Thursday 25 September 2014

The high cost of free tuition

A potted history.

In the beginning, New Zealand had tuition-free tertiary education restricted to an educational elite able to pass the difficult exams necessary for entry. They were well supported during their education, and they were few, and they were later hit with high top marginal tax rates as a way of ex post recouping some of the tuition costs.

Then, New Zealand shifted to greater access to tertiary education coupled with a reliance on tuition rather than high top marginal tax rates. Students were less well supported, but they did have access to student loans, and they could earn a return on their investment if they wound up choosing well, or at least a better one than their parents would have earned.

Then came a great wailing from the students. Why, WHY?! were they compelled to pay for their education when their parents' generation had not needed to pay? Nobody much pointed out that, under the prior system, entry was highly restricted. Instead, everyone viewed it as a great injustice that we did not provide open-ended public funding for anybody to achieve self-realisation by pursuing endless graduate studies in poetry. And so the 2005 election brought us interest-free student loans and Helen Clark earned a third term.

Soon after, the government realised that students respond to incentives. I recall having pointed out to the intermediate microeconomics students I was then teaching that, if they hadn't already borrowed the maximum at 0%, nothing in particular stopped them from doing so and opening up term deposits at RaboBank (or elsewhere) with a maturity coinciding with when their student loans were to come due. Unsurprisingly, most of the students had already figured that out: they were in the calculus stream of microeconomics and destined for great things.

Because of this kind of incentive, every new tertiary student imposed a huge potential liability on the government. Not only were they on the hook for the government's per-student tertiary subsidy, they were also now on the hook for all of the private portion of tuition costs, less that portion that might be repaid four or more years later without interest. Any increase in a university's private tuition charge then also cost the government directly via increased loan liability. I may be misremembering, but I believe it to be the case that we needed permission from the Tertiary Education Commission for tuition increases. The increase in demand from 0% loans could not be readily accommodated in disciplines with high salary costs and without capacity to increase tuition levies by more than the permitted cap.

Soon after that, the university funding models changed: instead of straight "bums on seats" funding, the universities instead found themselves having to submit business plans to the Tertiary Education Commission specifying how many graduates they intended on producing. Once its multiyear plan was approved, the university ceased receiving any government subsidy for any domestic students enrolled beyond 103% of that university's enrolment target.

Additionally, universities were penalised for students who failed to complete the degree for which they were enrolled. Student degree completion rates then mattered a lot. Auckland University quickly introduced entry bars as they suffered excess demand at going (and government-capped) tuition costs. This let them sort for the highest-achieving students. Other universities found other ways of ensuring that there were degree pathways for all potential customers.

And so we come to the demands in the recent election campaign for free tuition at New Zealand universities, but without a return to the exclusivity of days gone by. My fairly confident predictions of the outcomes of such a policy:
  • Government would be forced to constrain the resulting cost blow-out, either by restricting entry explicitly, or implicitly with more binding student number targets for universities;
  • There would be excess demand for tertiary entry at the regulated price of zero. A fairly severe pecking-order would quickly be established: the university seen as best would have first pick of students up to its capped allotment, and so on down the line. Where prices cannot ration quantity, excess demand has to be mopped up somehow. I'd expect really high effective entrance requirements at the institution with the highest excess demand.
  • Commerce disciplines would be killed. Absolutely murdered. The government has near infinite willingness to subsidise the bench sciences or anything that has cool kit allowing nifty photo opportunities for government Ministers. But I cannot imagine its countenancing sufficient funding differentials based on salary requirements. Any discipline whose costs are based on human capital rather than machines that go bing would be destroyed. Note that Glenn Boyle had already pointed to this problem in his great paper, "Pay peanuts to get monkeys". Because good recent PhDs in most bench sciences in the US are relatively cheap, and any recent PhD in economics or finance or accounting is really expensive because of out-of-academia opportunities, we have a hard time getting good business academics in New Zealand. This would be far worse under a zero tuition regime that would block universities from implementing higher tuition charges in disciplines like finance to fund salary differentials across disciplines. If you think that competent management is a potential stumbling point in raising New Zealand productivity, imagine how much worse it would get if we ruined the business schools. 
  • Universities would be increasingly creative in coming up with ways of charging student levies that somehow weren't really tuition charges. 
  • TEC would need to be more vigilant in guarding against the fake-student problem. If I recall correctly, one of the polytechs rorted the prior system by distributing CD-Roms to people on the street. If they ran the CD-Rom, they counted as a student. I can't remember what they had to do to get some kind of certificate for CD-Rom completion, but the polytech got a fair bit of government money out of the scheme. When the student doesn't have to pay an enrolment charge, these kinds of scams get easier.
  • While the policy would be lauded for being all great for poor people, it really wouldn't turn out that way in practice. Perhaps a few students who still couldn't afford to attend university despite student allowances and zero percent loans and existing government subsidisation of tertiary education and existing scholarship schemes would be able to get in. But the principal effect would be a big transfer to the kinds of middle-class kids who'd be going to university anyway. Highly income-targeted scholarship schemes can be progressive. Free tuition, not so much. 
  • We'd further mess up incentives at the margin to choose between trade schools and lower-value bachelors' degrees. 
  • Somehow, the whole thing would mean that the universities would have to lay off academics and hire more administrators. Just about everything yields this outcome, and free tuition is part of everything. Not having free tuition would likely do it too. Doubling university income would be about the only thing that wouldn't yield that outcome, but only because they could hire more administrators without having to cut academic staffing costs to achieve it.
Statistics New Zealand is starting to get better individually linked data tracking people from primary school through tertiary and employment. I would absolutely love to see some work taking not the average return to university over trade school, but the marginal return as established perhaps by a propensity score matching approach pairing students that seemed, on basis of characteristics and grades, similarly likely to choose university or trade school from high school but who took different paths. It would take a pile of permissions from StatsNZ, and sitting in their datacentre for a while, but boy would it be useful. 

At the end of it, you'd be able to compare outcomes, on average, for students who seemed similar at high school but who diverged in choice of trade school and university. And that might be useful in saying something about the merits, or otherwise, of free mass university tuition schemes.

Matt Nolan has a few comments on it here, but focuses more on general principles than on the abominations that would follow a zero-tuition policy in the real world.



  1. "Any discipline whose costs are based on human capital rather than machines that go bing would be destroyed. "

    Well, if it wasn't disguised as science or medicine. I think statistics and maths could survive reasonably well. Not that I'm enthusiastic about finding out.

  2. Indeed. And this would be accelerated under free tuition.

  3. For a while, we were trying to convince people that it should be STEEM rather than STEM, where the extra E is for Econ. We didn't get very far.

    I still love Seamus's line that the Soviet Union shows us exactly how great a system you can get when you pour money into base sciences but your econ's for crap.

  4. Actually, I got that line from Frank Milne, Professor of Finance at Queen's University in Canada.