Tuesday, April 07, 2015

Discounting and the Difference Principle


In what seems like a previous life, I slogged through John Rawls’ A Theory of Justice.  Rawls argued that we should judge policies that would treat different people differently by applying what he called the “difference principle.”  Broadly, the idea was that different treatment was justified to the extent that it offered the greatest benefit to the least advantaged.  In other words, deviations from equal treatment were only justified when they served to create more equality.  If you must treat the rich and the poor differently, treat the poor better.  

In reading about tuition discounting, I found myself drawing on long-repressed memories of Rawls.

Tuition discounting is the practice of accepting lower-than-sticker-prices from students, in order to entice them to enroll.  A school that posts a sticker price of $50,000 a year, but only asks students to pay $40,000 a year, has a 20 percent discount rate.  Discounts are usually presented to students as scholarships, but when the scholarships are funded entirely by the institution, they’re effectively discounts.  (Third party scholarships, by contrast, actually bring in money.  That’s true even when the third party is the college’s own foundation.)  When a college has too high a discount rate for too long, its survival comes into question.  Sweet Briar’s discount rate at time of death was over 60 percent.

Some colleges are getting more strategic about their discounting.  That can mean different things.  Among other things, it could mean:

  • “Plateau Pricing,” in which every credit beyond, say, the 12th in a given semester is free.  That creates an incentive for students to take 15 or more per semester, which tends to correlate with higher graduation rates.
  • Buy Three, Get One Free.  This can apply either to semesters or years.  The idea here is that a student who sticks around long enough gets a freebie.  Eat your vegetables and you can have dessert.
  • Price Freezes.  The price at which you enroll is the price you pay through graduation.  Again, the idea is to incentivize continuous enrollment.  

As a former colleague of mine once argued, it could even apply to a “graduation deposit,” like a security deposit.  You get it back, with interest, when you graduate.  I’ve never actually seen that tried, but theoretically, it could be.

Strategic discounting in these forms has real appeal.  It uses pricing to motivate desired behavior, typically in the form of on-time program completion.  Done well and transparently, it could nudge students in the desired direction by aligning their personal short-term incentives with their long-term incentives (and the long-term incentives of the institution).  

But every version of discounting-as-incentive falls prey to the same objection.  It tends to reward the students who are already the most resourced and capable, and therefore to punish the least resourced.  Which is where I’m reminded of Rawls and his difference principle.

On-time completion is easier for students with light external demands on their time and solid academic preparation.  It’s harder for students who have to work thirty or forty hours a week for pay, who have kids, and/or whose high school preparation was spotty or worse.  All else being equal, building in more rewards for on-time graduation means shifting resources from those who have the least to those who have the most.  

And that’s where too many of these discussions stop.  One side says that better results are more important than equity, and the other says that equity is the first principle.  Both agree that results and equity are basically opposed.  Both assume that a Rawlsian solution -- an incentive that has the most benefit for the least advantaged -- doesn’t exist.

The closest thing I’ve seen to a Rawlsian version of discounting is a scholarship with a GPA requirement.  The GPA requirement is there to prevent subsidized slacking.  But even there, to the extent that it’s easier to get and keep a high GPA when you have pre-existing advantages, it strikes me as missing the mark.

And this is where I’m hoping my wise and worldly readers will come to my rescue.

Have you seen, or come up with, an incentive-based discount that would actually benefit the least advantaged the most?  Is there a way to have both incentives and equity?