the lower marginal tax rates. The greater net reward for extra effort and extra risk-taking led to increases in earnings, in entrepreneurial activity, in the expansion of small businesses, etc. Lower marginal tax rates also caused individuals to shift some of their compensation from untaxed fringe benefits and other perquisites to taxable earnings. Taxpayers also reduced spending on tax-deductible forms of consumption.

A similar picture emerged for the group of taxpayers who faced slightly lower marginal tax rates of 42% and 45%. The reduction to 28% raised the marginal net-of-tax share of this group by 25% and their taxable incomes rose by 20%, suggesting that each 1% rise in the marginal net-of-tax share raised taxable incomes by 0.8%, quite similar to the estimate for the group with the highest marginal tax rate.

The substantial sensitivity of taxable income to the taxpayer's marginal net-of-tax share has important implications for the effect of tax-rate reductions on total tax revenue. For a 10% across-the-board reduction in all tax rates, a traditional "static" analysis implies that revenue would fall to 90% of its previous level. But reducing a current 40% marginal tax rate by 10% to 36% raises the net-of-tax share to 64% from 60%, a rise of 6.7%. If that causes the taxable income of those at that tax level to rise by 6.7%, their taxable income would fall to only 96% of what it had been. In short, the behavioral response of taxpayers in this highest bracket would offset 60% of the static revenue loss.

The effect of taxpayer behavior on revenue is smaller in lower tax brackets. Calculations using the National Bureau of Economic Research's TAXSIM model, which calculates federal and state income tax liabilities from survey data, indicate that a 10% across-the-board reduction in all federal tax rates would reduce revenue by about 60% of what a static analysis would imply—i.e., that the behavioral response of taxable income to the lower marginal tax rates would offset about 40% of the static revenue loss.

These calculations have important implications for today's deficit-reduction debate. Broadening the tax base by limiting the use of tax expenditures (the special tax rules that substitute for direct government spending as a way to subsidize health insurance, mortgage borrowing and other things) could raise substantial revenue. Doing so doesn't require eliminating any of those tax expenditures. In a study of recent Treasury data, Daniel Feenberg, Maya MacGuineas and I found that limiting each individual's tax reduction from the use of tax expenditures to 5% of that individual's adjusted gross income would raise revenue equal to about 10% of current personal tax revenue.

Combining that base broadening with a 10% cut in all tax rates would be revenue neutral in a traditional static analysis. But the experience after the 1986 tax reform implies that the combination of base broadening and rate reduction would raise revenue equal to about 4% of existing tax revenue. With personal income-tax revenue in 2011 of about $1 trillion, that 4% increase in net revenue would be $40 billion at the current level of taxable income, or more than $500 billion over the next 10 years.

The Joint Select Committee should insist on counting that revenue as the starting point for a serious deficit reduction plan.