The Puzzling Supply-Side of Skills

I’m totally new to blogging, so readers, please cut me some slack.

I plan to write about things that bother or puzzle me, and here’s the first one.

It is well known that the so-called “college premium” – the additional income college graduates earn over their lifetimes relative to the incomes earned by those with a high school diploma or less – has widened substantially over the past several decades. The puzzle is why that this clear payoff from going to and completing college has not induced more young people to go to college. Or, put differently, why has the high school dropout rate – which reaches 50 percent or more in some urban areas – not fallen?

Answering this question is important for many reasons, but among the most important to me is that if we had a better handle on how to dramatically cut the drop out rate, we could make a significant dent in reducing the poverty rate – and most likely the crime rate as well. Moreover, since President Clinton was clearly right when he famously repeated “You earn what you learn,” a lower dropout rate would narrow income inequality, while making our labor force more adaptable to continued (and seemingly accelerating) change.

A recent paper at voxeu.org by Joseph Antoni, Prashant Bharadwaj and Rabian Lange documents the college premium puzzle and offers a few answers : other things than money (such as what peers think) drive kids’ decision about whether to stay in school; parents may not be aware of the economic benefits of having their kids stay in school (this is hard to believe); kids are “liquidity constrained” and thus don’t have enough money to pay for college (but kids drop out well before they graduate and there are plenty of ways to finance even community college, while it also still pays to finish high school).

The authors admit that more study is required to fully explain the persistently high drop out rates. While I agree that certainly more research is always a good thing, more effort in the meantime should be devoted to experiments with ways to cut the drop out rate.

One good place to start is with incentives for teachers. We know that good teachers are essential not only for improving student achievement, but can also play a huge role in students’ lives – positively or negatively. A teacher, even in the earliest grades, who convinces her students that they can go to college and realize their dreams can give them the confidence and hope they need to do precisely that. Conversely, teachers who send subtle (or not so subtle) signals to their students that have no chance of finishing high school, let alone going to college, can also help about that unwelcome result.

If all this is true – and I believe it is – is not time to begin thinking creatively about ways to reward teachers based on the high school graduation rates of their students?

Thoughts – on reasons for the high drop-out rates and ways to reduce it?

Comments

[Gregg Gordon]

Very informative post, I agree with the analysis and very much like the idea of creating long term incentives. The current focus on standardized test scores helps but clearly isn't enough.

Seeing the issue from several different perspectives, I am convinced that a teacher's attitude and belief in her students is the most important factor.

Gregg

PS Congrats on the blog, I look fwd to future posts.

[Paula]

Kids drop out of high school because high school sucks! Kids will stay in school at the same rate that it is made less boring, prison-like, and socially humiliating.

[EZ]

Could it be that the correlation between college degree and higher wages is not a causal relationship? Maybe the people who decide to go to college would be ones who would earn the premium even had they not gone. On the flip side, maybe the people who do not go to college would not make as much as those who currently do, even if they did go. I am not sure that this is true, but if it is then we have to figure out some other way to reduce poverty other than simply attacking the drop-out rate.

[Michael F. Martin]

A good start to improving teacher quality might be permitting non-credentialed professionals to "guest teach" for short periods.

Getting a variety of perspectives from people who are out doing things in the world might get students more excited about going to school.

But in most states, I would expect the Teachers' Unions would defeat this type of proposal on the grounds that it would lower the quality of the teaching.

[chuckles]

If you want more fewer drop-outs, here's a radical solution.

Make high school shorter--without sacrificing academic rigor. (Reducing the time cost of HS is, I think, an efficiency gain by definition.)

It should be the case that in order to leave high school, you should have to pass an exit exam. (No one should leave high school, for example, without being able to read or perform algebra, etc., etc.)

But then, ALSO stipulate than anyone over 14 can get out of high school.

The harder you work--and the more you learn--the quicker you can "break free" of high school. Remember, for many students, school is very closely akin to incarceration or at least has high opportunity costs.

We should use a "carrot" (not "stick") approach to improve high school achievement.

To students, we say:
"Work harder, earn your freedom sooner."

All of this is contingent on a rigorous exit exam. If this exit exam is rigorous enough, it should allow employers to hire teenagers, knowing that they are skilled enough to learn their jobs.

Finally, people must be allowed to work full-time from 16 upwards.

[drtaxsacto]

A part of the differential could be in the pricing mechanism for college. Parents have a very tough time understanding the difference between Price, Cost, Subsidy and Net Price. (See the work of Gordon Winston at the Williams Center for the Economics of Higher Education at Williams College).

Reorganizing high school should also be a part of the solution. The model we rely on most in the US was built for kids more than a century ago - it does not fit this generation.

Finally, the economic differential between high school and college graduates is based on supply and demand at least at this point. There is some fairly good data that shows when the differential began to grow. The differential is worldwide - the net compensation for high school educated workers has declined in real terms in most countries around the world. Indeed, some people without a college degree are able to produce higher wages than the average but I think the data link (between education and income) is causal. In recent years there has been some discussion that the relative value of a BA has diminished in a similar curve to what happened to a high school diploma a generation ago.

[Mikko]

I think you need to look at the data more carefully. In your question, I believe, is a hidden assumption that all high schools are the same.

What if there are high schools that are very, very bad at delivering any help for pupils to learn - and that the drop outs are mostly from such places? Or perhaps we have cultural ideas that affect certain parts of population with anti-schooling mentality?

Just knowing the aggregate data, you can't know.

It would be interesting to see someone do a good statistical analysis on this by school, or by district. Is it people from better schools or worse schools that drop out? Or is the drop out rate the same regardless of the school...

[brainwarped]

Getting rid of teachers is an interesting solutions. Given that the majority of teachers are bad, getting rid of teachers will help decrease the amount of kids who drop out of high school due to teacher cues. Replacing teachers with video-taped lectures and professional advisers would make learning more fun, and give every student incentive to complete high school. The advisers would be required to push their advisees to their highest ability through inspirational mentoring. This would decrease the amount of high school drop outs due to parenting cues. But, I don't think the teacher unions would like this...

[rlitan]

You're right disaggregated data are needed. What I know is that dropout rates are disturbingly high in both rural/suburban schools and in inner city schools, but the problem is worse in the latter.

I would propose, however, that any teacher incentive scheme tied to high school completion should be targeted on schools with unusually high drop out rates.

[rlitan]

This is all true, but still the puzzle remains, why have not more students here -- and for that matte elsewhere -- responded to the growing college/high school wage gap by completing high school and going to college?

I don't claim to have all the answers. I agree with you that perhaps a reorg of high school may be part of the answer. But ultimately because the research consistently tells us that its the quality and motivation of the teachers that count, it would therefore seem that tying teacher incentive to the goal of getting more kids through high school also has to be part of the answer.

[rlitan]

The fact that the college/high school premium has risen does not necessarily mean that college education "causes" higher wages among higher college graduates, only that the demand for college graduates is growing relatively more strongly than it is for high school graduates (o0r those with less education). It's just a matte of supply and demand: the demand for college graduates is growing faster than the supply, and this is more true than for high school graduates. Hence the growing wage gap between the two.

[Dan Weber]

Perhaps the differentiation between the college-degreed and the non-college-degreed has increased because discrimination against that non-college-degreed has increased.

Imagine society is arranged like a pyramid, the pyramid divided into ten slabs. Generations ago, only the top(and smallest) slab got educated, and that slab discriminated against the slab beneath it, while people further down the pyramid didn't even notice.

People in that second slab would make sure they got educated, and their kids, too. After that was done, they then discriminated against slab #3.

Repeat this process. The differential grows over time, because the size of the slabs increases. (There is a much bigger border region between slabs 8 and 9 than there is between slabs 1 and 2.)

The bottom slab cannot be completely educated in the US because we're importing millions of non-native speakers a year.

(College certainly works on the "education" aspect for a whole lot of people, but I am increasingly fearful that for most others it's just perpetuating an arms race of certification-chasing.)

[rlitan]

Yes ~I agree there is probably a large amount of credentialing at work here -- college has become the new high school, as it were.

But again, if we as a society want to close some of the inequality gap, we're going to have to find a way I think to raise the fraction of people who get that more worthwhile credential, the college degree -- and that fraction has been stuck for years.

My suggestion so far has been to change teachers' incentives as a way of encouraging more people to stay in school.

Now, it could be that if more kids complete high school, we will simply ratchet up the credentialing "arms race" you mentioned and thus there will be no net effect on reducing inequality. I don't believe that, but others out there may, and if so, would welcome their views.

[Joshua Macy]

Well, if Bryan Caplan is right that college educations mostly serves as a signaling function, then it's not a return on education itself but a return on being the kind of person who can stick out a tedious multi-year task for a distant future reward. That is, college graduation is just a hard-to-fake demonstration to future employers that you're the kind of person who can provide value in their workplace. If you manage to figure out a way that even lazy, flaky people who can't force themselves to attend a few hours a week of classes or turn in any written assignment on time in a semi-readable form can get through college and get a degree indistinguishable from anybody else's you might well find that the returns drop as employers stop valuing the certificate. My guess is that employers would try to find ways of distinguishing among the degrees, and you'd start worrying about the education premium gap between more and less highly ranked colleges or majors.

I'd be curious, though, as to what the statistics are if you distinguish between skilled-but-non-college-credentialed labor (such as plumbers or auto mechanics) and unskilled labor.

[rlitan]

Joshua makes an excellent point about college (and same is true for high school) completion being a signaling device to employers rather than adding some valuable skills. But even that is entirely true -- and I doubt it (based on my own experience and that of my kids) -- then it would still be in society's interest, to the extent there is broad social concern about growing income inequalities, to have more people with the credentials that are favorably viewed by employers. Add to the supply and the college premium relative to high school completers and dropouts must fall.

But then Joshua argues that with more people with college degrees, employers will sort among the college graduates, by prestige of the university. Actually, we know they already do this -- just look at the starting salaries of graduates, or MBAs (even better example), from different ranked schools. But that's OK for my overall argument: it still should narrow the wage inequality for any individual that gets that credential relative to a world in which that person does not have that credential.

It all comes down to incentives -- for the kids, for the teachers, and for the parents. Right now, the social milieu and peer group pressure at high dropout schools provide non-monetary incentives for kids to drop out and/or the kids heavily discount any future prospects for themselves, given their socio-economic backgrounds.

But there are plenty of exceptions in high dropout school systems: kids from poor circumstances do break out of the trap. The challenge for society is to find more and better ways of incentivizing those who now are likely to drop out to stay in. I proposed earlier supplemental incentives for teachers tied to graduation rates of their students. RolandFryer of Harvard argues that small, but timely payments to kids themselves could help turn things around. Perhaps paying parents in low income areas for their kids graduating high school would provide reinforcing incentives.

Bottom line: it's long past time, in my view, to begin thinking out of the box, but using the insights of economics and the importance of incentives to address a very serious social problem.

[Dan Weber]

It seems that for the lower-class, encouraging them to chase a college degree merely as a credential is cruel.

There is already a long list of things that employers cannot ask about: religious background, family status, ethnic history. Simply add one more to the list: education.

If going to school makes you smarter, then it should still be good for you. But if it is just credential-chasing, then this change at least breaks the cycle.

(We'd probably have to make carve-outs for a few specific fields like doctors and lawyers and architects in order for this to pass.)

The dream of "universal college" leads to disastrous results. Lots of colleges -- particularly the ones at the bottom, that the lower class get into -- are under heavy pressure to simply become diploma mills, lowering their standards as much as they dare while still letting as many students as possible graduate. (Or, if they are the only school in town, keeping their rolls as full as possible.)

Using a degree as a credential poisons the entire educational system. Students become "customers" who are buying a grade or a degree. The teachers are supposed to give the student a grade as an honest evaluation have that process twisted because that evaluation is using by third parties. The best schools don't see this problem at all, because they still have elite students fighting to get into them.

Schools should be about education. There is still a place for credentials, but that system should be entirely separate. To a first approximation, imagine AP classes and AP tests. The AP classes tend to be very good, but the AP test, which is handled by an outside organization, is the real credential. Keeping the systems separate helps both systems stay pure to their true goals.

[rlitan]

Dan, you're absolutely right about college, and many are not suited for it, but at the very least, we should have everyone finish high school.

One of the problems getting there, though, is that what used to be voc-ed track in high school -- something that prepared graduates for real jobs -- is disappearing. It needs to be reintroduced and upscaled so that today's high school students who are not going on to college have a reason to stay in school. Perhaps if teachers and principals were paid, in part, based on high school completion rates,this would give incentives for them to lobby school boards to reintroduce upgraded voc-ed tracks in the schools.

[CK]

The ultimate causes of the gap are IQ and people skills. Neither of those are taught in college.

College graduation, like earning high wages, must be caused by IQ and people skills.

I'm continually appalled that economists (even Chicago School) think that more people going to college => more high earners => more wealth and growth and maybe even less inequality.

These economists refuse to believe in IQ because they are too PC.