Yesterday, the Tax Policy Center published a paper with “FAQs and Responses” pertaining to their recent study on Governor Romney’s tax proposal. This new paper is quite useful and helpful, and the policy community should be grateful that they wrote it. However, as blemishes are apt to stand out up close, in a few ways this document highlights my misgivings regarding their initial study.
One of the document’s most useful contributions is the authors’ new analysis of the tax exclusions for municipal bond interest and interest on life insurance. In a recent blog post, I pointed out that these two exclusions would likely be “on the table” in a Romney reform package—and the Wall Street Journal has recently verified this assertion by asking Glenn Hubbard, one of Romney’s top economic advisors. TPC, however, had assumed they were “off the table.” I went on to point out that eliminating these two exclusions could bring in “upwards of $90 billion,” much of which would come from those earning more than $200,000 a year, and I requested that TPC do a more thorough analysis.
To my great delight, they did. They found the value of these two exclusions to be $49 billion, $45 billion of which could be redistributed downwards. These numbers were derived from careful analysis that relied on assumptions consistent with those elsewhere in their study. The difference between the $49 billion that they find and my $90 billion, however, is primarily due to differences between those assumptions and mine: They assume that a rate-cutting corporate tax reform would be enacted simultaneously (eliminating some interaction effects), and that the revenue from eliminating the exclusions for corporations would not be included in the individual reform. For the sake of providing a rough upward bound, which seemed appropriate when discussing whether something is “mathematically possible,” I assumed neither.
TPC’s assumptions are reasonable, though, and accepting them brings us within $41 billion of distributional neutrality. That is awfully close.
First of all, there would be growth effects; maybe not by 2015, but they would come. As my colleagues Alex Brill and Alan Viard recently noted:
"Governor Romney’s tax plan for individuals would lower statutory tax rates on ordinary income while leaving tax breaks for saving largely untouched. His corporate tax reform plan [which TPC is assuming] would further improve the allocation of capital and foster economic efficiency. Overall, the governor’s plan translates into a reduced tax burden on savings and investment, which are key drivers of long-run growth."
Growth would contribute more revenue to the government, thereby making base broadening easier. What’s more, there would be feedback effects at the individual level with individuals supplying more labor, making more investments, and thereby increasing their own income. TPC envisions some feedback effects, but possibly far fewer than other economists might. Growth and individual feedback effects could make up much if not all of the distance to distributional neutrality.
Secondly, any complicated tax model like TPC’s has a margin of error. Hundreds if not thousands of assumptions must be made, and each time you make one you accept the hopelessness of capturing reality. (Sometime try making the imputations necessary to calculate the distributional effect of repealing tax exclusions for employer-provided health care.) I’d be surprised if $41 billion is outside of the margin of error of TPC’s model.
Now, it’s worth noting that enacting the reform that TPC has modeled would be difficult both practically and politically. Practically, it would be difficult to eliminate tax expenditure from the top down–first for the rich, then for the middle class and poor. But policy tools that that have gained support on both the right and left–such as the caps for tax expenditures proposed by President Obama and Martin Feldstein–could be modified to roughly achieve this effect by setting different caps for different income levels. Politically, it would mean eliminating cherished tax expenditures. But Romney’s selection of Paul Ryan as running mate indicates to me that he is willing to discuss with the public–and work hard to convince them to support–policies that might at first be unpopular.
In the end, I think it might have been more constructive for the Tax Policy Center to write a paper showing how, and under what assumptions, Romney’s plan could work rather than showing how, under their limited assumptions, it doesn’t. But I am glad TPC continues to refine the analysis and work to explain their thinking. Tax policy is a complicated and imprecise science and the better the stock of analysis, the better the policy making process. Of course, the political process is something else entirely. Here’s to hoping that a campaign touting these results as “proof” of anything will start to back down now.
Matthew Jensen is a Research Associate at the American Enterprise Institute.