Arnold Kling  

Kevin Carey on Innovation in Higher Education

PRINT
Being Single Is a Luxury... Fiscal Federalism's Failure...

Interviewed by Reihan Salam and me. The thought he expresses below is that online innovators will offer credentials that are initially inferior to college degrees but which will ultimately become superior. The analogy would be with Japanese cars, which initially were considered inferior because they were cheap. However, when people noticed that their Toyotas and Hondas were lasting longer than their Fords and Chevys, the status of Japanese cars increased.

Reihan has some post-interview thoughts, including:


At the very least, colleges and universities should be required to release data on whether or not students demonstrate a significant improvement in learning between enrollment and graduation -- and if they don't, they should be barred from receiving federal student loan money.

Full half-hour video here. Of all the video discussions I have recorded so far, I thought this was the liveliest. Comments welcome.

[UPDATE: podcast version.]


Comments and Sharing





COMMENTS (7 to date)
Justin Ross writes:
At the very least, colleges and universities should be required to release data on whether or not students demonstrate a significant improvement in learning between enrollment and graduation -- and if they don't, they should be barred from receiving federal student loan money.

Federal student loan money is "given" to the student, not to the university in any meaningful sense. Does the bank "give money" to a auto dealer when the consumer repaying the note buys a car? Not in any meaningful way. I would not be super thrilled with the government deciding what it means for a university to "demonstrate a significant improvement in learning between enrollment and graduation." This may be kind of meaningful in the 4th grade and maybe even high school, but far less so with the diverse and non-standardized paths students are allowed to take through different universities.

Universities are already competing for students who bring their federally subsidized loans with them. I see little advantage to denying a level playing field to schools which do not fulfill the government's perception of providing quality.

Greg writes:

how about .mp3 format???

Linda Sebach writes:

If institutions were penalized for graduating students who didn't learn much, they'd become less likely to admit students who, on their past records, were unlikely to learn much -- underprepared, often unmotivated. But their finances would take a spectacular hit.

Many of those young people shouldn't be going to college anyway, and don't graduate as it is; but the spectre of disparate impact would haunt any administrators who stopped admitting them. Not a pretty scene.

Chris writes:

Maybe I misunderstand what he means by "improvement in learning", but I didn't really go to school to improve my learning, but to learn more. Now, I didn't realize at that the time that my undergraduate degree (engineering) really taught me how to learn, but my Masters was much more focused on acquiring marketable skills and knowledge.

Regardless, I don't want the federal government interfering in establishing my personal goals.

Then again, I'm tired of having my tax dollars fund the college bills of students that have no objectives (party for 4 years) or whose objectives are non-nonsensical (I'll spend $150,000 on a 4 year degree that will allow me to earn $30,000/year).

Both are solved by getting the Federal government out of education.

English Professor writes:

"At the very least, colleges and universities should be required to release data on whether or not students demonstrate a significant improvement in learning between enrollment and graduation"

Colleges and universities can't do this because they do not collect such data and would be terrified of the results. Some students learn a lot in college, some absorb all the intended indoctrination, some learn very little of anything, and some pick up a series of misconceptions about what their instructors were trying to teach them. (This last is in many ways the worst--worse than learning nothing at all. You only recognize it when reading essay examinations. When estimating a student's "improvement in learning," this should be noted with a minus sign.)

Over the last decade throughout academia one constantly hears the rhetoric of "assessment"--that is, a rhetorical concern with whether students are learning anything. But in practice, this ends up focusing on everything but finding out what the students learned. When my department was told to put together an "assessment strategy" for the English major, we were given guidelines that stressed all the modern liberal educational pieties: were the students exposed to multicultural courses; were the women in classes comfortable with speaking up; and so forth. There were no questions about whether the students actually learned anything. I brought this up and proposed an exit exam that every English major would be required to take. You should have heard the outcry against this. My colleagues--the teachers of the courses--thought it unfair to test students in this way. They didn't want to know whether their students had learned what they were trying to teach them. O tempora! O mores!

And afterthought: one should also read Arum and Roksa's "Are Undergraduates Actually Learning Anything"
http://chronicle.com/article/Are-Undergraduates-Actually/125979/

Tom West writes:

The thought he expresses below is that online innovators will offer credentials that are initially inferior to college degrees but which will ultimately become superior.

I strongly doubt it.

1. The performance on any given credential vs. correlation between knowledge acquired in university is going to be moderately soft.

2. The correlation between knowledge acquired in university and actual job performance (as evaluated by worth to the company) is going to be (except at the extremes) really noisy and very rough.

3. The correlation between actual job performance (and value to the company) vs. perceived job performance is incredibly noisy.

Between these three, the only useful data you are getting (except in the extremes) is a college degree is roughly-sort-a-kind-a correlated with eventual higher perceived job performance. Hard to imagine that there's any metric that could do better enough to merit employers paying attention.

Especially when a college degree makes a fine filter and doesn't cost businesses a dime. Why spend more to learn about other credentials that aren't going to be measurably more accurate?

Now, if any business hiring a college graduate had to pay a small tax back to the government to cover the benefit the business (presumably) acquires from the government's expenditure on the employee's education, *then* you'd instantly see some low-cost credentials!

Comments for this entry have been closed
Return to top