Friday, January 30, 2009

Tax policy colloquium with Ed Kleinbard on the JCT's tax expenditures pamphlet

Yesterday we had our third session of the year, with Ed Kleinbard (Chief of Staff at the Joint Committee on Taxation) concerning the recent JCT pamphlet(s) on tax expenditure analysis.

This is a topic I've written and thought about a lot, such as in Rethinking Tax Expenditures and Fiscal Language, 57 Tax Law Review 187 (2004) (draft version available here), reworked (and shortened) as chapter 8 of my recent book Taxes, Spending, and the U.S. Government's March Toward Bankruptcy. In brief, my main points include the following:

(1) The distinction between taxes and spending is purely formal rather than economically meaningful. Hence, a tax rule can't "really" be spending, as tax expenditure analysis posits.

(2) Since, however, people mistakenly treat it as meaningful, tax expenditure analysis can improve information by addressing labeling games, the canonical illustration of which (for me) is David Bradford's joke about making the government $50 billion smaller by replacing $50 billion of military spending with the $50 billion "weapons supplier tax credit" ("WSTC") thereby causing both taxes and spending, as conventionally measured, to drop by that amount even though absolutely nothing has changed in substance. As we discussed in the session, if you look at recent legislation, the WSTC idea looks prescient rather than like a joke.

(3) In the context of a general, distributionally motivated "tax" system such as the income tax, the way to come up with a coherent framework for improving information despite the underlying fiscal language problem is to identify allocative provisions that have been stuck into this mainly (in its rationale) distributional system. E.g., no one would think that the WSTC is an aspect of adjusting relative burdens based on ability; rather, it serves to affect economic activity, i.e., by permitting the government to acquire the specified weapons.

Following our interdisciplinary philosophy at the colloquium, Alan Auerbach was the lead commentator on this law-based paper (obviously well within his expertise, of course), whereas next week I will be the lead commentator on a very interesting econometrics paper, Amy Finkelstein's EZ-Tax: Tax Salience and Tax Rates. Ed Kleinbard was, as always, a lively, delightful, and illuminating discussant, although given his current eminence at JCT I need to treat his remarks as off the record. It did seem clear, however, that there is a relationship between how the JCT proposed to revise tax expenditure analysis (links below) and my analysis, although for institutional reasons the two must and do look significantly different. This puts me in the position of being an ungrateful whelp, so to speak, if I cavil and carp at exactly how they did it, although this of course was also my duty as a commentator.

Recent JCT publications on the subject, all worth a look, include the following:

A Reconsideration of Tax Expenditure Analysis (the piece we discussed at the colloquium)

Tax Expenditures for Healthcare
(using the new structure to illuminate the issues in a much contested area)

Estimates of Federal Tax Expenditures for Fiscal Years 2008-2012
(using the new approach in lieu of the hoary old one).


jbrose said...

Here's what I don't understand about you're saying there's not distinction between taxes and spending. Doesn't saying taxes are the same as spending presuppose a base tax and then any deviation from that is the same as spending? But what is that base? If pushed far enough, you could say that the government should tax 100% of all economic activity and then any deviation from that, even a tax rate lower than 100%, would be spending?

Daniel Shaviro said...

No need to posit a base tax. The main points are (a) hard to say what the non-government baseline is, since everything we have reflects the existence of organized society, and (b) even from the baseline, all that matters is overall net cash flows (or other value flows) in one direction or the other, not the direction of particular cash flows within the overall series.

Also no normative implications regarding size of government (e.g., that it should be big rather than small or any such thing), although the view I'm stating does tend to undermine some versions or rationales for a libertarian view. (It would not undermine, say, a Richard Epstein consequentialist endorsement of something towards the libertarian side of the spectrum.)