In the Wall Street Journal, Phil Gramm, the former Republican senator from Texas, insists that tax reform not make the tax code more progressive than it is at present, but his analysis rests on a misconception:
First, under no circumstances should Republicans agree to make the tax system even more progressive than it already is, or to increase the number of people who do not pay income taxes. In 1980, the top 1% and 5% of income earners in America paid 19.1% and 36.9% of total federal income taxes. Today, the top 1% and 5% pay 37.4% and 59.1%. Meanwhile, 41.6% of American earners now pay no federal income taxes.
The more progressive the tax system becomes the more unstable the country’s public finances get. High-income Americans earn a large share of their income in bonuses, dividends and capital gains, all of which are highly sensitive to the business cycle. This means wide swings in tax collections that play havoc with government budgets. The removal of large numbers of people from the tax rolls makes the political system more unstable. Individuals and households that pay no income taxes have a diminished stake in limited government.
Gramm’s point about how a steeply progressive tax code generates a more volatile revenue stream than a relatively flat tax code is well taken. Yet this is precisely why it is important that the federal tax code be more progressive, as the federal government is able to smooth spending across the business cycle while state and local governments are severely constrained in their ability to do so. State and local governments are thus obligated to rely on revenue sources that are both stable and regressive, like retail sales taxes and property taxes. When state and local governments rely more heavily on progressive taxes, like taxes on capital income, they find themselves in dire straits. (See California.) The federal government, in contrast, is well-suited to handling volatility.
And as Ramesh Ponnuru has argued on numberous occasion, the notion that “the removal of large numbers of people from the tax rolls makes the political system more unstable” is problematic. Among other things, it ignores life cycle effects. Individuals who earn low incomes during one phase of life may earn higher incomes later on. Removing parents of young children from the tax rolls by expanding the child tax credit won’t necessarily create a constituency for state expansion, as these parents recognize that they will lose their tax-advantaged status once their children are adults. Individuals who earn low incomes throughout their lives are subject to a relatively low federal income tax burden, payroll taxes aside, yet these individuals have limited political influence, and it is not at all clear that not forcing these individuals to pay federal income taxes contributes to political instability. Indeed, one could argue that increasing the federal income tax burden on the least affluent would do more to increase political instability than decreasing it, as it might foster resentment. Gramm is making a specific claim that seems difficult to justify.
All that said, I don’t think we need a tax code that is substantially more progressive than the current tax code. It’s just that Gramm’s arguments tilt in the direction of a tax code that would be substantially less progressive, and that would be a mistake.