Bryan Caplan  

Overqualified: What's the Deal?

Rat-ta-ta-ta-tat... A High-Wage Equilibrium...

From Richard Rothstein at Cato Unbound:

College graduates are, in fact, not in short supply. Indeed, some college graduates are now forced to take jobs requiring only high-school educations. The Bureau of Labor Statistics projects that, for the next decade, only 22 percent of job vacancies will require a college degree or more. Forty percent will require only one month or less of on-the-job training, and could be filled by high school graduates or, in many cases, by dropouts — retail salespersons and waiters and waitresses, for example.
Rothstein's obviously right that people are often overqualified for their jobs. But can anyone come up with a microeconomic model in which this occurs? As far as I can tell, the prevalence of overqualification is a puzzle for both human capital and signaling theories of education.

Compounding the puzzle is the fact that despite overqualification, the return to education is historically very high. Lots of economists agree with Brad DeLong that this is a clear sign that we should invest more in education. But if many of the college grads who already exist aren't using their skills, what's the point of spending resources to make more of them?

One possible explanation is that "overqualified" college grads were low-quality students who didn't learn anything. But in that case, shouldn't we expect that encouraging more people to attend college will give us more of the same?

P.S. Yes, I know about IV estimates; I'm just not convinced.

Comments and Sharing

COMMENTS (33 to date)
Fabio Rojas writes:

Question: How much "overqualification" is temporary? If your model is that the overqualified are simply the dim range of college graduates, then one would expect that they would never move to more cognitively demanding jobs. If you see a lot of short term overqualification, then maybe all it is the effect of queing and sorting in the labor market. The weaker grads simply have to wait a little longer to get the "right" jobs.

From my experience teaching undergrads, I doubt these employees are in fact overqualified. A degree does not equal knowledge.

Yes, encouraging more people to go to college will give us more of the same. At the same time, our colleges don't seem to be turning out anyone who knows much more than what they should have learned in high school. As a result, jobs that required undergrad degrees are now requiring graduate degrees. As more people go to grad school, I expect it to dumb down as well.

Alex J. writes:

It seems that the overqualified-on-the-job graduates are not the graduates reaping the historically high returns to higher education. I think two factors may be at play here. One, the high returns to education might actually be high returns to high-IQ, highly motivated young people, the sort of people whom we would not be suprised to see graduate from college. Two, young people who neither accumulate human capital nor generate any valuable signal still have motivations to go to college: peer pressure and consumption. All of their friends are going to school, and this is an opportunity to spend 4 years out of the workforce with a bunch of similar young people. They may incur a heavy debt of student loans, but these are the same people who often incur a heavy credit card debt also. Plus, student loans are subsidized, as is a college education generally. (I might add, undergraduate educations are subsidized by parents, and perhaps university endowments, as well as various levels of government.)

Felix Chow writes:

"Over-qualification" suggests larger costs than benefits at the margin. However, there are a lot of harder to measure benefits that come from a university education.

One might enjoy the higher social status of a college graduate as compared to someone with only a high school diploma. College graduates may have more bargaining power in the marriage/relationship markets. There is also enjoyment from the experience of learning and making life long friends. Adding positive externalities to society of having a well educated population, I suspect people are not that over-qualified at all.

GU writes:

I think Felix is on to something. As the U.S. has become more and more prosperous, the wealth has trickled down, along with the attendant expectations that greater financial well-being bring.

Basically, families that would have never sent their kids to college 30 or 40 years ago now do so, in large part because those families are much better off financially. Kids who grow up with comfortable middle class lives expect the status that a college degree supposedly confers, whether they would really benefit from it. Easy government money and parents all too willing to subsidize make this choice easy.

As other commenters mentioned, this phenomenon has really stripped the BA/BS of any real status or meaning, forcing those who really want to differentiate to go to grad school (admittedly, many of these folks would have gone anyway).

Jody writes:

Isn't this akin to a traditional cause of a recession?

Specifically, companies mistakenly overinvest/overproduce in the expectation that current (perceived) trends will continue. When the trends don't continue, the manufacturers got too much and have to cut back production.

In this case, the companies were students and the overinvestment was human capital, it's another example of the effect of the aggregation of erroneous information in a market.

Typically, this would eventually clear by letting inventory drop (a nice feature of market with feedback), but in this case, DeLong et al are continuing to advertise erroneous information which is diluting the feedback thereby hindering the ability of the market to clear (i.e., cut down on college grads for a few years). This means that this overqualification trend will continue and deepen.

For fun, note that another factor hindering the clearance of this market is government subsidies (low cost student loans) propping up an overvalued sector of the economy.

FC writes:

Then there is the false equivalence between degrees when used as a ticket to employment and status. A bachelor's in science or engineering is far more difficult and content-rich than a bachelor's in theatre or fill-in-the-blank studies. But they cost the same and confer the same civil service preference.

It's not that "'overqualified' college grads were low-quality students who didn't learn anything," but that they are high-quality students who didn't get a degree in anything useful. The proportion of high quality students that choose liberal arts majors versus careers that consider liberal arts degrees worthwhile is high. There is a simply a surplus of general talent that did not receive specific training relevant to our economic system.

Mr. Econotarian writes:

I've been interviewing a large number of recent electrical engineering graduates.

Suffice it to say that most did not appear to have held on to what should have been fairly basic EE knowledge.

One problem seems to be that students work on group projects, and it is clear that only one person in the group really "gets it". While this may be a reasonable model of the working world, it doesn't help individual students become knowledgeable.

I would suggest that engineering schools require fairly complex individual thesis projects before graduation.

Mensarefugee writes:
One might enjoy the higher social status of a college graduate as compared to someone with only a high school diploma. College graduates may have more bargaining power in the marriage/relationship markets. by Felix Chow

Classic Stadium analogy comes to mind. If we could do away with all of it, we would all be better off. But there is no incentive for the first guy to start. Pathetic eh?

Mr. Econotarian writes:

I've been interviewing a large number of recent electrical engineering graduates.

Suffice it to say that most did not appear to have held on to what should have been fairly basic EE knowledge.

One problem seems to be that students work on group projects, and it is clear that only one person in the group really "gets it". One problem seems to be that students work on group projects, and it is clear that only one person in the group really "gets it".

Hmm... maybe Computer Science is better after all. All the snide deriding about Immigrants pulling it down aside...

Steven McMullen writes:

This mystery is not difficulty to solve.

The return to education measured is an average return, not the return that can be expected from everyone. Standard human capital theory says that the return that the marginal student gets from their college education will equal their marginal cost (opportunity cost).

This is also the average return across all colleges. The return to attending some colleges is likely higher than others.

As others have stated, the marginal cost is likely quite low once grants, loans, in-state tuition, and parents are factored in.

Finally, where is the evidence of over-qualification? In the 2000 census roughly 24% of US residents over age 25 have a college degree or greater. This lines up pretty nicely with the 22% of new jobs requiring a college degree.

Steven McMullen writes:

Also, I am sorry to say that your critisism of the IV-results for the returns to education is unfounded.

(see comments for that post)

1) Many jobs today require a college degree by law. Lawyers, doctors, nurses, therapists, many engineers, civil service, teachers, military officers, etc. all require a degree by law. Even for jobs where accreditation agencies do not explicitly require a college degree, the employer must often require one anyway because of the Supreme Court's decision in the Griggs case.

2) Few parents want to believe their kid is too dumb to benefit from college.

3) For a student, college is usually heavily subsidized, both by the state and parents. Few turn down four years of subsidized hanging out and partying.

4) The university places the same role in today's society that the church used to play. It sets the national ideology. Thus many people will go to college for the same reason people used to buy indulgences. They have been taught that a college education is necessary for a complete life.

Dr. T writes:
Then there is the false equivalence between degrees when used as a ticket to employment and status. A bachelor's in science or engineering is far more difficult and content-rich than a bachelor's in theatre or fill-in-the-blank studies. But they cost the same and confer the same civil service preference.
Not in the federal government, which has screwed scientists for decades. Science degrees give you lower GS ratings, not the higher ratings one would expect. For example, a BS in Medical Technology gets you a GS-5 starting position. That's the same GS rating held by high school graduate phlebotomists (who draw blood). People with a BS in Sociology or Psychology can start in Human Resources as a GS 7. People with a BS in Criminology or Psychology can start at the FBI as a GS 9. The physician who heads an entire VA hospital Medicine Service is a GS 15. The associate medical director (an administrative position) is a GS 16.

With policies like this, it is no surprise that students are spurning difficult majors in science and engineering. The extra work does not pay off. One can have orangutan-level intelligence and get a business degree at some colleges, and many of those orangutans with BAs get jobs that pay more than what scientists make.

MouseJunior writes:

Part of it is probably due to the decline in academic performance at the sub-university level. Employers end up hiring college grads for low skill jobs because the people who would normally fill them aren't making it out of high school with even a minimal skill-set.

If you can't do basic math, you can't close out a till at the end of the day. If you're not competent at basic English, getting a job doing data entry is often out of reach.

John Fast writes:

1. Bryan, I find your original post to be ironic considering that you already pointed out that most "education" is really just signalling.

2. I also agree with all the posters who believe that education has been "dumbed down" so that -- very roughly speaking -- a college degree today is the equivalent of a high school diploma from 30 or 40 years ago. Part of that is due to the decline of government schools (probably best-explained in Losing Ground) and part is due to having a much greater part of the population going through college, rather than only the most qualified.

The G.I. Bill was a great idea, particularly because, until then, university admissions were based primarily on ancestry and/or social class rather than merit. Unfortunately people then got probably-mistaken idea that most people should go to college.

The 1960's were a not-so-great idea and they -- especially the system of draft deferments for the Vietnam War -- gave people the idea that everyone should go to college. That system also promoted grade inflation because professors felt that giving bad grades was morally equivalent to a death sentence.

3. IMO the best solutions would be to reform lower education (K-12) along the lines Murray suggested in Losing Ground, and to promote alternatives to college for those who aren't really qualified for it. And, for the record, we should obviously keep the idea that college admissions should be based on merit, not social class; perhaps having fewer college students -- by weeding out/diverting applicants who either aren't qualified, or don't really need a degree for their careers -- will allow more scholarships for the truly qualified...

Make college harder to get into and harder to pass and they will make high schools harder, which will make middles schools harder, which will make elementary schools harder, which will result in people actually getting educated. Also, split middle and high school into college-bound and trade schools -- in other words, make every middle and high school a magnet school. With trade schools, you will make education relevant to the majority of students who do not find education to be relevant at all.

Peter St. Onge writes:

I think most return-on-education literature sees all college majors as undifferentiated. So a sociology degree or master's of fine arts are lumped together with, say, nursing and law degrees.

In a society with increasing wealth we might expect students to substitute towards more enjoyable degrees (and the more enjoyable careers they lead to) and away from harder degrees, which may lower average productivity of a college degree. And so we end up with increasing proportions of psychology and sociology human capital, and less engineers, skewing the average wages.

DanT writes:

Most likely components of a microeconomic model for job overqualification:
1. Systemic overqualification: Employers may be required (or may prefer) to hire those minimally qualified for a position, but pay no premium for overqualified staff. If there are many potential employees, at the time of hiring there will be no overqualification for individual positions. If even one individual in a single position pursues additional education while holding a job (to improve his own future job opportunities), there will be overqualification. Due to the disjoint nature of educational credentials, being overqualified for one position does not necessarily make an individual qualified for the next higher position. The desire of individuals for improvement and disjoint nature of educational credentials, combined with inefficiencies in changing positions in even a large labor pool, leads to systemic overqualification. (RESEARCH OPPORTUNITY: turn this into a mathematically rigorous model)
2. Measurement error: Bureau of Labor Statistics may only be measuring posted job vacancies. Commercial companies are increasingly not posting job vacancies (which provide valuable information to competitors), instead using Internet resume repositories and direct referrals - especially for higher skilled positions. This will skew the BLS report to underreport high-skilled jobs. If this is true, BLS needs to update their job estimation methods by querying companies confidentially. (RESEARCH OPPORTUNITY: measure job vacancy listings versus company surveys on job opportunities not listed.)
3. Assumption Error or Voluntary Overqualification: The assumption is that graduates always seek a job using their degree. Graduates may voluntarily not seek such a job for various motivations, either temporary or permanent. (RESEARCH OPPORTUNITY: survey recent graduates not using their degree)
4. Temporary Circumstances: Recent graduates temporarily take jobs to pay the bills while they search for a better job. (RESEARCH OPPORTUNITY: measure whether college graduates have a job using their degree 5 years after they attain it.)

All of these contribute to the overall model to some degree.

Gary Rogers writes:

The fact that you are overqualified for a lower level job does not imply a qualification match for a higher level job. For example, someone with a Phd in history is overqualified as a sales clerk, but not necessarily qualified to be a computer programmer. The higher the education level, the harder it is to find an exact employment match.

True enough. And then there are absurd situations like I keep finding myself in. I was told by an English dept. looking for someone with a Ph.D. that I was overqualified for the position (my Ph.D. is in the Humanities, M.A. in English, B.A. in Recombinant Gene Technology). David Boaz at the Cato Institute told me that it was decided I was too "unfocused" -- which, since I am an interdisciplinary scholar, is just another way of saying "overqualified" (or, perhaps, "we really just don't know what to do with you"). It seems I have somehow managed to find a way to educate myself out of any job. Some days I really feel like saying, "For the love of God, will someone out there please exploit me!"

Axel Molotov writes:

Two comments:

One, an oversupply of college graduates could be a result of a shortage of highly qualified workers, i.e. those with graduate degrees or highly developed entrepreneurial skills. It is no secret that this country is currently experiencing a 'brain drain', where a lot of highly qualified foreign nationals are either returning to their countries or choosing not to come here at all. Most of us have read about the H1 Visa shortage (H1 is for specialized workers). Highly qualified individuals might generate a demand for college-educated employees, and, in turn, increase the amount of job vacancies that require a college degree. Trickle-down economics it may be, but the correlation between the current shortage of highly specialized workers and the surplus of college educated yet unspecialized labor is hard to miss.

And two, for all those practically calling for a reduction of college-educated individuals - either through raising college admission standards or what have you - consider the alternative. In my opinion, it is much better for a country to generally have a higher level of education. In other words, I would rather have retail salespersons and waitresses have college degrees than be dropouts.

John Fast writes:

1. I may want to "revise my remarks" (as they say in the Congressional Record) based on Steven McMullen's very significant remark that

In the 2000 census roughly 24% of US residents over age 25 have a college degree or greater. This lines up pretty nicely with the 22% of new jobs requiring a college degree.

I'd like to look at that in more detail, such as looking at college graduates as a percentage of the actual workforce -- so we should exclude retirees, for example. And I'd also like to confirm what percentage of adults 18-24 attend college, in case college education is far more (or less) common among the current generation than for the population as a whole. And, finally, I'd like to know what percentage of total jobs -- not just new jobs -- require a college degree.

But his point certainly makes a lot of sense to me!

2. Here's my current microeconomic explanation for the overqualification phenomenon (assuming it really exists): The most qualified people get the best jobs. Others who are marginally qualified for those jobs are forced to seek lesser jobs for which they are overqualified. Employers who have a choice will prefer (presumably because of signaling) to hire someone who is overqualified.
Someone who is overqualified is either significantly more productive than someone who is marginally qualified -- in which case they will be paid somewhat more; or else they are not more productive, and the overqualification is merely signalling their "conscientiousness" so the employer knows to hire them, but doesn't pay them any more than he would pay a less-qualified employee.
Either way, overqualified employees in Nth-tier jobs squeeze out marginally-qualified employees who then get (N+1)th-tier jobs for which they are overqualified; repeat as necessary.

Tom Whiston writes:

The problem is that college educations in the United States are heavily subsidized by government intervention. This creates a false incentive to get a college education. This is the same problem China is facing, they have Engineers with PhD's that wash cars. In the United States, specifically at GMU we have people like Tryia Venkatraman and Malda Al-Sarayji in grad school who don't even deserve to be there. Malda is a great example because she was unsuccessful in completing the graduate program at mason and Tryia is just an embarrassment to the program.

It is what it is...

Dipper writes:

well maybe all these jobs now are graduate jobs.

We've doubled the supply, but not changed the jobs landscape, so jobs that were once done by people who had A levels and didn't go to university are now done by the same people except they've gone to university. Hence the jobs are graduate jobs.

happyjuggler0 writes:

What job isn't filled by people who are overqualified? Seriously.

What qualifications does one need to be a waiter beyond elementary school level math and English? We have high school graduates doing these jobs that only need a 6th grade education to fill. Oh, the horror!

One simply is overqualified when one learns about WWII in high school, or Columbus in the first grade, and doesn't use that knowledge. Is this a problem? If not, then why is it a problem that someone with a BMW degree (i.e. a degree in Black studies, or Marxist studies, or Women's studies) is working as a cashier at the local gas station?

When college trained geologists, or those with Bachelor's in chemical engineering, etc. can't land a job at any wage then I'll concede we have too many people going to college. Until then all that the "overqualified" stat shows is that some people spent four (or more!) years in college getting a useless degree when they could have spent those four or so years getting a degree where they could actually learn an employable skill.

In short, we have too many people majoring in the wrong subjects. Perhaps we ought to limit subsidized student loans and grants to those who major in logic based subjects like science, technology, engineering, math, law, medicine, economics, business, etc.

Tom Whiston writes:


Or maybe we should have no government subsidies for any education. See Henry Hazlitt, Saving the X industry (a chapter from Economics in One Lesson). Government subsidies divert resources from where they are valued the most. This hurts that industry and promotes a miss allocation of resources because the government subsidy takes from one sector of the economy to promote another sector. In this case, labor markets are being skewed and individuals are investing too much money in their education.
Thanks to the government subsidization we have whole departments that wouldn't exist without government intervention. These departments promote nothing but socialist ideologies and create a hostile learning environment if you would happen not to be sympathetic to their ideologies: English, Education, Sociology, law and most of the others you cited above.
The answer is to get rid of government subsidies all together and the welfare state.

Tom Whiston writes:


sorry may have overstated the departments that wouldn't exist without a government subsidy. Math, Engineering, etc. Nonetheless there still should be no government subsidization because science is a quest for truth. When tainted with government subsidies, science is no longer a quest for truth, but a quest for who can get the government entitlement for their research. This makes scientists not quest for knowledge and truth but to become a tool for politicians for their own political gain.
See Dr. Bennett's work for good examples.

quadrupole writes:

The real problem here is that the risks have been misjiggered.

Consider the following reform:

1) Require the university to assume the default liability for student loans.
2) Allow for modified bankruptcy dischargment for student loans as follows:
a) The lending institution gets their money immediately from the university.
b) The IRS collects from the students 10% of their income up until 15 years after graduation and remits it to the university to repay the debt.
c) The student cannot be held liable for any part of the debt beyond 15 years after graduation in a chapter 7 bankruptcy proceeding.

So what does this do? It gives the university a pretty big stake in giving students a useful education while balancing this with a strongly encouraging students to repay their student loans. As long as the university equips you to earn a pretty good living, it's pretty low risk for them. If 10% of my earnings over 15 years will cover the cost of the education, the university is pretty nearly guaranteed that even if I do default they'll get their money back. But if they send me out with a fill in the blank studies degree that provides no value add... well... they're likely to loose money.

It also allows the universities to more realistically evaluate their departments. If department A has a lot of defaults among it's students, then maybe department A is a liability to the university.

It's all about the incentives.

Matt writes:

Brian points out that investing in higher probability of failure lowers return.

That is a choice problem for investors. If they could buy shares of students efficiently, then the market would provide educational completions in smaller increments, spread the risk.

That is, if my rate of return of a Princeton grad is 7%, then at efficiency, my rate of return for investment in an auto tech grad should be the same.

The best environment for investing in education is when you have variety in students and variety in schools.

Leslie writes:

Quadrupole's reform would simply make it easier for those more financially well-off to get into college. If a student from a wealthy family applies to a university, the university has an incentive to admit that student over one who needs to take out loans. This would simply be a regression to a class-based society -- cyclical because no one would have the opportunity to move up. And no legislation should force students to take out loans for college.

charles darwin writes:

if the internet is any indication:

colleges will be obsolete in 5-10 years.

all a college does is sell Admissions tickets.

1. yes, you can come to class here.

2. no, you cannot come to class here.

digital age
digital lectures
digital knowledge
digital education

1. Real lecture:
Room 101
Tuesday morning
$50,000 tuition
18-30 in age.

2. Digital lecture.

any classes
any knowledge

free to world
free knowledge
free education for all

colleges will be obsolete in 5-10 years.

Myriam Robin writes:

to charles darwin:

i doubt colleges will be obsolete. from my experience as a current university student, much of what you learn there is self-taught anyway. it's the environment of college that's more influential than the actual teaching. this can't be conveyed in a digital lecture, you'd be missing out on the best part.

Comments for this entry have been closed
Return to top