Sunday, September 21, 2014

Bringing the States Back In

The hot topic in higher ed policy circles right now is the idea of tying federal ratings of various colleges to their eligibility for higher or lower levels of financial aid for students.  In theory, the availability of potentially greater amounts of financial aid would act as an incentive for colleges to do the things that would make them look better in the ratings.  If that happens, obviously, then getting the ratings right matters a great deal.  If you do something as stupid as simply tying eligibility to graduation rates, then you’re encouraging grade inflation and/or leaving high-risk students behind.  

Some smart people published pieces this weekend attacking the concept.  Susan Dynarski pointed out, correctly, that it makes no sense to punish or reward colleges for factors they can’t control, such as the economic means of their students.  Andrew Kelly pointed out that it’s much harder to measure “value added” than to measure raw outcomes, but the latter is largely a function of who shows up.  If we want to incentivize the addition of value, we need to reward it.  (This critique could also apply to Paul LeBlanc’s proposal for a free, national, competency-based online degree.  In the absence of incentives to add value, I foresee many states defaulting to the freebie and reducing their community college spending to zero.  Arizona has already come close.)

I agree with much of Dynarski’s and Kelly’s arguments, but I’d add another factor.  Although the really big money in financial aid comes from the federal government, community colleges in particular are actually run more by states (and, in some states, by counties or districts).  Any attempt to work around the states is bound to create perverse incentives.  We need to bring the states into the analysis.

Over the past few decades, as Dynarski pointed out, states have disinvested substantially in community (and state) colleges.  Colleges have responded in part with internal austerity, and in part by raising tuition and fees.  

But wait, you say.   If per-student spending has been essentially flat, why the austerity?

That’s where Baumol’s cost disease comes in.  If your spending is flat, but your real costs increase every single year, then you have to exercise the austerity you can.  That’s the primary driver behind the shift to a heavily adjunct faculty.  Paying the real salary and benefit cost increases for full-time employees against a flat total requires underpaying others.  Since labor is the bulk of the budget, labor has taken hits.

In any event, though, states remain the missing variable.  If we leave states to their own devices, and offer to increase financial aid for students at certain public colleges, what do we think will happen?

My guess is that states will use the new federal money to replace their own.  They’d treat it as a windfall.  Colleges would only be able to capture the new money by raising tuition and fees.  Colleges that do well in the ratings would basically hold steady; colleges that do less well would suffer.  So we’d have increased federal spending, increased tuition, and increased student debt.  This is not a happy picture.

Instead, I’m thinking we need to knit the states into the system.  If we want to incentivize colleges to keep students’ costs down without just resorting to chronic austerity, one way to do it would be to incentivize states to reverse the trend of disinvestment.  A clean and simple way would be to offer to match operating aid to public colleges, contingent on certain basic controls.  If a state knows that money spent on public colleges brings in “free” money from the feds, and cuts to public colleges mean leaving federal money on the table, then the incentives are finally aligned for states to do the right thing.  The actual multiplier could be adjusted, but it should be high enough to make a material difference.

Performance, I would argue, is better assessed on the state level anyway.  Nationally, the states with the highest community college graduation rates have the weakest four-year sectors, which makes sense; when the local high performing high school students attend the community college by default, its graduation rates will naturally be higher.  In contrast, a state like Massachusetts, which has a robust four-year sector, doesn’t have any community colleges with graduation rates approaching those of, say, South Dakota.  Punishing Massachusetts for having a strong four-year sector would be silly, and would do nothing to incentivize performance.  Better to judge performance in more defined contexts, so we’re actually measuring what we intend to measure.

Obviously, this proposal is only a start.  But at least it recognizes states as relevant actors with their own interests.  And it gets us away from the shallow and unproductive comparisons of colleges in different regions of the country, with substantially different needs.  Let the Feds set some basic guidelines to avoid perverse incentives, but situate performance within each state.  And let the states that step up get more help than the states that don’t.  Until then, we’ll be stuck rewarding all the wrong things.