Tax policy colloquium, week 2: Michael Simkovic's The Knowledge Tax

Yesterday, we discussed the above paper, which builds on Mike's co-authored empirical work on the economic value of a law degree, along with other empirical work that Mike and others have done suggesting that higher education degrees (college and up) offer highly favorable pre-tax (and to a lesser degree, after-tax) rates of return, even taking into account of opportunity costs (foregone earnings while one is in school).

The underlying research reaches two main conclusions.  The first is that lifetime earnings are sufficiently higher for those who get higher education degrees (despite the time away from being more than part-time in the workforce) to offer an above-market rate of return on tuition, etcetera.  The  comparison here is to investment returns on different forms of capital other than human capital.  The second is that a causal arrow runs from higher education to higher earnings - in other words, that it's not just selection bias, as in the case where those motivation and abilities will tend to produce higher earnings in any event also happen to pursue higher education.

Both of these conclusions, perhaps the second in particular, are controversial in the literature. Based admittedly on just superficial inquiry, Simkovic's work on these issues strikes me as plausible and well-reasoned, but I admittedly don't know enough to form a definite opinion.

In reading The Knowledge Tax for our session, I thought it reasonable to accept the empirical conclusions for argument's sake (which again, is not to suggest skepticism about them), because the main issue presented by the paper concerns their further implications. The paper argues that part, though not all, of the higher pre-tax return associated with higher education reflects that human capital investment is treated less favorably by the U.S. fiscal system (and in particular, though not exclusively, the U.S. federal income tax) than other investment. It thus argues that more favorable treatment of higher education would increase tax neutrality, economic efficiency, and long-term growth, due not only to the standard efficiency arguments for tax neutrality between particular alternatives, but also positive externalities.

There is of course a difference between devising "neutral" rules, at some particular margin, and those that are aimed at inducing particular behavioral responses, but if the rules are currently biased against something with positive externalities, then up to a point the two modes of analysis may have some tendency to travel together.

I thought the paper made a good case that we should mainly think of higher education expenses as investment, rather than as consumption (an old debate, of course).  But it's harder to draw firm conclusions about the current degrees of relative bias.  Obviously, the question of how favorably the U.S. federal income tax system actually treats non-human capital investment of varying kinds is quite complicated.  And one tax benefit of pursuing higher education, in lieu of working currently, is that the opportunity cost isn't taxable.

E.g., say I could have earned $50,000 this year, but instead I earn zero and rely on loans, savings, or family resources to spend an additional $50,000 going to law school.  I get implicit expensing of the $50,000 foregone earnings, even though the law degree may have future value that goes forward for decades.  This is highly favorable treatment from an economic standpoint.  On the other hand, I never get cost recovery for the $50,000 of tuition, even though it may reasonably be viewed as a cost of generating earnings.  There is of course no general answer to the question of how alternative investment choices would have been treated by the tax system, given the crazy quilt of possibilities.  So figuring out what's treated better or worse than what is quite tricky.

It's also fair game to ask how we think people who are considering higher education actually decide, and to what extent they are focusing on the long run, and sophisticated aspects of it such as the tax treatment of alternative types of future income.  As I discuss in my recent paper on behavioral economics and retirement saving, people often act as if they are myopic, even if that is not exactly the internal mental process. Now, when one decides to pursue higher education, evidently there is some sort of departure from the blinkered, short-term focus of the classic myope.  But still, even if people are taking account (or act as if they are taking account) of long-term earnings potential, it is possible that some aspects of the myopic or as-if-myopic frameworks will continue to influence them.  And this is potentially relevant to how they respond to reasonably expected after-tax earnings under alternative choices, and to tax rules that would change the overall treatment (especially down the road).

There are also important institutional issues to think about, e.g., does the education sector respond to demand (e.g., given that much of it is nonprofit), and how do relevant labor markets function - e.g., those for lawyers and doctors.

So it's a rich topic, to which the paper makes an interesting (if inevitably inconclusive) contribution.