College Choice Futures

Students choosing a college consider many issues, but especially:  "Which college is best for my future?"  Someone asked me if we could make decision markets to advise this decision.  (A decision market estimates an outcome given a choice.) 

The choices are nicely distinct, as is the choice time.  We could bet conditional on whether you start in a certain season to attend a certain college (perhaps also in a certain school of that college).  We could also consider the choice of not attending college at all.  (Breaking this condition into parts might be trickier though.)

Of course students would have to post info on themselves for bettors to consider, perhaps anonymized to protect their privacy.  And to make sure we could pay off bets, we would need students to report later on their success.  Perhaps as the price for advice on their college choice, students would pay a fee, which would later be refunded with generous interest, if the student reported as promised on their success.  (All bets could be conditional on this later report.) 

OK, but what outcome(s) would count as "success"?  Income is an obvious choice, but you might have to wait ten or twenty years, until well after college, grad school, and medical internships, to see clear income results.  One might find a way to rate immediate post-college outcomes in terms of future income potential, but this seems tricky.

If we waited long enough we might rate "fame" or "accomplishment", such as awards or media citations, but many judgment calls would be needed to make these comparable.  We could look at whether the student married and how many kids they had when, but it is not clear how important this consideration weighs on students choosing a college.

P.S.  Most bettors would probably prefer to bet on bundles of students, such as all females from New England attending Stanford.  This is quite feasible.  Next week: helping colleges pick students.

GD Star Rating
loading...
Tagged as: ,
Trackback URL:
  • Peter St. Onge

    Viz the income measure, I’m curious why private lenders don’t already do this; they could charge differential rates on student loans for different schools and majors. Perhaps there are regulatory reasons why this isn’t done?

  • http://profile.typekey.com/sentience/ Eliezer Yudkowsky

    Onge beat me to it. Or “I’ll buy 1% of your income in 10 years for $X if you try for major Y.”

  • http://profile.typekey.com/sentience/ Eliezer Yudkowsky

    Additional thought: Bets on large bundles of students mean that bettors don’t care as much about variance as the students do. Most people would rather have a guaranteed salary of $80,000 than a 50% chance of $120,000 and a 50% chance of $45,000. (“Revealed preferences” otherwise may reflect unrealistic optimism which is an example of what this market is trying to get rid of.)

  • David Robinson

    What many high school students today might have been more interested in (I probably would have when I was in high school) is what colleges they would get into.

    This situation is even more ideal for a betting market, since the outcome is very simple (accepted/rejected/waitlisted). It’s true that unlike most decision markets Hanson recommends, this wouldn’t provide any particularly useful information regarding decisions (after all, the student could always apply and see if he got in or not, so the best any market could do would be to let the student know his chances a few months earlier).

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    Peter and Eliezer, selling shares of future income is an interesting option.

    Eliezer, yes, a risk-adjusted income measure might advise better, at a cost of more complexity for traders.

    David, yes, “where would I be accepted” futures would be much easier to implement, which might well more than compensate for being less useful as a decision aid.

  • Silas

    Re: regulatory reasons, I suspect that selling shares of your future income would have enforceability problems. If nothing else, a lot of voters would view that as deeply offensive and “exploitative”, and thus ban it. (Incidentally, why don’t governments issue equity in addition to debt?)

    If it could work, that would have the added benefit of a derivatives market in labor, which would allow you profit from knowledge of future changes in wages, or to hedge against your labor’s loss of value.

  • http://quitacet.net quitacet

    the singapore government already does something like this with its pre-college employment contracts that bind students to a six-year post-college civil service/state enterprise career: it is betting on future accomplishments based on current ones. Their success at picking winners has been somewhat mixed depending on who you talk to, partly due to inherent incentive problems.

  • Pingback: Overcoming Bias : College Admission Markets