Why We Shouldn’t Tax Grad Students

A Tuition Waiver is Just a Coupon

Section 1204 of the new tax bill would impose tax on most tuition benefits provided to graduate students (by repealing section 117(d) of the existing tax code). A number of otherwise smart people seem to be confused about what’s going on here (see, e.g., Tyler Cowen, who apparently doesn’t realize he’s just argued for deducting all education expenses). It’s just about price discrimination. Let me try to explain.

An income tax system does not generally tax coupons. I bought 2 pairs of pants on Saturday at 30% off, I don’t plan to report the savings on my 1040, and I’m not expecting an audit (though I’m sure my snazzy pants would impress the auditor). Why not?

Because coupons don’t offer any evidence about my ability to pay taxes. Efficiency and fairness both suggest we want to impose the same utility burden on every tax payer — everyone should, at the margin, experience the same loss when paying their last dollar of tax. We can’t observe utility, but we can see what resources each taxpayer has available to support the government — the resources they consume, plus the resources they save. Further, we usually can only tax observable resources — measurable evidence of the consumption value taxpayers get from their stuff.

Coupons and other forms of price discrimination (like, say, getting the $1 ticket to ride the Bolt Bus, or the $39 “wanna get away” fare on Southwest) don’t provide evidence of any new resources for the person who made the bargain purchase. The best evidence we have for how much value an item has for a particular person is how much they paid. You could try to insist that I would have paid $30 to go to New York in an ugly blue vehicle crammed full of millennials, but how can you know for sure?

So, too, the tuition waiver. Schools offer discounts to some people in order to get them to attend. Presumably, that person demanded a lower price because they placed a lower value on the education that was on offer. Perhaps some consumers have consumer surplus as a result, but you can’t run a tax system trying to prove that.

Things may be different if a school is also employing me or one of my relations. In that case, the grant of the scholarship probably replaces a portion of the cash compensation I would have otherwise received. Although it’s difficult, we have to attempt to tax most of the value of in-kind compensation, otherwise too much of the tax base would slip away through a net of loopholes. For many taxpayers, section 117(d) is a somewhat unprincipled exception to that rule: it allows universities to give tuition breaks to their employees and the employees’ kids (via the cross-reference to section 132(h)). Let’s just say that Catholic schools in particular make good use of that provision.

But graduate students are a lot more like Bolt Bus riders. Sure, they may teach a class or help out in a lab. But the value of these services is often considerably less than the value of their tuition. Further, the graduate student almost certainly would not have been working for the university in that capacity if she were not trying to earn a degree. So there is no concern about salary slipping through the tax net.

I’m fine with repealing most of 117(d), but taxing graduate students is confused, or ideological, tax policy.