Yesterday’s paper was Reed Shuldiner’s Taxation of Risky Investments, another entry in the burgeoning recent legal and economic literature about the differences between income and consumption taxation, which, the literature has concluded, are significantly smaller than had previously been thought. It used to be thought that a pure income tax reaches, while a pure consumption tax exempts, all “capital income” (whatever that is). The literature now concludes that the difference between the two systems, at least under “ideal” versions of each, is restricted to the fact that the income tax reaches but the consumption tax exempts the risk-free return to waiting, which in recent decades has tended to be quite low (as in 1 to 3 percent a year). Much ado about nothing, therefore?, the literature is prone to concluding provocatively.
This conclusion has been pretty much universally accepted among law professors and economists who write about it, but ignored by most others, not just outside the academy but also, for example, when economists measure income tax distortions, typically under the assumption that a much higher interest rate than the risky rate should be assumed in assessing timing differences. The arguments are a bit too complex and elaborate to cover comfortably in a blog, but my own take on them can be found here. Suffice it to say that the skeptics, who like to think they are too hard-headed and practical to accept the new line of reasoning, in fact have to posit that up-front consumption tax refunds, if the consumption tax uses expensing , play a vital role in enabling cash-constrained investors to snap up extra risky investments. What about the fact that some consumption tax variants don’t provide up-front refunds, and some income tax variants do? And what about the fact that you can make your pre-tax investments riskier, thus reversing the effect of the tax on your risk position, by simply buying riskier items as opposed to more oft the same? “Get out of here, sonny boy,” is pretty much their reply, which is why they haven't published much on the point. (They tend to be among the older people in the field, and thus more set in their ways of thinking.)
Shuldiner, so far as I can tell, is on the fence about this, reflecting in part social affiliations. He happens to have closer relations than most of us in our shared age cohort (mid to late 40s on down) with people in the older group who are feeling bitter and passed by. The Colloquium session will have done some good, at least for him, if we persuaded him not to write the piece that they want him to write.
So clarified, the paper that I anticipate this will become should still interestingly (for those of us in the biz) flesh out some aspects of adding realism to the models that are used in generating the broader conclusions. Shuldiner also argues, persuasively I think, that "so what" is a bit too strong a conclusion however provocative, although something can matter a fair amount and yet still matter less than people once thought it did.