Should we let graduate students in private universities form unions?

That is the topic of my latest Bloomberg column and my answer is no, here is one excerpt:

One core reason to have unions is to boost the real wages of needy workers. But graduate students are not employees in the traditional sense. They are receiving training, often on very favorable terms. Typically a university is investing large sums of money to make those students employable and successful, usually on the academic market; the University of Chicago says it invests more than $500,000 per doctoral student. If those students demanded and received higher wages for their teaching, the university would not necessarily increase its investment in them at all; it could simply reallocate existing funds. Thus it is misleading to think there is a real bargaining situation here.

Think of a university as an investor in these students, and toward that end it must choose between boosting their academic quality through better training, or paying them higher stipends and teaching wages to ease their immediate financial concerns. The incentive for the university, which cares about its broader and longer-term reputation, is to invest in the quality of those students but pay them smaller amounts (though enough to live on). In contrast, the incentive for a graduate student union would be to push for higher wages, given that the other university investments are less visible and hard to monitor.

At the margin, society is better off if the focus is on the training, which enhances productivity in the long term, rather than on higher wages and stipends for students in the short term.


In general, when considering this issue, ask yourself a question: When it comes to bringing about change, do today’s universities have too many veto points or too few?

Some researchers have pointed out that graduate student unions don’t seem to have harmed the public universities that allow them (such unions, which are permissible in many states, would not be affected by the federal government’s decision). The evidence may be compelling in the short run, but the real costs are likely to come later — by slowing down or even preventing beneficial changes to the U.S. system of higher education. Furthermore, state labor laws dramatically limit what public employees can negotiate for. Unionized graduate students at private universities unions would not face similar restrictions.

Recommended, do read the whole thing.


Worst of all, it could take money out of the budget that's needed for professors' salaries.

We see trade unions playing important role for advancing / defending liberalism in places like Algeria and Sudan. I don't like their anti-liberty stance on school choice and I don't trust them to back immigration (see AMA for a crumby example), but I do think on net they are a salutary political / civil-social force

why people keep using failing countries to suport their points of view ?(failing for reason diferents of the discussion topic)
why not denmark, german, sweeden, spain, portugal, etc with a higher live standard. Countries with unions (and yes, with unions in university and schools).
Algeria and Sudan? really?, why not Yemen or Siria o Lybia, or guatemana, nicaragua, venezuela, there are plenty of failing countries for diferent reasons.

Not to mention Solidarnosc in Poland, which was instrumental in freeing that country from Soviet tyranny. Odd how an allegedly Marxist government tried strenuously to ban a trade union...

I suppose it’s because trade unions are democratic and Marxism isn’t?

Especially in the humanities and social scientists, there are too many Ph.d students receiving too many Ph.d's, and that makes job market a nightmare. The real exploitation happens when grad students turn into adjuncts who get paid absolute crap. At first I thought that higher wages for grad students would incentivize even more people to go to grad school. But actually, the more expensive it is to employ a grad student, the fewer spots the university will offer. So maybe that's good! Cheap labor is one reason why schools take on more grad students than they can possibly place.

Right, we have an overproduction of elites problem, so reducing the number of marginal Ph.D. candidates who have little chance of ever making tenure due to having to pay the best candidates more money would seem pretty sensible.

It's not an overproduction of elites, it's departments that know their grads have poor job prospects but overselling them to naive students in order to gain cheap labor and critical mass for fighting academic turf battles.

We require nutritional facts labelling from those selling popcorn, universities should be required to disclose the expected income at graduation and 10 years later, along with the expected debt load and repayments.

Phd students arent exactly known for their naivety.

Kinda depends on the discipline, I'd say.

22-year-olds generally are known for naivete about finances and their own chances of success. That's why some of them sign up for overpriced credit cards, and some of them sign up for grad school.

Savvy comes from spending your whole life in school?

Wit. A human adult who willingly chooses an English PhD program is an elite?

Most humanities and social sciences PhDs are a form of entertainment. People who do these courses could have been making decent money working as plumbers or grocery story managers. They choose to do these PhDs because they like it not because it provides decent job prospects.

Otherwise, supply of grad students would dry up and employment prospects would improve. Grad students have rational expectations and they are aware they are not going to make money. They just to love to complain.

I'm not in favor of telling people what they can and can't do. Framing the issue in terms of what we "let" them do is exceptionally misguided.

I agree.

Is it a good idea for grad students to form a union? Probably not.

Should it be illegal? No. Forming a union should be a basic right under the freedom of association. That doesn't mean that universities should be required to negotiate with that union, or legally prevented from taking actions like expelling students that go on strike...

As I understand it, we have laws that provide for unions --- granting them exemptions from anti-trust, preventing employers from firing people for organizing unions, etc. So, the question is whether those union and labor laws should apply to graduate students. Union laws are already government intervention to infringe freedom of association and "telling people what they can and can't do".

There's just an enormous amount of labor law that goes well beyond freedom of association. Such as freedom to not associate, for example, and freedom not to be "represented" by a third-party organization if one does not wish to be. And, of course, a right not to pay agency fees for such representation.

Should all graduate students in the bargaining unit be compelled to accept the union contract (which covers not only wages and other compensation, but also work rules)?

Labor law evolved within the context of industrial employment, where work rules, seniority, and job classifications are at least comprehensible. Is it unreasonable to ask whether this model (and especially its compulsive features) is a good fit here?

Indeed, they're adults.

People sometimes do mistakes, but better have the freedom to do mistakes instead of being perfect in a cage.

Should grad students be "allowed" to organize to negotiate with a university? I don't understand a pro-freedom argument to say no, especially in the context of supporting businesses' ability to aggregate resources in nearly all cases.

Should students be allowed to organize to negotiate with a university and bind all other grad students to pay into and benefit from that negotiation? This is a harder question. One reason society might allow this in general is a belief that employers of all sorts can earn excessive rents and there is a societal imperative to mitigate this. I'm sympathetic to this argument in graduate education. Another reason may be to redress a perceived market failure. That many graduate students make negative-expected-value decisions to enter certain fields is sad, but hard to argue as a market failure.

The article and discussion would be clearer if we better distinguish between these two scenarios.

Would a graduate student union of any sort be beneficial to its members, to universities, and/or to society at large? If you think so, go spend more time with graduate students.

Does Tyler really think that other employers don't train their workers and invest in them? He's arguing against unions writ large.

'He's arguing against unions writ large.'

For almost 4 decades, at this point. You honestly were unaware of one of the major donor selling points of the GMU econ dept and law school, along with several private policy institutes that tend to use those professors to confer a veneer of academic respectability to political aims?

Well, now you know. It also reveals anyone talking about Prof. Cowen or Prof. Tabarrok as lefties that support unions to be abysmally ignorant.

Not that someone like Prof. Cowen cares - the more people who think he is a centrist, a reasonable man only interested in a dispassionate process of taking steps to make the world a much better place, the more people who will not look critically at what is actually being espoused, or ignore it completely.

Really, the purpose of unions is to boost the wages of needy workers? How about, it's to secure for the people who actually produce value their fair share of the profits? Sheesh.

Actually, the purpose of a union is to boost the wages and privileged of members of the unions, at the expense of non-members.

You know that an industry is a rent seeking one when Unions can be considered as an option. In a competitive industry companies are just trying to survive and if they pay their employees more than their competitors will go out of business, and if less then they lose their talented employees to other companies and also go out of business. Most employees realize this, so don't organize much in the private sector anymore. Of course there is friction and momentum which hides the immediate and short term effects of this (many people make mistakes based on short term impacts - a company can well afford a few years of underpaying it's employees for instance, especially if they are large firm in a small town).

That doesn't mean of course that there is not competition for jobs in protected industry, in fact the appeal of these jobs means there is probably more competition than if it is not protected. So if graduate students do form Unions, they may find they attract more higher quality candidates from elsewhere, which may work for the initial phase of students (since they are already on the gravy train) but will be to the detriment of the following lot. Indeed we may end up with rent exhaustion, where thousands of students engage in extended wasteful competition for just a few roles as rent seekers. As rent seeking positions other people are able to argue that places should be shared according to other metrics other than merit. For instance according to racial categories; after-all if the position is rent seeking should society share rent equally in all categories? So Asian students are going to find they will have a lot less graduate student positions.

What happened to this once underrated blog?😑 First the Baumol debacle and now promoting illiberalism?

'and my answer is no'

The only reason to read the column would be if that answer was anything other than the banally predictable.

1. Let graduate students choose if they want higher wages or better training?

2. Maybe they'll use the union to bargain for better training?

3. I'm deeply sceptical of 500,000 figure.

Still, in practice I'm probably with Tyler because the unions would no doubt be dominated by the whiny and end up more about student activism.

Re. D’s 3. I’m also skeptical. Someone who believes this needs to explain the economic incentives. TC says universities are concerned about their reputation. This doesn’t pass the smell test. If an institution wants to improve its reputation by paying out a half million dollars per student, only to attract more half-million dollar students, they would quickly go bankrupt. I believe the alleged $500,000 is more than recouped by the elimination of full-time faculty salaries and benefits that cheap graduate labor enables. That grad students earn a living wage is just a lie. Aren’t grad students supposed to be doing research? And doesn’t full-time faculty shun teaching because it interferes with research? And what about the undergraduates being taught by students? Does that help the university’s reputation?

That $500,000 number is based on the (inflated) sticker price of coursework.

The same logic applies to student athletes and whether they should be paid and allowed to unionize. Graduate students and athletes are no more university employees than girl scouts are employees when they sell cookies. Girl scouts are beneficiaries/customers, not employees, of Girl Scouts organizations. Girl scouts sell cookies to defray the out-of-pocket costs that they would otherwise have to pay to participate in scout activities. Similarly, RA and TA-ships are just ways to decrease graduate students' out-of-pocket costs of their education. Finally, most college athletes would not be able to afford the lavish facilities and professional caliber coaches if they had to pay for those resources themselves. Fortunately for those athletes, some past administrator was able to figure out how to pay for that stuff through television rights, ticket sales, and merchandising. On the flip side, the benefits consumed by scouts, students, and student-athletes also should be exempt from taxes, which I think for the most part are. The one exception might be that graduate students stipends are subject to income tax (though not Social Security tax?).

Child labor laws should not apply to girl scouts nor should labor and union laws apply to graduate students nor student-athletes.

How many big time college football teams are there? It looks like there are 130 now, up from 117 last I checked maybe a decade ago. If colleges had to pay players, say, $50k per season, what would the number fall to? 64 teams? 32? Would that be so bad?

Poor Nick Saban might have to scrape by on a mere seven figure salary in that scenario. What kind of monster would suggest such a thing?

It would be bad for the student athletes that could have played for the 65-100 teams that no longer would exist when the field got cut from 130 to 32-64.

How many fewer drunken coeds would be raped by football players if the number of college football teams fell to what the market would bear?

The emotional attachment to college football is peculiar. The players are recruited (essentially hired). Except at the smallest schools (and perhaps the service academies), they don't share classes, housing, much of campus life, etc. with the real students. They are not, in any real sense, actually representative of the schools or students.

My school hired better players than your school seems a weak reed to carry much weight.

An economics professor forgets about offer and demand when discussing salaries.

Universities can offer such bad contract terms to PhD students because there are lots of students willing to sacrifice their present income for future gains. Of course, there's a great difference between willing and being able to sacrifice the present income. Graduate students from relatively well-to-do backgrounds are willing and able to forgo income. Students from lower income families are just screwed up.

Also look at the international composition of graduate students. What would happen if the Chinese (or others) were a bit less willing to work for low salaries? Would universities offer better contract terms?

Continuing with the offer/demand of PhD students, compare the average salaries of PhD students between the US and some countries in Europe. US: 30K, France: 29K, Germany: 40K, Switzerland: 50K.,11_IN226_KO12,23.htm

Job conditions for Europe: PhD is around 4 years and teaching is optional. PhD labor is scarcer compared to the US because there are so many good paying jobs that few young people want to spend 4 more years with a relatively low salary.

Good points:
1. Politics. Yeah, unions could and probably will be more involved in West Bank issues than they should. Solution is not to oppose unions, it is to carve-out politics in this particular case. "Financial support" argument is not particularly strong one (easy parallel with ordinary taxes, you pay taxes in Student Union that opposes Facultistan and Administraria and does some pointless red tape because politics), but politically motivated exclusion of students from such unions could be (exclusion from graduate education would be even worse).
2. Unions will oppose good things that are contrary to the interests of current union members. Sadly every powerful organisation does that, it is somewhat a further question of what powers unions should and should not have.
Good point probably should have been summed not as classic "unions are great but not in MY backyard", but as "graduate school is non-generic industry, it deserves non-generic union law, here's why and how".
Bad points:
It is either "universities should admit more people" (and train them, and pay salaries) or "universities will cut costs on quality if they are obliged to pay more elsewhere". You cannot hold both positions simultaneously.
Evidence of well-to-do students as if it was random sample. If you make admittance conditional on being a millionaire, your students (if any) will all be millionaires even before you've started to teach them. Wouldn't it be splendid?
Veto points. Do students already have veto points? Then reform them so that they go through the sensible union procedure instead of what mess is it there now. If they don't, union will be in favor of eliminating veto points at the administration and faculty, because it opposes their interests.
Good point to raise, but more discussion is needed somewhere else:
How fast should we move and how many things to break along the way in education? One example where I would say we need to look before we jump is online education. Nobody knows what equitable property rights arrangement should be. But universities have all resources they need and educator have way fewer (if you take non-english educators, power difference is even larger). My university (just for example and, therefore, nameless) pushes total ownership for any course that was produced with its resources. They might have a point (and certainly the power), but in that "long run" it will hurt them. If they were compelled (by a union) to pay royalty on those courses, short-term they would have produced less, but long-term incentives to develop courses and not to just to sit on the accumulated stock would be higher. \\Alas, graduate education is special in that it has less courses and those courses are less pliable to online-isation.

I did not attend grad school but I did apply for a few moonshots (at elite programs, where I was not accepted so I chose law school instead). What I do know from that experience is that elite law schools accept far more students than elite grad schools accept (in economics, which is where I applied), many times more. I assumed that training a grad student was more intensive than training a law student. The two may not be comparable, so my comparison my prove nothing, but if my assumption is correct, then a grad student is not like an employee and more like an apprentice.

Even though I was not accepted at one of the few elite economics programs where I applied, I was interviewed by several, an experience one does not have when applying to a law school. Again, I assumed that I was interviewed because the investment in each grad student is so great that it was essential that the program choose applicants who would be the right fit for the program. An aside, I chose the few elite economics programs where I applied based on one or two faculty members who I wanted to train under. Law school isn't like that at all. Sure, a law student might become a professor's assistant, but that is rare and the experience is not like being trained by an economist. Also, I do see that my analysis might be viewed as reason for treating the grad students as employees (apprentices are more like employees than "students") and allowing them to form unions, but again it's the enormous investment in the grad students that distinguishes them from employees.

In the trades an apprentice in the union gets both training and pay. It takes four to five years of training before becoming a Journeyman. I don't see how this would be any different. I was talking to an electrician just this week, and he was telling me how the union was moving away from allowing people to enter directly as an apprentice since too many quit the program in the first year (and the union wasted those resources training them) so the make them do a year as a laborer first.

I work as the "ta-coordinator" for a political science department at a large public research university. There is no doubt in my mind that the ta union reduces the quality of teaching. Every minute spent on teaching must be accounted for -- e.g. time spent in lectures, grading, emails, office hours, etc. In practice, there is never enough time to do a competent job at grading-- for example, grad students will typically have 15 minutes to grade a ten page paper which is (possibly) enough time to evaluate it, but definitely not enough time to offer detailed constructive criticism. The union contributes to the idea that the real purpose of universities is to over-produce research, and that teaching should be treated as an unfortunate necessity. On the other hand: if a grad student is falsely accused of improprieties, they are assumed to be guilty by everyone except the union...

In many fields, grad school is not significantly differentiated from experience that could be obtained in private industry. And yet private industry provides that on-the-job training with reasonable pay.

Unions have been under political attack for quite some time now. Those attacks have been largely successful and the number of unionized workers has consistently dropped over the last 40 years.

How has diminished union power helped ordinary Americans?

Less organized crime, for one thing.

Move all of higher ed to a free market system. Let educators sell their education, let students purchase the best education packages they can.

"Think of a university as an investor in these students" ... "the University of Chicago says it invests more than $500,000 per doctoral student"

The University of Chicago is a government institution. If they are investing $500k per grad student, that's ultimately government money they are investing.

In addition to government money there are undergraduate tuition and philanthropic gifts.
And the $500K figure would only be approached by counting the tuition the university does not actually charge the grad students it supports as teaching and research assistants.

Tyler's done got his! Doan u be thinkin' bout takin' from his stack! He SLAP you up!

In situations like this, I find Isaiah Berlin’s “two concepts of liberty” to be quite helpful.

“Should we let...” And there you are. The closer libertarian principles come to eroding one’s extreme privilege, the less applicable the principles. TC, it’s not up to you do decide for other people what is best for them.

Was about to post the exact same thing. What a strange thing for a “libertarian” to say.

Would the answer be the same for postdocs? Arguably they are still being mentored but are required to be far more independent and don't by default always get much from the institution outside the specific faculty member's group (if they have a formal mentor at all).

I remember working in a science lab as an undergraduate and there was this post doc working in the lab. Nice guy but his experiments didn't work out and he was constantly called out in group meetings by the professor because his experiments failed. And then one day, poof! The post doc disappeared and I overheard that he had been let go. That's it. Career ruined. He may as well be flipping burgers at McDonald's.

I was considering perusing graduate school in the sciences but after that incident, I thought better of it and became a medical doctor instead. There's an oversupply of doctorate graduates so professors have complete power over them. A PhD is for the birds.

You didn't mention my favorite argument for the union:

If there's excessive entry into PhDs, because people are overconfident or have otherwise misspecified beliefs about the value of graduate school, or because graduate schools actively mislead them, then raising entry barriers via a union could be good!

This is embarrassing: "It must choose between boosting their academic quality through better training, or paying them higher stipends and teaching wages to ease their immediate financial concerns." I'm sure Tyler knows that UChicago Econ and other programs provide little to no training/mentorship (many other universities are just as bad). Most current Econ department students earn under $27,000 a year before tax. Please come to Hyde Park and try to live on that.

47 comments so far and only one (bigrob) claims to speak from any experience. Very typical of the comments page on this blog. As it happens, I do have experience of this issue, and my advice to prospective grad students is not to make enrollment decisions based on whether there is a grad students union or not. I was in the Teaching Assistants Union at Wisconsin, was even a union steward, and dropped out of the organization, though I think my dues were still collected. I think the union did little harm, and also little good -- maybe it did some harm to the people who wasted their time on union activities instead of their dissertations. There are a great many situations in the economy where a union does not have much leverage over pay and working conditions, and in my judgment most unions in most grad schools will not have much leverage.

it is always amazing to me how the stark logic of libertarianism suddenly discovers normative concerns when one;'s self-interest is at stake.

>When it comes to bringing about change, do today’s universities have too many veto points or too few?

Is this a joke?

Universities exist to teach facts. Not to "bring about change."

If you conflate these things, you are the problem.

Without regard to the utilitarian value of unions, how can you advocate barring freedom of association?

So what other people do we think are incapable of "wisely" investing their compensation?

Might I suggests professors (disclosure, I am technically one, but don't get paid for it)? We really should reduce their wages and invest the rest in things that will better help them pursue their goals. Like say $20,000 off the top and contribute it towards professional grant writing. I certainly know far too many professors who are terrible at investing their time & money efficiently on this process ... best to make the decision for them.

As someone who had to go through the post-graduate "training" and "investment" scam, and is on the other side of the desk today I am completely unconvinced. No, large institutions are terrible at "investing" in their trainees effectively.

In my corner of the world, we have these lovely trainees. We make them work 80 hours a week and no more than 30 hours continuously without sleep (though almost half of programs go over even these insane limits). We make it nigh unto impossible to bear children during training. We demand our trainees give up all control over: their schedules, their work conditions, their city of residence, and even which discipline they wish to specialize in. We make it difficult for them to sleep, to eat healthy, to have any semblance of work/life balance, and to enjoy the milestones of adulthood like their peers. We place them under heavy stress and give them massive responsibility with insufficient authority. We place them in highly stress inducing situations, repeatedly, and then make it impossible for them to seek mental healthcare of their own.

And what do we have to show for it? Doctors who are somewhat better than their peers in the G20 ... but vastly more jaded and embittered who demand higher compensation when we finally let them loose.

And what are these programs hoping to get out of their investments? Higher rankings. More alumni donations (generally correlated with shepherding people through the higher payer sub-sub-specialties).

So no, "investing in their students" is a crock. If they actually invested $50,000 per student they would put it on a line item Powerpoint on
the webpage and use it as reference for negotiation. Even if the mix of "investments" is generally useful ... there are far too many non-traditional students (who maybe don't need the institutions "investment" into their social life), students with atypical goals (who maybe don't need "career" advice or counseling for something they don't plan to pursue), or who have more pressing needs (who might want to burn something that is only going to return 5% per annum rather than fixing their car on the credit card at 20%). It is just another instance gatekeepers mythmaking to cover building their own petty empires.

Comments for this post are closed