New light on old battles

Last Thursday at the NYU Tax Policy Colloquium, our guests were Joe Bankman and David Weisbach, presenting their recent paper on income vs. consumption taxation, available here under the March 2 date. Profuse apologies for getting the title backwards on the cover sheet.

Bankman and Weisbach push a pro-consumption tax line that I have also pushed in print. The basic idea we share is that a consumption tax can be just as progressive as an income tax, while being more efficient and perhaps also doing a better job of ranking people (on a lifetime basis) in terms of how well-off they are.

One of the big points of contention concerns the burden of the consumption tax on future consumption. Income tax advocates commonly complain that, say, Bill Gates and his heirs won't pay consumption tax on their huge fortune until it is actually all spent on consumption, if it ever is. Consumption tax advocates reply that, assuming a perpetual fixed-rate tax, the wealth is already bearing the burden and has merely gotten to defer payment at a market rate of interest (meaning that the deferral does not reduce the present value of the liability). No difference than if wealthy people paid more tax today but did enough extra borrowing to fund the cash flow. And hard to deny if you accept, as most do, that a consumption tax is neutral as to when one consumes, thus supporting the conclusion that wealthy people aren't paying less, in present value terms, merely because they defer spending their wealth. In any event, this is the argument that I and others (such as Bankman and Weisbach) have made.

In the course of the discussion, largely due to Alan Auerbach's efforts as Colloquium co-convenor and lead discussant for the session, it became clear how uncontroversial all this would be if people in fact invariably spent all their wealth before dying. E.g., if Bill Gates actually were guaranteed to spend his entire fortune before heading to that virtual cyber-place in the sky, then his deferring payment at a market interest rate would be accepted by nearly everyone, I think, as good enough to support viewing the unpaid tax as being borne by his wealth today. But of course Gates is not going to consume it all. Instead, at least if he didn't plan substantial charitable bequests, it presumably would all be going to his children.

Since this is what makes income tax advocates cavil at the equivalence claims, it seems clear that the treatment of bequests is at the heart of the income vs. consumption tax debate, even though I and others have been accustomed to describing the debates as wholly separate. Inherited wealth is at the heart of the dispute, even though one could have an income tax or a consumption tax with or without an estate or inheritance tax. (The latter would be paid by recipients of bequests, with the rate structure depending on how much one got rather than on the size of the overall estate. The optical reason for this is that it combats calling the thing the "death tax." The substantive reason is that, if inheritance of concentrated fortunes is the concern, the tax should depend on the degree of concentration that persists.)

Let's back up here a second. How could one possibly support an inheritance tax if one favors consumption taxation? They're often thought inconsistent, because taxing bequests implies taxing saving given that it happens over time.

But in fact there is a separate thread here. Henry Simons, in his "Personal Income Taxation" book, famously urged that gifts be double-taxed (non-deductible by the donor, but included by the donee). In doing so, he was addressing the definition of the consumption piece of the income tax base. He argued that it's clearly consumption by the donor, who does it voluntarily in lieu of spending the money, say, on restaurants or vacations, while at the same time financing current or future consumption by the donee (who indeed is better-off than one who had to render services to get cash). And Simons was clearly correct, in terms of how we might most logically think of the gift as affecting the welfare of the two parties.

Why not double-tax gifts, then? Leaving aside administrative problems (especially when we think of all the gratuituous transfers of services inside a household), the best argument against, developed at some length in work by Louis Kaplow, is that there's an altruistic externality we might not want to discourage, from the fact that a gift dollar in effect purchases $2 of consumption value (by the donor and the donee).

Let's cycle back to the consumption tax. Bequests other than accidental ones (i.e., those left without donative motives because the decedent didn't live long enough to use it all up) "should" be included in the consumption tax base, as consumption by the decedent, if we are looking just at how well-off the decedent is compared to people with the same budget line who simply had different consumption preferences. Then we would also tax the heir, like any other donee, on spending the bequest on consumption. Again, this would seemingly be the right rule in the absence of the altruistic externality. All this leaves unsolved the question of whether taxing gifts just once is the right response to the altruistic externality, as opposed to being, e.g., either too big or too small a benefit (perhaps more likely the latter).

So the treatment of bequests and other gratuitous transfers raises issues that are separate from the income vs. consumption tax question of taxing saving.

Why would we have an estate or inheritance tax, on top of having decided that generally taxing gratuitous transfers once is the way to go? The argument is presumably one of negative externalities to bequests, which worsen the relative position of non-recipients. Again, whatever one thinks of this, it's distinct from the income vs. consumption tax debate.

Final point trying to tie all this together: if we want to tax gratuitous transfers at once, taxing the bequest at death as consumption but then giving some sort of credit to the donee against future income or consumption tax liability as to the amount received would address the multi-generational timing point that income tax advocates hold against the consumption tax. And it would not increase the tax burden on saving or bequests if we could make the tax present value-equivalent to deferring it (via the offset against future tax liability). An issue worth exploring?