So, according to the Tax Policy Center, we start out with an $86 billion hole in Romney's tax plan. But the Tax Policy Center's own calculations show that that $86 billion hole can, in fact, be filled without raising middle class taxes.

TPC's study assumes that pro-growth tax reform cannot produce any economic growth. TPC acknowledges that, according to an economic model created by Harvard professors Greg Mankiw and Matthew Weinzerl that assumes tax reform will produce economic growth, "the tax cuts would result in revenue reductions of $307 billion (instead of $360 billion)." In other words, economic growth could fill $53 billion of that $86 billion hole.

That still leaves us $33 billion short. But TPC also acknowledges that its study assumed that Romney would not touch "the exclusion of interest on state and local bonds and the exclusion of inside-buildup on life insurance vehicles." According to TPC, eliminating these exclusions could raise $45 billion in revenue.

So economic growth ($53 billion) plus nixing these two exclusions ($45 billion) equals $98 billion. That's $12 billion more than the $86 billion needed to prevent a middle class tax hike.

William G. Gale, co-director of the Tax Policy Center and one of the authors of the study on Romney's tax plan, told me this morning that under these two assumptions Romney's tax plan would maintain revenue neutrality without raising middle class taxes. "Under those assumptions and policies it would be revenue neutral," Gale wrote in an email, "but remember the tax expenditures are eliminated from the top down and that is not administratively feasible. So you also have to assume tax expenditures are eliminated in an infeasible manner to avoid the tax increase on households with income below $200,000."

The first aspect that stands out is that Governor Romney has yet to detail what tax expenditures he’d repeal, but TPC assumes that many items are either “on the table” or “off the table.” While some of these assumptions make a lot of sense, others make less.

For example, a couple of items that TPC assumes are off the table are the exclusion of interest on tax exempt bonds and the exclusion of interest on life insurance savings. While Governor Romney has professed a desire to keep rates on savings and investment low, maintaining these exclusions is not necessarily what he meant. In fact, both of these exclusions largely benefit the wealthy, and, according to the Treasury Department, added together their repeal would net upwards of $90 billion that could be redistributed to lower-income individuals. That would go a long way towards balancing out the supposed $86 billion windfall for the rich and tax hike on the middle class and poor, and it could make the impossible suddenly possible.

While my analysis of those exclusions is nothing if not rough, TPC could do a much better job with the use of their model. So, my dream study would show all of the different combinations of items that could be included to make a revenue-neutral tax reform distributionally neutral rather than placing items either on or off the table and then claiming, given those assumptions, that distributional neutrality is impossible.

More base-broadeners are on the table. TPC originally claimed that the Romney plan would raise taxes on the middle class by $86 billion. After a critique by AEI colleague Matt Jensen, who pointed out additional opportunities for base-broadening, TPC downgraded its estimate considerably. Specifically, Jensen pointed out that the exclusion of interest on state and local bonds and the exclusion of inside buildup on life-insurance products could yield more revenue. TPC then acknowledged that repeal of these provisions would raise approximately $45 billion from high-income taxpayers, reducing any need to tax the middle class by the same amount. The result: A purported $86 billion tax increase on the middle class shrinks to $41 billion.

Incorrectly Ruling Out “Step-Up”

The authors also made a considerable error in ruling one policy out of bounds. They assume that the Romney plan would not change the step-up in basis of capital gains at death. This critical error significantly biases the report’s results.

All proper tax reform plans should eliminate the death tax, as the Romney plan does. Under current law, heirs of taxable estates inherit assets after paying the death tax. When they sell those assets later, they calculate their capital gains tax by subtracting the asset’s value at the time they inherited it rather than the price that the original owner paid. This is known as “step-up” basis. This makes sense because the heir already paid the death tax on the asset. Using the initial owner’s basis would make the double taxation of the death tax even worse than it already is.

However, keeping step-up in place after eliminating the death tax would mean that some of the value of inherited assets escapes taxation. There is no economic justification for keeping it in place after death tax repeal. Plans that properly repeal the death tax typically employ a system, after repeal, whereby heirs pay capital gains tax when they sell inherited assets using as the basis the price paid by the person from whom they inherited the asset. This is called “carry-over” basis.

Different Assumptions, Different Results

The authors assume that Governor Romney’s plan would raise taxes on middle-income and low-income taxpayers by $86 billion, which would result from closing certain tax preferences on these taxpayers. However, if they had instead eliminated certain tax preferences that tilt toward high incomes that they originally ruled off the table, such as the exclusions of interest on life insurance savings and municipal bond interest, a minimum of $45 billion of the $86 billion would fall on high-income taxpayers, according to their follow-up analysis.[8]

The Office of Management and Budget estimates that step-up will reduce tax revenue by $24 billion in 2013.[9] Assuming that assets inherited after death follow the same distribution as long-term capital gains as reported by the IRS in 2009, 78 percent of that $24 billion, or $19 billion, would count as offsets for incomes over $200,000.[10]

In total, fixing the step-up error and putting the exclusions of interest on life insurance savings and municipal bond interest back on the table would shift at least $64 billion of the $86 billion tax increase (about 75 percent) that the report assumes would fall on middle-income and low-income taxpayers to high-income taxpayers.

In addition, this amount is decidedly conservative. Strong evidence suggests that the authors significantly underestimated the $45 billion that would be transferred up from middle-income and low-income taxpayers by eliminating the exclusions of interest on life insurance savings and municipal bond interest.[11] Furthermore, the $19 billion tax reduction for step-up in 2013 would likely be even higher in 2015, the year the report uses for its analysis.

Even assuming the $45 billion estimate is accurate and assuming step-up would not grow in value, there are several ways to close that remaining gap of $22 billion. For example, choosing tax preferences other than the ones the authors chose and combining them with other policy changes that they did not consider in their analysis could close the gap. These could include phasing out personal exemptions for high-income taxpayers or capping their itemized deductions. Some of these options might not be sound policy, but they would easily raise enough revenue from taxpayers earning more than $200,000 to close the $22 billion gap. In fact, President Obama’s fiscal year 2013 budget includes a cap on itemized deductions for incomes over $250,000 that would raise considerably more than $22 billion.[12]

Making these changes in the authors’ assumptions would undo their headline-grabbing conclusion that Governor Romney’s tax reform plan would cut taxes on the rich and raise taxes on middle-income and low-income taxpayers. Instead, with these changes in place, their analysis would show that the Romney plan makes growth-promoting policy changes in a revenue-neutral manner and does not raise taxes on middle-income and low-income taxpayers.