A quick analysis based on class shows that the math simply doesn't add up, particularly for the poor and middle class.
The big news in campaign trail policy wonkery last week was the Tax Policy Center's white paper by Samuel Brown, William G. Gale, and Adam Looney arguing that it is mathematically impossible for the Romney tax plan to meet its described goals. Ezra Klein has write-ups here and here, and James Pethokoukis has analysis here. Since Romney hasn't released his plan, Brown, Gale, and Looney cleverly put together the best case scenario and crunch the numbers -- and conclude they don't work.
How is that? Romney's plan has three goals. It starts by lowering tax rates by 20 percent. It then seeks to keep raising the same amount of tax revenues as it did before by removing tax expenditures, or the variety of exemptions, deductions, or credits in the tax code that function as government spending. As the wonks would say, it wants to "lower the rates and broaden the base." However, and this will be crucial, it excludes expenditures related to investment income and savings from being available for these cuts. Finally, it wants to maintain the current level of progressivity by making sure that the top one percent pays no less in taxes and everyone else pays no more. The Tax Policy Center analysis shows that it is impossible to do all three: enacting the Romney plan requires cutting taxes on the top one percent and raising them on everyone else.
In order to better understand why this is impossible we need a quick, back-of-the-envelope class and distrbutional analysis of how tax expenditures work in the United States. Tax expenditures are thought to be regressive, benefitting those with more resources. The general argument for why is because tax expenditures are closely linked with employment compensation or spending, so those who have jobs and get paid more or spend more benefit more. Being able to pay less in taxes disproporationately benefits those better off and those with the resources and ability to take advantage of often complicated tax planning. A privatized welfare state administered through these coupon-like mechanisms, compared to public ones, involve less compulsory risk-pooling and more individualized risk-bearing, which tends to benefit those who are better off.
But we can get more granular than that, and we need to in order to understand why Romney's plan fails. Let's take a quick look, using this great New York Times chart based on Tax Policy Center numbers, at who gains from different types of tax expenditures in the United States.
This chart looks at five types of tax expenditures and then at the distributional consequences for each class. Let's grab this stick by the other end and look at what sets of tax expenditures benefit three classes of people.
The first are tax expenditures that go to the working poor. These are focused on refundable credits, where almost 60 percent of them go to the bottom 40 percent of Americans. Low-income workers are, by definition, struggling to find decent wages, and these tax expenditures are meant to help boost wages at the bottom end. The big driver here is the Earned Income Tax Credit, which is a credit for low-income workers. The Tax Policy Center has this "on the table" for being able to be cut under Romney, and it is telling that the GOP hasn't said whether they want to cut it and seem to be dropping hints that they might want to go after these "lucky duckies."
The second are those that go to the middle class and upper-middle class. I don't mean middle class as the median person, but more along the lines of people whose work requires having at least some college education and who often are defined by longer-term attachment to an employer or middle management positions. These are focused on itemized deductions and tax exclusions. As seen in the chart, over 50 percent of these go to those between the 80th and 99th percentile of income. The big drivers here are two goods that are closely associated with middle-class life: health care provided by employers and a home mortgage. Both mortgage interest and employer health care spending are subsidized through tax exemptions.
And the third set are those that go to the top one percent. These are focused on special treatment for capital and dividend income. Dividends and capital income are taxed at a lower rate than wages, and as those incomes are predominately earned by the top one percent, these benefits tend mostly to benefit that group. The top 0.1 percent earn more than half of this expenditure, with the top one percent taking home a total of 75 percent of the benefit. Because of this differential, people working in certain elite financial positions often claim that their wages come from money rather than labor, and thus qualify for this exemption.
Tax policy doesn't create the conditions for each of these groups, but it helps sustain them. Making low-wage work more bearable, keeping the middle class in long-term employment relationships and making sure they are property owning members of their communities, and increasing the financialization of the economy and the explosive wealth of the top one percent all are boosted by the system the government uses to identify and collect taxes.
So with this framework in mind, what's the problem with Romney's plan? What it wants to do is lower taxes on each group and make up that difference by reducing the tax expenditures each group receives. But remember that he doesn't want to touch the tax expenditures in the third set, all the ones for savings, capital gains, and dividends, which go overwhelmingly to the top one percent. So he wants to lower taxes on the one percent, and he has to make the lost revenue up by cutting a set of tax expenditures for them that largely go to either the working poor or the middle class.
Now it is true that the top one percent benefit from exclusions as well, but this is largely a function of tax preferences for retirement savings, which the Tax Policy Center excludes from Romney's plan. The rest of the major exclusions and credits don't do very much and can't make up the shortfall given runaway inequality. Qualified retirement plans have caps on them, and health care premiums do not scale with inequality at the top end. Child tax credits are a fixed amount and don't scale at all with income. There's simply very little in this space that could be done.
So, and I'm not seeing this emphasized enough, if you are going to "lower the rates and broaden the base" for the rich, you need to actually broaden the base of the tax expenditures that the rich receive. This will be true for all of these plans going forward, and especially for Romney's. Otherwise, as the Tax Policy Center found, the exercise can't actually work.
One question I have, and I'm surprised the paper doesn't touch on, is whether the Romney plan is mathematically impossible, period. Would broadening the base on the third set of savings and investment income actually make the plan work? The 99 to 99.9 percent gain an average percent change in after-tax income of 3.5 percent, and the top 0.1 percent gain 4.4 percent, while everyone loses 1.1 percent under the Romney plan.
According to another Tax Policy Center paper, "Distributional Effects of Individual Income Tax Expenditures: An Update" by Eric Toder and Daniel Baneman (p. 7), eliminating the tax preference for capital gains and dividends would reduce after-tax income by an average of 4.5 percent for the top 1 percent. That would get Romney most of the way there, and perhaps removing exclusions for savings would make the entire plan work. However, if you lower rates while broadening the base, reducing tax expenditures brings in less money. So bringing in the tax expenditures for the top one percent may still not allow the plan to work. It's likely that these exclusions for savings and investments would be expanded, not cut, under the Ryan budget and what Romney eventually ends up doing, but it is worthwhile to see if this plan could work under any set of base broadening.
Mike Konczal is a Fellow at the Roosevelt Institute. Follow or contact the Rortybomb blog: