Last week, economists Martin Feldstein and Larry Summers sparred over tax reform on a panel discussion hosted by the Brookings Institution’s Hamilton Project. It is widely agreed that Washington is overdue for and likely headed toward comprehensive tax reform in the next few years. Any compromise on a long-term deficit reduction package will necessitate more revenue and a quarter-century has passed since the last major scrubbing of the tax code (though the Tax Reform Act of 1986’s revenue- and distributional-neutrality makes it an inappropriate benchmark for reform). Most of the reform proposals have focused on broadening the tax base—i.e., eliminating or curbing tax expenditures—rather than raising marginal tax rates to increase revenue. Tax reform is also typically couched in the objectives of simplifying the tax code, reducing economic distortions, and creating a “pro-growth” tax code.

Summers raised an excellent point about the tax reform discourse, particularly the way tax simplification is spun by many advocates. Simplifying the tax code is often falsely equated with reducing the number and scale of marginal tax rates. The House Republican budget, for instance, proposed consolidating the income tax to two brackets—of just 10 and 25 percent—as the cornerstone of “simplifying the tax code and promoting job creation and economic growth.” But in the age of tax return software, Summers argued that tax rates don’t add to complexity; the complicating factor is calculating taxable income. Even before TurboTax, citizens managed to navigate a much longer schedule of marginal tax rates; the tax code had at any time between 24 and 26 marginal tax rates between World War II and 1978. The tax code was also much more progressive, particularly at the top of the income distribution, and income growth was evenly shared over this period. At present, compressing the number of marginal tax rates (currently at six) would only further decouple the contours of the tax code from the increasingly uneven distribution of income gains.

Reform is often framed in terms of broadening the tax base and simplifying the code in tandem, but Summers noted the two are often contradictory objectives. Base broadening inherently increases the amount of taxable economic activity, which in some circumstances can actually make the tax code more complex. Summers gave the example of repealing the exclusion on owner-occupied home sales from capital gains taxation, which would broaden the tax base and raise revenue but unquestionably increase complexity for many filers. Summers downplayed the need for simplification—at least as it’s often thought of—in this tradeoff, and instead endorsed base broadening in ways that increase revenue and the perception of tax fairness.

The complex nature of the tax code favors—both in effect and appearance—those with numerous financial planners and accountants, including dozens of Fortune 500 companies paying nothing in taxes. The tax code needs to be cleaned in ways that inhibit tax gaming and avoidance—best accomplished by reducing tax loopholes that allow for income shifting, such as the preferential rates on capital gains and dividends or the tax deferral for U.S.-multinationals’ foreign source income. That largely amounts to base broadening. Flattening the schedule of marginal income tax rates, on the other hand, would not improve tax compliance or simplify the tax code but would markedly reduce both tax fairness and revenue. When push comes to shove, much of what is peddled as “tax simplification” is nothing but an attempt to further undermine tax progressivity and reduce effective tax rates for upper-income households; doing so would amount to a tax cut, not tax reform.