Some new studies by the Tax Policy Center show, once again, that Mitt Romney's tax plan is mathematically impossible. The plan, you may recall, is this relative to the current policy baseline in which the Bush tax cuts are fully extended:
— Cut rates 20 percent across the board from the new baseline.
— Make up the lost revenue by reducing deductions and loopholes but not the tax preferences for investment income.
— Do so without raising taxes on the middle class.
The verbal argument that this is impossible is, I think, clearer than the study based argument but I'll get to the latter. But in words, if you want to change the tax code something has to change about the tax code. You can either change the amount of revenue it brings in, or you can shift around who bears the burden. But if your goal is to neither shift the burden nor alter the distribution, then you can't change the tax code. It's Romney's Trilemma—our of base-broadening reform, revenue neutrality, and distributional neutrality you have to pick two. Romney argues in the debates by switching between which two he's picking.
For his latest stunt, Romney has floated the idea that instead of ending specific deductions he would just generically cap deductions at $25,000 a year. That's a politically clever way to do it, and since the cap is high it's a very progressive way of raising the revenue. But Romney's Trilemma still bites. This base-broadening, distributionally progressive reform doesn't raise enough revenue for Romney's proposed across-the-board rate cuts. The Tax Policy Center says it gets you about $1.3 trillion over ten years when Romney's rate costs would cost $5 trillion. And that's $5 trillion over and above the price of fully extending the Bush tax cuts.