What is the incidence of a tax on tuition waivers?

Here is some basic info, in 2011-2012 145,000 graduate students received tuition waivers.  Monday I suggested such a tax is a bad idea, but who would bear the burden?  Let’s say there are three parties, the universities, the graduate students, and third-party funders who support research and graduate students.  Those third parties may be for instance Harvard donors or the National Science Foundation.

The short-run, first-order effect is that the grad students pay tax on their waivers and fewer of them pursue postgraduate studies.  And if grad students are dead set on attending no matter what, they bear a relatively high burden of the tax.

That said, there is more to the story.  Universities seek to attract graduate students for multiple reasons, with two possible options being “to enhance their prestige” or “to boost revenue,” or some mix of the two.  It will matter.

To make up for (some of) the tax, and maintain the flow of students, universities will opt for some mix of lowering their tuition and increasing stipends and increasing non-taxed forms of aid, such as quality of office space or teaching opportunities for grad students.  If universities seek to boost their prestige, they will be quite keen to keep up their “Q,” and not eager to lower Q, even with higher P as recompense on the revenue side.  In that case a relatively high share of the burden will fall on universities.

In contrast, if universities pursue revenue, they are more willing to live with a lower Q if accompanied by a higher P.  More of the burden will fall on students, because the accompanying enrollment-maintaining compensations from the universities will be accordingly lower.

I don’t know of a paper estimating the effects of taxing student fellowships, an innovation from the Reagan tax reforms of 1986.  Can any of you lend a hand here?  It didn’t seem to much slow the growth of graduate education as far as I can tell, so perhaps the burden there was born by universities.

Now enter the third parties.  Donors might give more funds to universities to help make up for taxed tuition waivers.  If you are a Harvard alum, for instance, you might wish to see Harvard carry on its great traditions with yet another generation of Ph.d economists who initially received tuition waivers.  In other words, you want prestige as an alum and that requires keeping up the flow of Q, number of quality students, through the program.  Donors will give more resources to the universities, or to the students (through other vehicles), to help make up for the new tax.  In words, to the extent the donors covet prestige, more of the tax will fall on them.  This is a tax on prestige-seeking!

My intuition is that the schools with a strong donor base will put in much more effort to raise money for graduate students, and they will meet with a fair degree of success.  (Note that Harvard’s now-bigger fundraising campaign will to some extent distract the attention of the president and other senior leaders from other programmatic activities at Harvard; in the longer run that could harm Harvard stakeholders.)  But schools below the top tier don’t so much have this option, so they will decline in resources and status relative to the very top schools.  This is p classic case of how imposing new burdens leads to higher market concentration and cementing in the status of the elites, in this case the educational elites.

Throughout, I am assuming the universities cannot evade the tax outright, for instance by relabeling the categories of tuition and tuition waiver to avoid the bite altogether.  But that is another possible equilibrium, if the details of the law so allow.


So, still ignoring the scholar athletes? Or do the products of the NCAA system use another name when they do not pay to play for a well known university football or basketball program?

"So, still ignoring the scholar athletes?"

Did you read somewhere that athlete tuition waivers would be treated differently? That's news to me.

Two quick thoughts on this proposal:

1. Universities could be somewhat constrained in lowering their tuition as a tax offset strategy, especially if they are bound to relatively fixed university-wide rates (e.g. public institutions on a state-set tuition formula) or if doing so would create a large gap between grad tuition per credit hour and the standard undergrad rates that aren't affected by the tax (and I suspect very few universities would be willing to drastically slash their ugrad tuition rates just to save a few grad students some money on taxes)

2. If the bulk of the tax's incidence could be placed on the student, it might also be thought of as less of a revenue device and more of a disincentive for seekers of low-value PhDs. It isn't a very popular subject to raise among persons who are close to its source, but there seems to be a massive glut of new PhDs issued in several academic disciplines (especially the humanities) relative to the jobs in these same areas. One way to reduce that glut is to raise the price of getting a PhD.

To this latter end, if the tax's incidence fell primarily upon the student then the removal of the credit would likely affect peripheral PhD students first - i.e. the type of student in who spends 7-10 years finishing a poetry PhD from a weaker program with very little in the way of job prospects. That may not be an entirely bad thing.

1) I'm not aware of a university that does not already have different rates for undergrad, masters, and PhD. While I'm sure there might be a few, it seems to be relatively common practice to have different rates based on level. Additionally, there are other ways used to raise the tuition costs already used, such as, classes that are four credits instead of three, in or out of state, or lab fees.

Then there are all the other fees that get mysteriously added to pay for peoples' pet projects. The athletic department convinced the board to add a $10 per credit hour and cut tutoring services to pay for returfing the grass inside of the track a few years ago at my local CC. The cost for the field that was never used was over a million. This behavior is considered industry best practices by university administrations. The universities will take care of a tax change to students just fine, as they've demonstrated a high degree of skill when it comes to manipulating all the relevant parties.

I'm wholly confident that the average university administration will be able to use the tax change to simultaneously raise more funds from the donors, get more out of the students, cut services, and raise admin salaries. In other words, business as usual.

"The short-run, first-order effect is that the grad students pay tax on their waivers and fewer of them pursue postgraduate studies. And if grad students are dead set on attending no matter what, they bear a relatively high burden of the tax."

Is this bad?

On your first point, is this tax only on the fellowship that covers tuition costs? If so, what's to prevent an Ivy League school from just eliminating tuition all together, since that's basically what they do now. I didn't know anyone at a top school who actually paid tuition. My sense is that they put a price on it for fundraising purposes. It's a way to tell donors that educating the top PhD students in the world costs $X per semester, so give us more money to help cover those costs.

> disincentive for seekers of low-value PhD

This tax plan is brilliant.

Less education, more government debt, and a transfer to your pocket?

Yet another dope who conflates attending a college with education.

Writes a commenter who undoubtedly earned an engineering license without actually bothering to go to a 4 year college - https://en.wikipedia.org/wiki/Regulation_and_licensure_in_engineering#United_States

Or who practices medicine without having put with any of that pesky medical college fooforaw.

The option to earn an PE without a 4 year ABET accredited engineering degree in the US no longer exists. It was phased out several years ago.

It is a pretty extreme position that college generates and transfers no knowledge.

But lets keep our eye on the ball, this isn't about students paying their fair share, this is about dividing the loot.

In related news, "Republicans against colleges and universities" is a real and growing thing.


Given this analysis, it’s hard to see why one would think such a tax is a bad idea.

Too bad it ain't net-net revenue enhancing. It is in service of benefiting other groups which are now more highly valued under the law, but while still failing to improve the debt to GDP ratio.

In the best of all possible worlds this would be a revenue-neutral or slightly revenue-positive tax reform

In such a situation we would be looking at how the change in incident tax was a transfer from who to whom. Perhaps we would say that graduate students deserve a break versus workers or vice versa. But what is even going on here? Leaving aside the great loss of revenue, where does this transfer go?

From graduate students to hedge fund investors? Or merely from students to established families?

Do you believe that the mortgage deduction is primarily a benefit to homeowners or banks?
Do you believe that the tax advantage to PhD students is primarily a benefit to students or universities?

I thought I made clear that I viewed this as a shift in burden. Certainly it is a shift to graduate students in a straightforward way.

Meanwhile, what? 145,000 students aren't going to make up $1.5 trillion in new deficit spending, so ..

Paraphrasing your position:

A student decides to pursue a PhD, and the tax system shifts the burden to the guy at McDonalds. The public agrees because 'education good'.

My position:

A student decides to pursue a PhD, and the Universities claim that they are giving free education while treating them as indentured servants. The public agrees because 'education good'.


When a bill is proposed that says that in-kind education benefits should be taxed, you see burden to student; I see burden to universities.

It isn't straight forward since the PhD students are effectively working for the university at extremely low rates doing both research and teaching. The universities could just be honest and say that they have adjunct positions with research that require on the job training. At the end of five years the adjuncts will have the option to test for a full time qualification.

Whether or not this is going to make up for a deficit isn't relevant, at least to me. When was the last serious attempt for debt/deficit solutions? Clinton doesn't count since his solution was to raid SS.

"A student decides to pursue a PhD, and the tax system shifts the burden to the guy at McDonalds. The public agrees because ‘education good’."

That is not really true, is it? A student is taxed lightly, and a full time adult at McDonald's might even be earning EITC.

The bulk of this transfer is to higher earners.

Fine. Substitute McDonalds with the employer of your choice. The point, which your avoiding, still stands. The burden is being shifted to universities, which incidentally are high earners, not students.

The students will literally pay this tax on their Form 1040 (or newfangled post card).

I understand the academic argument that since it comes out of wages it is "really" paid by the employer, but I also see the handwaving involved.

Tell me, when you did your taxes last April did you say "no worries, my employer is paying all this not me!"

And I do find it odd that I am a lonely voice for fiscal responsibility.

I swear, during the Obama years there were more of us.

Now you just deserve mockery. We are talking about taxing on supposed free doctoral education not giving away more benefits funded by the taxpayer.

This is fiscal responsibility, and the exact opposite of Obama.

Nope. You are falling for the Bright Shiny Object and ignoring the principle (stated!) purpose of this tax bill.

I mean it is pretty obvious. Students, especially grad students, are a bunch of liberals. Ding them, because they are just a political out-group (like Californians), as you do the one thing this is really supposed to do:

Cut taxes by raising federal debt, and buy votes for the next election.

BTW, if you want to know why "fiscal responsibility" can include "counter-cyclical stimulus" watch the new MR video.

The assumption seems to be that tuition waivers are a form of disguised compensation for services provided by the grad students. But aren't tuition waivers also used to attract the most promising grad students - and undergraduate students as well. How does one distinguish the two? Anytime a vendor (the university being a vendor of education services) charges different prices for the same products or services, there's an implicit quid pro quo. I've commented that elite colleges use "scholarships" to attract the best undergraduate students, but they aren't really scholarships, they are in effect partial tuition waivers; hence, elite colleges are like the airlines, with everyone on board paying a different price for the same product or service. Just as the airlines charge different prices to attract different customers, so too do the elite colleges. Substituting a "scholarship" for a tuition waiver elevates form over substance (and has implications like those mentioned by Cowen in his blog post). Of course, tax professionals are adept at avoiding taxes by changing the form of the transaction. Thus, tech companies can put a patent in a file drawer in a tax haven, apportion lots of income to the patent, and avoid U.S. taxes. It may be a sham, but members of Congress treat the tax avoiders as heroes. Would Congress treat colleges and their students as heroes if they were to change the form of tuition waivers? I drifted away from tax practice in part because so much of it is form over substance (that and in the 1980s tax planning was mostly about the time value of money (i.e., tax deferral as opposed to tax avoidance) when interest rates were so high). We live in a time and place where all too often reality is whatever one wishes it to be, where tuition waivers become scholarships, where real news and fake news are indistinguishable, where asset prices will rise forever. We are only fooling ourselves.

They might have to do something really radical like, say, pay more valuable grad students more. God, I almost sprained something trying to come up with that idea.

Principles matter. Either ALL in-kind income should be taxed, or none of it (although there is no real reason it should skate). Of course waivers should be taxed - as should employer-provided health insurance and other benefits. To hear the post-grads, especially doctoral candidates, talk, they are already impoverished so the waivers are unlikely to move them into taxable territory, and minimally so for those who are affected.

Tax-free benefits are in fact a subsidy from those who don't enjoy them to those who do. The same people who whine incessantly about "fairness" also squeal loudest when asked to pay their own fair share with no special treatment. Those who cry loudest for "democracy" guard most jealously the privileges they enjoy which would never pass on a stand-alone vote. Hypocrisy should be its own reward.

Should your credit card rewards be taxed?

As far as I know, they are not a fringe benefit of employment or a substitute for salary. It is simply the reduction of a service fee that is available to anyone whether they are employed by the credit card company or not.

"Income" is not actually dependent on "employment."

As rayward says, once we get into price reduction as income a lot of things could start looking funny.

Are you one of those people who think that the 10 percent extra on that box of cornflakes is actually free?

To be clear, I am 99% concerned about $1.5 trillion in new debt as a result of "tax reform." Tax reform should take debt the other way, especially in a period of high profits, high employment, etc

I am 1% concerned that going after grad students is a chickenshit way to balance an awful plan.

"so the waivers are unlikely to move them into taxable territory". No, that's not right. The standard stipend of a PhD student in a program that pay stipends is around $20k to $30k a year, perhaps a little more for certain program at the richest university. At this level of income you're basically not taxable. But the "official" tuition, waived in these program, is often $40k to $50k. If this becomes taxable, the grad student will have a taxable income of $60k to $80k. For a single filler, that puts you well in a taxable territory -- with a federal tax not far from $10k.

I believe that in the top tier universities, university will pay the cost, because there is already serious competition among programs to attract good graduate students. The university which are cash-constrained will thus have to admit fewer grad students.It is possible that certain programs fall below "the critical mass" and are terminated.

Indeed. If the tuition is taxed at high end universities then the take home salary might be $10,000 after taxes. Now if that same student was in a STEM field then the choice would be to get hired at a tech or biotech firm immediately...or live on $10K a year for another 4 years. The economic choice is pretty clear. It may be the STEM courses at many Universities that are hit hardest.

I can see Universities rapidly losing their graduate programs, which means fewer RA's for research, and fewer TA's to help bear the teaching burden in large classes. You may be right that MIT, CalTech, and Ivy League schools will find a way. But State Universities may not. And that is where whole programs may disappear.

This might also apply to Resident Assistants in the dorms that receive room/board as their principle compensation (they might get a small additional stipend). R&B is now about $18,000/year so the stipend might just about cover the tax.

If there are fixes this will most likely result in an increase in tuition for the majority of students, or increases in dorm fees.

Government subsidies for university education ought to be focused on courses that produce graduates that are of positive benefit to a country's infrastructure, such as engineering, science subjects or medicine. "Fun" subjects such as art, sports and media studies should be at students' own expense.

Agreed. I'm surprised there is so little debate on this point.

Most PhD programs should be cut.

Maybe we are looking at the wrong actor who receives the benefit of tuition waiver.

It is the faculty member who is the beneficiary. If there were fewer graduate students, the department wouldn't grow. So, the department has the incentives to acquire more foreign graduate students to keep the faculty busy. And, they are cheap labor for doing research.

How many departments would wither if tuition waivers were not granted?

I am in favor of giving scholarships or financial support to students. Domestic students. We are subsidizing the education of foreign graduate students to keep the departments busy.

Think you have that backwards. Foreign grad students arrive as top performers after ~22 years of investment from their home country, and many stay after graduating. But, international means competing globally, so the massive tax increase of the tax plan would make the US less competitive. Outside of the top programs, STEM grad student slots already go wanting from lack of qualified applicants, not displacement of domestic students.

So we get to subsidize them.

Lack of domestic applicants means we have overcapacity.

Which means we are doing for the faculty.

This is being done because the GOP has decided that universities are against them, so why not punish them by taxing endowments and tuition waivers. If it was just about taxation fairness, then why not tax free airline rides given to families of airline employees, or other similar untaxed services?

This post, and the comments, are all from the point of view of elite schools and STEM (+economics, which thinks it is really a STEM field). Universities give tuition waivers across fields through university or provost fellowships or for programs like creative writing that would hardly exist without waivers. The difference between being taxed on a $15-20,000 stipend (which most often requires teaching undergrads) and $40-50,000 (stipend plus tuition) is likely to be prohibitive for many students. Pure principles are fun to espouse, not to mention calling people hypocrites and whiners, but we do not live in, nor will live in, a world where all in-kind benefits are taxed. I have a suggestion: once you have nailed down taxation of employer-provided health benefits, you can move on to taxing graduate students who cannot pay. After the mortgage-interest deduction is eliminated.

The theme here is "sure I did nothing to deserve my tax reduction, but OTHER PEOPLE should still pay more!"

"Not giving me special treatment is special treatment for you."

The whole proposition here is a payout for Trump voters, and to preserve Republican power in 2018.

I am old enough to remember Republicans saying that running up debt to buy votes was a bad thing.

"The Republican tax plan would add more than $2 trillion to United States debt in 10 years and only boost gross domestic product by up to 0.83 percent, according to a Penn Wharton Budget Model (PWBM) analysis."

Even worse numbers.

That means it would add about $2 trillion to GDP per year after 10 years. I wouldn't vote for it, but it's not that bad

Here is the link to the full report. As I read it, the 0.83 percent is cumulative.

"TCJA raises GDP in 2027 between 0.33% and 0.83% relative to its projected value in 2027 with no policy change."


Apologies for not posting the link the first time.

You are a worthy foe!

Making it harder for graduate schools to price discriminate seems like a good idea. Treating the income from teaching, working as a research assistant, etc as taxable income also seems like a cleaner approach.

Of course we don't tax senior citizens for the "income" that they receive from senior discounts that some companies give. Nor do we tax people for the discount on pharmaceuticals that their insurance company may negotiate. Not because the government wouldn't like the additional income, but because collecting the tax would be difficult and might just end the practice. Plus it would appear that society accepts these discounts.

But the more colleges price discriminate, with increasing list prices, the less sympathetic the general population becomes. The more colleges shroud pricing in convoluted rules and discounts for favored groups, the less sympathy the general public has for the winners of this "game."

Reduce that value of the waiver, reduce the list price, and you reduce the tax. Then pay the graduate students for the services they provide, with higher salaries for more gifted candidates. This would still leave wealthy schools with the ability to offer higher salaries to gifted students if they choose. Life is often unfair.

Medical students would seem to be an especially difficult area. It is usually much more expensive to teach medical students. Tuition probably doesn't reflect the true cost to the school of running the program. Worse, drawing the line between education and services provided by the externs can be difficult. But then again medical schools, in general, grant fewer tuition waivers. (Their are some MD-PhD programs or agreements where you agree to work after graduation, military etc. Of course your total earnings from the MD-Phd, even after scholarships, can be lower.)

Medical residents do pay taxes and social security on their earnings. (If they work over 40 hours per week, I think.) And different programs pay different salaries. This isn't considered outrageous.

If the tax on tuition waivers encourages colleges to decrease price discrimination, on balance, I favor it.

Grad students are already taxed on teaching income and fellowship income. This is taxing a discount, effectively.

The universities will bear almost 100% of the incidence. What's overlooked us that most of the tuition waved grad students are participants in huge profit centers. Theyre the core of the TAs that make the 500 person mandatory online classes run.

Nominally they are "taught" by a professor who doesnt do anything but upload a syllabus amd a few pre recorded lectures. In reality all the grading and admin is done by a TA running a 60 person session. With no facilities to use this is a huge moneymaker.

You're literally talking about $100,000 in revenue per grad student per semester. Having the university comp $2k on a W2 is nothing.

One thing to consider in this is what tuition actually financing, in at least some graduate programs. After year one, many students take no courses per se. Instead they enroll in "Dissertation Research" hours. In the first year of graduate school, for example, the "cost" of my tuition and stipend was paid by a federal program to finance STEM PhD's. Later, the "cost" of my tuition and stipend was paid by my mentor's grants. In this context, it is worth considering how much tuition reflects the genuine cost of educating students and how much of the tuition price includes a large transfer of monies to universities and/or researchers. Indeed, even as a postdoc, I am still enrolled in "classes" every semester, for which my university charges tuition, which is paid on my behalf by the state. These classes include such things as "Lab Research" and a host of classes which would normally be considered new employee orientation.

From the outside looking in, that's insane. Worse even than our health care system (and presumably for the same reasons).

OTOH, it does suggest that the net effect will be modest - just different numbers in various, internal-only spreadsheets.

One response could be that elite institutions simply raise the stipend to cover the tax. Of course the stipend will have to rise 33% more than the amount of the tax, because the increase will itself be taxed, and the student is now in the 25% bracket. In STEM fields, where the cost of graduate students is largely covered through research grants from NSF/NIH/industry, the result will be fewer grad students per research dollar, and less research accomplished. So the cost of this tax will be born by society through reduced research and innovation, with all of its multiplier effects. You could achieve the same results more efficiently by simply cutting the budget for NSF and NIH. Is that really what we want to do? We already spend considerably less of our GDP on government supported R&D than our major economic competitors.

Universities seek to attract graduate students for multiple reasons, with two possible options being “to enhance their prestige” or “to boost revenue,” or some mix of the two. It will matter.

Another reason is to have cheap bodies to teach large introductory classes. This reduces faculty workload, obviously. I suppose, indirectly, it reduces expense.

I don't think it's possible to talk about this when we have no idea of the value of these waivers. None. Yes, there is a nominal price set by the school, but that's all it is. It's a "Manufacturer's Suggested List Price," not an actual market price. Suppose tuition waivers were rare or non-existent. What would the market price of graduate education be? That, I think, is where to start.

How is a tuition waiver even "income"?

It never appears in your bank account.

This is an incredibly bad idea.

The same way loan forgiveness is I imagine.

Income is more than just money, it is any sort of compensation.

Loan forgiveness does appear in your bank account, or did when you took out the loan.

I think tax incidence is enormously important by few voters even think about it. Politicians like Bernie Sanders use that effectively to fool the voters.

There should also be a clientele effect. The taxation is more painful to students in higher marginal tax brackets (eg, with outside income or working spouses). So fewer of these students replaced by students in lower marginal tax brackets. Also, perhaps some redistribution among states. States will likely conform to federal tax treatment, so there may be higher demand in lower-tax states.

In top-tier graduate research programs, all domestic research grads have waivers. In the fields I think you're discussing when mentioning Harvard and the NSF, a domestic student paying for their education is doing it wrong.

Restoring the status quo of an untaxed waiver is an internal accounting exercise of making graduate tuition in the research program zero, controlling access through the admissions program, and assessing the cost of a researcher's classroom training directly as a cost against the PI's budget.

Student researchers already lost in practice whatever independence this reaccounting costs them in theory.

Question for anyone who's read the plan in detail. The plan changes the deductibility of tuition reductions that are given in exchange for TA/RA work but not those which are given as pure scholarships.

Is there any restriction on doing the following, where the current nominal situation is 40k in tuition reduction and 25k in stipends that come with RA/TA requirements and let's say that the student has to work 8 hours a week for three 10 week semesters?

40k tuition reduction is given as a no strings attached scholarship that a student receives only on the basis of maintaining academic good standing.

25k stipend remains conditional on RA/TA work. That still leaves the hourly rate for the RA/TA work at around $100/hr which should be more than sufficient incentive to keep students doing the gruntwork.

My read is that 25k stipend would be taxable (as it is now) and that the 40k tuition reduction would be an untaxed scholarship.

Am I reading the proposed changes wrong?

I predict the incidence will fall almost entirely on the universities.

As others have noted, the tuition waiver is closer to an accounting fiction than an actual transfer of goods or services. Grad students are employees of the university, and don't receive classroom instruction. Their pay is already marginal, and there's not a lot of room to cut without dramatically changing the cost-benefit equation.

If the accounting fiction ends up causing a real tax consequence, I predict that the Universities will simply change the name they give to the "waiver."

Every person who has written about this assumes that all universities use this waver system; they don't. I have been paying tax on my "fringe benefit" income - that is, "research credit hours" - for three years now. While the school I did my master's at did use the waver method, my current, very large, university does not. And yet, somehow, the grad school here hasn't shut down...

Comments for this post are closed