One of the many ways in which the Tax Policy Center's analysis of the Romney tax plan went out of its way to be fair to the candidate was its inclusion of "dynamic scoring." And when analysts used this method -- where, as Suzy explained, any economic growth caused by a tax change is considered when estimating its cost -- found that the plan would cause $53 billion in increased revenue.

That's nowhere near enough to offset its $360 billion annual cost, but still a substantial chunk of change. Given that tax revenue is expected to be 17.7 percent of GDP in 2015, the year for which these estimates apply (assuming current policy continues) and GDP is projected to be $15 trillion, this would mean the taxes caused $300 billion in economic growth, or a 2 percent increase.

That's a big number! Dynamic scoring is controversial, so it should be taken with a grain of salt, but for perspective, if the United States were growing 2 percentage points slower at the moment, we'd be in a recession. The problem is that if you make the tax plan revenue-neutral by cutting deductions and credits, that economic growth effect goes away. One of the best recent explanations of this was written by Alex Brill and Alan Viard of the American Enterprise Institute, who concluded that "base-broadening" reform of the kind Romney is proposing would not increase growth much at all. The reason: It doesn't remove the incentive to not work created by the income tax.

Brill and Viard propose a thought experiment where a country where residents spend half their money on apples and the other half on oranges. Suppose the country initially had a 40 percent tax that only applied to oranges, and then moved toward a 20 percent tax that applied to both apples and oranges. This would be revenue-neutral, and it would improve efficiency by eliminating the incentive to buy apples rather than oranges, just as eliminating real deductions for mortgage interest and health insurance would remove incentives to buy those specific goods. But it wouldn't increase one's incentive to work, because one is paying a 20 percent tax overall in either case.

Base broadening tax reform works like this if it doesn't change revenue, Brill and Viard conclude. It may be a good idea that increase economic efficiency, but it doesn't spur people to work more, and thereby prompt economic growth. Given that Romney advisor Glenn Hubbard took to the Wall Street Journal Thursday morning to reiterate that Romney would "ensure that tax reform is revenue-neutral," this is a problem for the plan.

How does the Romney team get around this? Well, for one thing, they're backing away from the revenue neutrality pledge, saying, “we are not proposing to ensure revenue neutrality solely through base broadening.” But if revenue neutrality is not solely ensured through base broadening, it'd be ensured by a combination of base broadening and spending cuts, which are also contractionary. Still, Romney adviser -- and Viard and Brill's AEI colleague -- Kevin Hassett insists that the plan would spur a 1 percent of GDP increase in growth. That translates into an increase in revenue of about $26 billion, so the vast majority of the revenue lost would have to be made up for in cuts to either spending or deduction.