Monday, March 03, 2008

Tax policy colloquium session on my tax & accounting paper

Last Thursday at the NYU Tax Policy Colloquium, we discussed my paper from last fall, "The Optimal Relationship Between Taxable Income and Financial Accounting Income: Analysis and a Proposal." Kevin Hassett did his last co-leading gig of the semester unless required to pinch-hit later on. I'm very grateful to Kevin for the great job he did throughout the semester as a very stimulating colleague and discussant. Plus it's been great to talk regularly to someone who disagrees with many in my circle on at least a few contemporary political issues. Same-mindedness and orthodoxy are the enemies of creative thought.

Kevin began the day unenthusiastic about my admittedly tentative proposal, under which publicly traded companies' taxable income would be adjusted part-way (say, 50 percent) towards an adjusted measure of the financial accounting income of the same affiliated group of companies. But in the course of the colloquy he acknowledged to moving in the direction of greater sympathy for my approach, in particular because it tries to address the downside to a full-fledged "one book" approach, which I locate primarily in legislative politics.

Some of the flavor of the discussion at the colloquium session is captured in a new subsection I added near the end of the paper, addressing particular critiques that I have heard often.

"1. Why not simply increase penalties and regulatory oversight? Doing so might be a good idea whether or not the taxable income adjustment was adopted. Moreover, insofar as it reduced the magnitude of the problems posed by tax sheltering and earnings management, it would indeed tend to weaken the case for adopting the adjustment, given the various tradeoffs presented. Nonetheless, even with optimal auditing and penalties, the adjustment would have benefits. For example, it would reduce the managerial incentive to waste resources engaging in transactions that are legally permissible, and thus that would survive heightened scrutiny, and yet that serve no good social purpose beyond advancing the managers’ income manipulation goals. Examples include creating hybrid financial instruments that are debt for tax but not accounting purposes, and engaging in tax shelter transactions that have just enough economic substance to withstand IRS review.

"2. Why not instead directly improve the systems’ income definitions? This as well would be independently desirable, and might reduce the social gain from adopting the adjustment. Nonetheless, it would still leave room for the adjustment to improve matters. Any plausible rules for defining taxable and accounting income are likely to leave room for the exercise of interpretive discretion, which managers would be expected to use in a self-interested fashion to reduce the former measure and increase the latter one. This problem can only be addressed via the relationship between the measures.

"Consider again the case of the tax shelter transaction that has just enough economic substance to withstand IRS review. Such cases may exist even with optimally designed economic substance rules, given the tradeoffs that underlie choosing the proper level of stringency. Accordingly, the taxable income adjustment, which would reduce the tax benefit from engaging in such a transaction, is not simply or even primarily a substitute for directly seeking improvement in the income definitions used by either system.

"3. How can going halfway towards a one-book system be a good idea, if going all the way is not? The experience of countries such as Germany that have moved away from one-book systems may support the inference (with which I agree) that adopting a one-book system in the United States would be a mistake. Why move halfway towards something not worth doing in full?

"The core reason, in my view, for avoiding a predominantly one-book system (even with specified exceptions, such as for foreign subsidiaries) is that it would put the U.S. Congress more directly in the business of defining financial accounting income. My proposal is designed to minimize this danger, and concentrate the incentive effects on corporate managers rather than on politicians whom it would not succeed in reining in. Insofar as the proposal would nonetheless result in increased legislative meddling in the definition of financial accounting income, the case for adopting it would be weakened."

I have now sent the paper to a bunch of leading student-edited law reviews and am hoping for the best.

No comments: