Mitt Romney's tax plan is mathematically possible -- but only if the rich get richer at a level we have never seen before
Mitt Romney's tax plan is a logic puzzle. The details barely exist, but there are just enough of them to infer what the nonexistent details would be if they did exist. Think of it like the LSAT, just with more numbers. Pick up your number two pencils, and let's see what we can figure out.
II. Eliminate the Alternative Minimum Tax (AMT) and the estate tax
III. Close enough loopholes to make tax reform revenue neutral
IV. Maintain rates on savings and investment and eliminate them altogether for the middle class
V. Keep the mortgage-interest, healthcare, and charitable giving deductions for the middle class
VI. Have high-income earners will pay the same share of overall taxes that they do now
VII. Not raise taxes on middle-income taxpayers
The nonpartisan Tax Policy Center (TPC) has a head start on us. They looked at the first four conditions above -- Romney only laid out the others later -- and concluded that the numbers don't add up for 2015. There aren't enough tax expenditures for the rich to pay for the tax cuts for the rich. The result is a net tax cut for high-earners to the tune of $86 billion -- meaning taxes would have to go up by $86 billion on everybody making less than $200,000 for the plan to be revenue neutral.
That's a bummer. But is the Romney plan really unsalvageable? That depends on four big assumptions. First, what does Romney mean by middle class? Second, what taxes is Romney talking about when he talks about preserving rates on "savings and investment"? Third, how does Romney's corporate tax plan factor in? And finally, how much economic growth should we project? These assumptions are worth real money. Romney's annual revenue hole is either as small as $41 billion or as large as $144 billion depending on our answers here. Let's consider them in turn, and then see what we can piece together.
1. Who's middle class, exactly?
Former Reagan adviser and Harvard professor Marty Feldstein claims TPC got it wrong -- that Romney's tax math works without requiring a middle class tax hike. Feldstein argues that cutting tax expenditures for households making $100,000 or more would pay for their tax cuts. This is incorrect. Brad DeLong points out that there isn't enough money in those expenditures to pay for those cuts. But there's a bigger issue. Feldstein claims that Romney's plan would work by closing loopholes for households making between $100,000 and $200,000, but Romney defines those households as middle class. Feldstein inadvertently corroborates TPC's conclusion -- Romney's tax plan does require a middle class tax hike to work.
2. What's savings and investment?
TPC assumed that Romney would not change the tax treatment of savings and investment when he said he would not change the tax treatment of saving and investment. But maybe he will! Some conservatives have said Romney might consider ending the tax-exempt status of municipal bonds and inside-buildup of life insurance contracts. Even if that's true -- which is just speculation -- that wouldn't fill Romney's revenue hole. TPC analyzed these potential changes, and calculated that Romney's plan would still cut taxes for the rich by $41 billion.
3. What about corporate taxes?
Romney wants to overhaul our corporate tax system in two steps. The first step is cutting the tax rate from 35 to 25 percent, preserving recently added research credits and expensing provisions, and enacting a repatriation holiday. The second step involves lowering rates further, and moving to a territorial system -- meaning overseas corporate profits would not be subject to U.S. tax. Romney would pay for this second change by closing corporate loopholes, but he would not pay for the first change. TPC assumed both parts would be paid for, so it didn't look at this in its analysis -- but if it had, this unfunded change would have made Romney's revenue shortfall $96 billion worse. Thanks to this handy chart from the Congressional Budget Office that shows which income groups bear corporate income tax liability, we can estimate that 60 percent of this $96 billion would go to households making $200,000 or more. That's another $58 billion in cuts for the rich that needs to be offset.
4. What about growth?
Even under TPC's aggressive growth assumptions, Romney's plan was mathematically challenged. This wasn't a case of TPC being too timid with dynamic scoring -- it got its dynamic scoring numbers from Romney adviser Greg Mankiw. Not that we should expect revenue neutral tax reform to catalyze much growth. A 2011 paper by Alan Viard and Alex Brill of the conservative American Enterprise Institute concluded that a broader tax base would negate most of the supply-side effects of lower marginal rates in revenue neutral tax reform. In other words, people's incentives don't change when their taxes don't change even if their tax rates change.
Still got your number two pencils out? Now we're ready to tackle this logic game. Romney wants to cut rates and cut loopholes but keep everybody's taxes the same. That's the implication of a revenue neutral plan where the rich pay the same share and the middle class pay the same amount. It's just a complicated way of saying nobody's tax bills change. But we're back to the same old problem: the rich pay a lower effective federal tax rate under Romney's plan, so they won't pay the same share. Unless they have more money than we've assumed.
But there is one way that Romney's plan works mathematically: Income inequality explodes. If enough growth goes to the top 5% of earners, they will get rich enough to fill the revenue hole. How much richer would they have to get?
That depends on the size of the hole. There are four basic scenarios here. The shortfall could be $41 billion if Romney ends the special treatment of municipal bonds and life insurance buildups and we ignore his corporate tax plan. It could be $86 billion if Romney preserves the special treatment of municipal bonds and life insurance buildups and we ignore his corporate tax plan. It could be $99 billion if we take the first scenario and add the $58 billion of corporate income tax cuts for the rich. And it could be $144 billion if we take the second scenario and add the $58 billino of corporate income tax cuts for the rich. The chart below looks at how much richer the rich would have be -- compared to the TPC 2015 baseline -- for Romney's plan to add up under each of these scenarios. The answer: between 3.2 and 11.3 percent.
(Note: These changes are relative to how much TPC projects the top 5 percent will earn in 2015).
A lot of assumptions went into these calculations, so let's lay them out. First, I assumed that Romney would not raise or lower taxes on anyone making under $200,000. In other words, he would close just enough loopholes to pay for the 20 percent marginal cuts and $38 billion of corporate tax incidence for the non-rich. This would mean that any revenue hole in Romney's plan comes from the rich. Next, I assumed that the top 5 percent grow pari passu -- that is, households making $200,000 to $500,000 grow at the same rate as households making $500,000 to $1,000,000 and at the same rate as households making $1,000,000 and up. Then I reverse engineered the effective tax rates the rich pay under Romney's plan -- along with the original $86 billion revenue shortfall TPC found -- using the 2015 income levels from this TPC distributional table and the data in Tables 1 and 3 of TPC's analysis of the Romney plan. Finally, I divided the revenue hole in each of the above cases by the weighted effective tax rate the rich pay to figure out roughly how much more they would have to take home to make the numbers work. These assumptions are obviously not all true, but they are close enough to give us a reasonable answer to our question.
That answer is more inequality than we have seen before. The proof is in the Gini coefficients. Those measure inequality on a scale of zero to one. A rating of zero indicates perfect equality where everybody shares all the income, and nobody else makes more than anybody else; a rating of one indicates perfect inequality, where one person has all the income, and nobody else makes anything else. We already have the most unequal society of any rich nation, and TPC's 2015 projections imply it will only get worse. Even if the Bush tax cuts expire, our post-tax Gini coefficient will rise to 0.531 from 0.45 in 2007. That would increase to 0.544 under Romney's tax plan, and as much as 0.557 in the $144 billion shortfall case. It's the difference between us merely having Rwandan levels of inequality and having Bolivian levels of inequality. For comparison's sake, remember that Denmark and Japan are the world's most equal societies with 0.25 Gini coefficients.
The chart below looks at post-tax Gini coefficients for each of the 2015 tax scenarios. The only question is how much our republic is getting banana-ized.
(Note: Thanks to Michael Linden of the Center for American Progress for helping me calculate these Gini coefficients).
There's one word you've probably noticed again and again throughout this piece: assume. That's what we have to do again and again when it comes to Romney's tax plan. The details are mostly not there, but there are just enough of them to deduce some of the rest.
The upshot is this: Romney's tax plan does not work under remotely plausible growth projections. It either increases middle class taxes or increases the deficit. If Romney is serious about doing neither, then he has to be unserious about his growth projections. The rich have to get almost impossibly rich to make up for the lost revenue in Romney's tax plan. Realistically, their incomes would need to be 7.7 to 11.3 percent higher than TPC predicts -- that is, we should not ignore the corporate income tax cuts. To put that in perspective, that's between $377 and $548 billion additional dollars flowing to the top 5 percent of households.
Romney may not like this, but that just means he does not like his own tax plan. These numbers are the inescapable conclusion of a plan that relies on a giant magic asterisk to add up.