The first romp was a post on estate tax economics that reviewed an NBER paper by Wojciech Kopczuk. In this post I review one by Louis Kaplow, which begins with the familiar idea that every type of tax imposes some distortion: tax discourages the thing upon which it is levied. With different modes of taxation available–income tax, estate tax, value-added tax, etc.–a goal is to achieve a redistributional end with minimal distortion.
The notion of externalities plays a role in tax theory, as it does in many areas of economics. The overproduction (underproduction) of things with negative (positive) externalities is a welfare loss. Hence Pigovian taxes and subsidies can be welfare improving and the distributive consequences can be remedied through the income tax. The combination is a pure correction of the externality. The welfare gains of the correction can be distributed to make everyone better off. Notice that this combination of tax changes is revenue (and distribution) neutral. From this perspective, the revenue effects of transfer taxation–the very aspect Kopczuk thought most important–are irrelevant. (All this assumes changes to the income tax don’t impose a different set of undesirable externalities and disincentives, which probably isn’t fair to ignore.)
<Press the pause button.> In two short paragraphs I’ve summarized six pages (roughly 25%) of Kaplow’s paper. At this point I thought, “Oh, come on! Such rarefied tinkering with the tax system isn’t really possible is it?!?” Well, to Kaplow’s credit he read my mind (anti-causally!) and writes, “The foregoing discussion is obviously Panglossian, naive, or . . . choose your preferred adjective. No suggestion is made that reality operates so predictably or efficiently.” (I had to look up Panglossian. Plus, note Kaplow’s humor! Also, all quotes © 2010 by Louis Kaplow.) Nevertheless, Kaplow points to the 1986 and 1993 U.S. tax reforms that implemented changes in the spirit of what he sketched out. We’re talking theory here so let’s give Kaplow some slack. <Press play.>
Next Kaplow turns to transfer taxation. As did Kopczuk, he considers the externalities of transfers. He claims that there are positive externalities in the form of donee benefits. That is, the donor is not compensated for the benefits experienced by the donee. This argues for gift subsidization. (I did get a vague sense of double counting and Kaplow points to a paper by Peter Diamond in which that claim is made.)
On the other hand, to the extent that the donee, now wealthier by the amount of the gift, substitutes leisure for work there is a negative externality on the treasury: lost tax revenue. (I guess we are to ignore the fact that the donor will or did substitute leisure for work in order to save the sum that is gifted. So didn’t the treasury “get theirs” already? Hmm…) Kaplow goes on to describe a few other sources of negative externalities that I covered in my review of Kopczuk’s paper.
Given all the above, and particularly since tax changes can be made revenue and distribution neutral, the essential question is how to tax most efficiently (i.e. how to maximally internalize externalities), making use of all available taxation modes. In the specific case of transfers, it is possible that externalities exist at all levels of wealth (e.g. the positive externality on donees and the negative one on the treasury). Should transfers be taxed (or subsidized) at all levels?
In the final part of his paper, Kaplow discusses a few other considerations such as: transfer motives (covered in my review of Kopczuk’s paper); the fact that the most inter-generational transfers are in the form of human capital (how ought that be folded into the analysis?); the inequality of starting points brought about in large part by differences in human capital inheritance; the effect of transfer taxation on savings incentives (which can be augmented by adjusting taxation of capital gains); and charitable contributions (transfers from individuals to charitable entities outside the family are also significant).
As in my reading of Kopczuk’s paper, at the end of Kaplow’s I was left feeling like I’d missed the forest for the trees. There are clearly a lot of important considerations in transfer (or any) taxation. But they can’t be fully understood in isolation. For revenue and distributional neutrality to hold, increasing tax via mode A requires decreasing it via mode B and vice versa (more complicated still, the multiplicity of taxation modes is potentially unlimited). Given this, the absolute value and sign of externalities that arise via tax mode A don’t matter except relative to those that arise via mode B. Neither Kopczuk nor Kaplow provided in their papers a comprehensive theory within which to make a full analysis across all externalities of all modes. (This is not a critique of the papers as that’s not what they were about. I just yearn for more and will keep reading as I find relevant material.)
On the other hand, such a theory seems like it would be almost hopelessly … well … theoretical. There are practical limitations of knowledge and tax collection, not to mention an explosion of heterogeneity, that would seem to reduce the applicability of much of the theory, if it exists. To his credit, Kaplow acknowledges what I yearn for, both on the theoretical and empirical fronts.
In all, a more complete analysis of optimal taxation as a whole (that is, both income taxation and transfer taxation) would be a daunting task—one that, with regard to many of the factors highlighted here, has not really been attempted.
… There is much empirical work to be done if the many open questions are to be resolved. Moreover, the present analysis suggests that the pertinent list of empirical questions is rather different from those that have received the most attention to date.
OK then. Good start. I eagerly await the sequel.
Addendum: I confess to not having read the very long footnotes that accompany the body of Kaplow’s paper. However, I did notice that he references his book on taxation theory. See his website for details.