What happens when statistical models don’t work?


While much of the blogosophere I follow has been debating the math of newspaper payment models (which has been especially amusing, as most of the people debating did everything they could to avoid taking a college math course), I’ve been wondering about another marketplace that is about to learn whether or not its economics and “business model” are flawed. (Or, if not flawed, it may provide some insight into how bad (or good) the economy may really be.)

I’ll admit, I have a personal interest in this economic indicator, as the 18-year-old who serves as my in-house focus group allows me to have an up-close and personal look at this marketplace — higher education admissions.

Specifically, what, if any, effect does the current economic meltdown have on college admissions this year? The New York Times has a story today that explores the dilemma facing college admissions people. (But first, let me say how much I respect and admire, yea, love, those who do this type of work at one of the schools to which the 18-year-old has applied.)


“Typically, they rely on statistical models to predict which students will take them up on their offers to attend. But this year, with the economy turning parents and students into bargain hunters, demographics changing and unexpected jolts in the price of gas and the number of applications, they have little faith on those models. “Trying to hit those numbers is like trying to hit a hot tub when you’re skydiving from 30,000 feet,” said Jennifer Delahunty, dean of admissions and financial aid at Kenyon College in Ohio. “I’m going to go to church every day in April.”

In the next few weeks (especially April 1-15) through the fall matriculation at the nation’s colleges and universities, there will be a great opportunity for economists to study if “selective” colleges and universities are immune from economic cycles that the current meltdown has shown affect other “can’t lose” investments.

Could this year prove to be what Barbara Fritze, vice president of enrollment at Gettysburg College, describes as being “like the dot-com bubble burst for higher ed. We’ve been in this growth mode for a period of time. Now there’s a real leveling going on.”

Or what if it’s worse than the dot.com bubble burst? What if it’s like the sub-prime, Wall Street meltdown burst? Or what if it’s not like the meltdown, but it is a part of the meltdown like, say, the falling prices of real estate, financial markets and commodities. Many of those elite universities have seen their endowments fall in the same way 401Ks and real estate have dropped. And the New York Times story doesn’t even mention the effect the global economic meltdown might have on the rate of acceptances by full-tuition-paying international students who apply to such universities.

Perhaps there is, as we’ve been led to believe, an endless supply of over-achieving applicants to these selective institutions — along with a limitless source of tuition funds from parents and scholarship sources. Or perhaps there isn’t. Either way, this is going to be an interesting story for economists (and parents) to watch, even if you don’t have a high school senior.

(Personal note: Hang in there college admissions people. We’re all pulling for you. You’re the greatest. Really.)

[If you don’t recognize the image that accompanies this post, it’s a poster from this really awful film.]

  • Hudge

    Applications from Tennessee to my Alma Mater from are up this year, and total applications are up 12 % – which I hasten to emphasize says nothing about how many will be offered a place, nor how many will accept. The school has a need-blind admissions policy, and this year may very well test its ability to meet demand.

    Anecdotally, based on interviews with a large number of applicants over the years, I believe that not all the applicants are over-achievers; though most are very bright and ambitious, and quite a few have achievements that are almost scary at their age, most have to realize they are playing an academic version of the lottery. And they do it with multiple colleges.

    Out of around 18,000 applicants something less than 1,000 will actually matriculate. And more than half of those will come from the early admissions pool, which is much smaller (and you are required to swear on your grandmother’s gray head that if accepted, you will come). We interviewers usually scream once a year that someone we really thought was terrific material didn’t even make the waiting list (frustration over having none of one’s interviewees admitted is one of the leading causes of attrition in the ranks of interviewers.)

    Even so, if powerball had those odds, I’d probably be doing direct deposit of my paycheck, so maybe they’re making a shrewd investment of their parents’ application fee dollars, hoping to hit the jackpot. But I suspect there is also a large amount of status grabbing here, for parent and child, which keeps boosting the numbers.

  • @Hudge – I feel certain your top-shelf-brand Ivy League alma mater will do fine, as always. However, I think that we’ve seen the economy collapse even more dramatically from the time those applications were sent in. The notion that endowments are going to continue to grow every year has been completely obliterated by the recession. Next up may be the “applications are up” every year metric, which, frankly is not a relevant metric when the common application and $50 can allows you to apply to any school. Again, I have no doubt that the Ivy League schools will accept a small percentage of those who apply — and those students will, no doubt, be stellar — but even those schools are not immune from the economy.

  • Great stuff Rex,

    The intersection of social networking, the economy, and the student/parent dynamic makes this year all the more interesting. I know that our network of sites (www.usphere.com) has seen much more interest in state schools and in community colleges in years past. (In fact, we’ve launched two blogs in the past week to celebrate niche programs — http://www.mbasphere.com and http://medschoolsphere.com.)

    Tack on the fact that you can bolt together a good, respectable, and career-launching degree from a few places (online, community college, etc.) and the old predictive models will continue to be pushed and pulled and tugged all over the place.

    The US News rankings, which each year talk about yield and acceptance rates and all, should look nutty next year.