Back in 2007, the economists Kenneth Scheve and Matthew Slaughter co-authored an essay in Foreign Affairs calling for “A New Deal for Globalization?” The thesis is that the gains from economic openness are very large, e.g., the authors cite a Peterson Institute study which suggests that trade and investment liberalization to date have increased U.S. annual income by $500 billion to $1 trillion, and they suggest that a successful Doha agreement (no longer in the cards, sadly) would have added an additional $500 billion. Yet the dislocation caused by economic openness is sufficiently great that the U.S. needs to consider a more robust form of redistribution to less-skilled and mid-skilled workers. And so they tentatively propose a dramatic revision of the payroll tax:
Truly expanding the political support for open borders requires a radical change in fiscal policy. This does not, however, mean making the personal income tax more progressive, as is often suggested. U.S. taxation of personal income is already quite progressive. Instead, policymakers should remember that workers do not pay only income taxes; they also pay the FICA (Federal Insurance Contributions Act) payroll tax for social insurance. This tax offers the best way to redistribute income.
The payroll tax contains a Social Security portion and a Medicare portion, each of which is paid half by the worker and half by the employer. The overall payroll tax is a flat tax of 15.3 percent on the first $94,200 of gross income for every worker, with an ongoing 2.9 percent flat tax for the Medicare portion beyond that. Because it is a flat-rate tax on a (largely) capped base, it is a regressive tax — that is, it tends to reinforce rather than offset pretax inequality. At $760 billion in 2005, the regressive payroll tax was nearly as big as the progressive income tax ($1.1 trillion). Because it is large and regressive, the payroll tax is an obvious candidate for meaningful income redistribution linked to globalization.
A New Deal for globalization would combine further trade and investment liberalization with eliminating the full payroll tax for all workers earning below the national median. In 2005, the median total money earnings of all workers was $32,140, and there were about 67 million workers at or below this level. Assuming a mean labor income for this group of about $25,000, these 67 million workers would receive a tax cut of about $3,800 each. Because the economic burden of this tax falls largely on workers, this tax cut would be a direct gain in after-tax real income for them. With a total price tag of about $256 billion, the proposal could be paid for by raising the cap of $94,200, raising payroll tax rates (for progressivity, rates could escalate as they do with the income tax), or some combination of the two. This is, of course, only an outline of the needed policy reform, and there would be many implementation details to address. For example, rather than a single on-off point for this tax cut, a phase-in of it (like with the earned-income tax credit) would avoid incentive-distorting jumps in effective tax rates. [Emphasis added]
Note that $256 billion is an annual price tag, and that Social Security’s status as a “self-financing” program is very much in jeopardy. The potential political appeal of the Scheve and Slaughter proposal isn’t as obvious as one might think. Workers earning less than the national median are relatively disengaged from politics, though of course that is not a permanent condition, and a sharp increase in taxes on households earning more than the median would presumably spark a backlash. In a perfect world, perhaps we would merge the payroll tax and the income tax into a unified progressive tax of some kind, as Megan McArdle and others have proposed.
Regardless, I think that Scheve and Slaughter are on the right track in focusing on the payroll tax. One of the more consequential potential payroll tax reforms is Robert Stein’s proposed reform of the federal income tax, which would effectively remove large numbers of low-income and middle-income households with children from the payroll tax rolls:
To correct for this inadequate treatment of households with children, the existing dependent exemption for children, the child credit, the child-care credit, and the adoption credit should be replaced with one new $4,000 credit per child that can be used to offset both income and payroll taxes. (This amount is set much closer to the $3,250 figure than the $8,500 one mostly to reduce the plan’s negative impact on federal revenue.)
The new child credit would accomplish several significant policy goals. First, it would offset the anti-parenting bias created by Social Security and Medicare. Second, the credit would help simplify the tax code by getting rid of other exemptions and credits that apply to children. Third, and very important for many families, it would end the bias against families with a stay-at-home parent now caused by the child-care credit (which applies only if both parents are working for pay). And finally, it would reduce effective marginal tax rates for many middle-class families. [Emphasis added]
The political basis for Stein’s reform is different from Scheve and Slaughter’s, but it also addresses the payroll tax burden in a potentially appealing, and more to the point socially and economically beneficial, way.