The Current Cost Debate Will Do Nothing Except Hurt Students

Next year's tuition increases are coming in, and the average seems to be between 3 and 4 percent  at elite private colleges and universities (Stanford, Bryn Mawr, Princeton, Dartmouth, MIT, Ithaca College Marquette). Tuition will be up across a wider range at publics (9 percent in Colorado, 10 percent in the Louisiana and North Carolina systems, but no increases at Purdue, the University of California or Cal State).  The Consumer Price Index is forecast to rise around 2.3% for 2013.

The national outcry against the tuition and student debt escalation has had no obvious effect on the sector's longstanding tendency to raise tuition at 2-4 the rate of CPI. See the gruesome visual below.

The public has an obvious right to be upset about absent college affordability at a time when college remains the necessary condition for a decent stable income while at the same being no longer sufficient.  Students are taking on debt to buy a high-priced ticket to "Intern Nation," nicely covered in the New York Times piece over the weekend on the "No-Limits Job" that college grads are now heir to.

What's bad for students has been good for Wall Street.  The Wall Street Journal reports this morning that "Student-Loan Securities Stay Hot" even as student default rates climb.  "Demand for the riskiest bunch" of student-loan backed securities sold last week by SLM Corp, formerly known as Sallie Mae, "was 15 times greater than the supply."  The riskiest securities have the highest yields, but investors don't have to care, given the special impossibility of defaulting or erasing student loan debt.  Meanwhile, the New York Fed reports that 90-day delinquency rates have risen from 24 to 31 percent since 2008, and that student debt nearly tripled in the last eight years.  When the Fed breaks out the numbers for Intern Nation--graduates of the past eight years--they found that "the delinquency rate jumped to 35% last quarter from 26% in 2008."

The only remedy for current debtors would be widespread debt relief.  Movements led by Strike Debt are seeking exactly this (see a discussion in Dissent, an interview with a leading organizer, NYU professor Andrew Ross, and a Forbes view with some good links).  But I know of no support from university administrators for student debt relief, even as tuition hikes continue to inflate the debt bubble.

For future students, budgeteers and their political bosses are looking in the wrong place to cut costs--namely, in undergraduate instruction. Good evidence for this came last week from retired UC Berkeley physics professor and independent UC budget maven Charles Schwartz, who has published an important update of his estimates of the cost of instruction at UC.

The basic problem is that universities have always bundled various costs together, so that the costs of instruction are hard to define.  The mainline higher ed accounting body, the National Association of College and University Business Officers (NACUBO), continues to support this bundling.  Other organizations, the Delta Project most notably, have separated  instructional from research costs (e.g. Figure 9), but methodological complexities remain.  Public universities like UC do estimate the per-student "cost of education," but research and other non-educational expenses appear to inflate the alleged instructional cost.

To take Prof. Schwartz's example, the University of California, UC's Office of the President estimates educational cost per student as $17,400 (2011-12). Average tuition is a bit over $13,000 per year, and up to a third of that is returned to financial aid, so UCOP estimates that net student tuition covers just under half of instructional cost. Prof. Schwartz, in contrast, estimates that average cost per student is actually $6910.  This is 40% of UCOP's estimate.

Prof. Schwartz concludes as follows:

• From the perspective of UCOP’s Budget Office, looking at the cost to UC: UCOP says Student Fees (net) cover 49% of their cost and I say it is 127%.

• From the perspective of students and their families, looking at the sticker price: UCOP says Student Fees (gross) cover 74% of their cost and I say it is 191%.

These low instructional cost estimates should lead to two things:

First, analyses and critiques of Prof. Schwartz's methodology and findings.  His many budget reports, extending back at least two decades, are works of scholarship in the sense that they define their methods, identify and publish their data, and explain each calculation on the way to the conclusion.  They deserve to be treated with the academic standard of rigorous peer review.

For example, Prof. Schwartz uses Activity Based Costing, and his estimate of instructional cost depends in part on a time-use survey that suggests that UC faculty devote 21% of their time to teaching.    I think this estimate is low as a UC average: it may be two or even three times higher for ladder faculty outside of the natural sciences and engineering fields that have lower teaching loads to reflect their obligations to their sponsored research.  A recalculation using higher teaching share estimates would raise the per-student instructional cost somewhat--though the adjusted figure would remain far closer to Prof. Schwartz's than to UCOP's.

Unfortunately, this series of reports has been ignored rather than critiqued by budget officials.  The Academic Senate has been part of this syndrome--and I include myself in this observation during the period when I chaired Planning and Budget committees on my campus and at systemwide.  Research faculty correctly believe that teaching and research are intertwined, meaning that research expenditures wind up back in the classroom and should be counted in part as instructional expenses.  I believe this too--as does Prof. Schwartz, as far as I know--but this would lead to a further adjustment of his cost estimates.  If we kept his methodology and make these two adjustments, instructional costs might get up to around $9500 / per student. (I'm eyeballing here, not re-running all the details).

This is still well below UCOP estimates and below in-state tuition.  So Prof. Schwartz's basic argument would remain intact.  We should be having this debate about actual instructional costs, starting with the Schwartz calculations: it is a major national policy and political issue.  And if students are racking up life-damaging debt to pay far more in tuition than they get in teaching (broadly defined)?  This is exactly the impression senior managers hope to block by avoiding debate, but this strategy stopped working long ago.

Second, bloated teaching cost estimates have fueled the national bid to fix colleges with technology. Bob Samuels has made this point in so-far futile efforts to instill realism in state agencies about real university costs. Are MOOCs supposed to save UC 15% at a $17,400 per-student instructional cost? What percentage of courses need to be removed from faculty and grad students to close UC's looming $1.5 Billion structural deficit? What if UC spends half of what UCOP says on teaching?  So far, these questions aren't being addressed at all by MOOC advocates like Coursera's Daphne Koller, whose presentation to the Little Hoover Commission was financially data-free.  The items on Prof. Samuels' list of real cost drivers have far more political clout and cultural capital than teaching does, and  it is largely for that bad reason that they aren't being seriously addressed.

Online boosters need to be held accountable to actual current teaching costs, not to ones that have stayed inflated for political reasons.  At the moment, the debate based on opaque teaching costs is offering students the worst of both worlds--no meaningful containment of the real costs that drive tuition hikes, and more austerity for their instruction.