Yesterday we had our first NYU Tax Policy Colloquium of the semester. This is year 13 for me, shockingly enough, meaning I've been doing it for more than a quarter of my life (certainly a strange thought).
My co-convenor for the first 7 weeks is Kevin Hassett of the American Enterprise Institute. This brings a strong conservative voice of the intellectually honest genre to the table, not a bad thing at a major American law school or indeed for me. I've spent time at AEI in the past and have often considered myself more center than left (because I like redistribution but think markets are important and find political processes & centralized decision-making suspect). Then along came Bush, causing me to foam at the mouth and feel much more left. So a counter-balance is as good for me as I think it is for everyone else in the class or at the sessions.
We discussed Lily Batchelder's work on inheritance taxation, which I've blogged about in the past. But this is the first time I've seen it discussed by someone who is strongly opposed to the bottom line, which is that an optimal tax policy set of tools would include this instrument. I've suggested in past blog entries that points in favor of Lily's approach include the following:
--Including gifts and bequests received, and interacting their tax consequences with consideration of the recipient's other resources, uses more distributionally relevant information than any other alternative on the table (i.e., don't tax bequests, use an estate tax, or tax accessions without regard to the recipient's other resources). This is only an argument for having bequests affect bottom line tax liability, not for having a positive as opposed to a negative tax rate on them.
--Evidence about accidental bequests and lack of donor planning suggests that this is an area where the tax draws less of a real planning response than one might expect under standard economic models. (By real response I mean adjusting one's work and saving in response to the tax, as distinct from hiring an estate lawyer to arrange various rigmaroles.)
Kevin pushed back effectively against this view, which is not to say I always entirely agreed with him. Two of the main points raised were as follows:
1) A soundbite-style misreading of Lily's work might interpret it as follows: most bequest dollars are accidental (i.e., incompletely annuitized taxpayer died before spending everything), such bequests can efficiently be taxed at 100%, hence a very high tax rate is fine. Kevin notes that this line of reasoning would be defective. Even with incomplete annuitization and consequent accidental bequests, in a rational planning model a prospective decedent who also had some altruistic bequest motives would leave more if the residue would go to kids than if it went to the government. This is true, but I concluded the differences on this issue are semantic. Relatively inelastic accidental bequests would affect the analysis in the direction that Lily suggests.
2) Given that gratuitous transfers unfold over time, rather than simply being lateral (e.g., if I don't eat the apple this period, I make a gift of it to someone who also eats it this period), taxing bequests involves taxing returns to capital, leading to the "exploding tax rate" problem with wealth and capital income taxation over long periods generally. Perhaps I am too much of a conceptual purist in wanting to say that the lateral and inter-temporal issues are theoretically distinguishable - in practice taxing the former means taxing the latter. But there are questions of how well very long-term rational planning models capture actual human behavior. E.g., even if a low-rate annual income tax adds up over 30 years to an 80% tax wedge between consuming today and in the future, how responsive are people to this?