Bryan Caplan  

Why Is the National Return to Education So Low?

Jeff Ely on Price Controls Dur... Hayek and Hurricanes...
Like Garett, I'm a huge admirer of Lant Pritchett's "Where Has All the Education Gone?" (World Bank Economic Review, 2001).  My favorite part of the paper is when Pritchett presents three stories that might explain his results:
I discuss three possibilities for reconciling the macro and micro evidence and explaining the differences across countries. The first possibility is North's (1990) metaphorical piracy: Education has raised productivity, and there has been sufficient demand for this more productive educated labor to maintain or increase private returns, but the demand for educated labor comes, at least in part, from individually remunerative yet socially wasteful or counterproductive activities. In this case, the relative wage of each individual could rise with education (producing the micro evidence), even while increases in average education would cause aggregate output to stagnate or fall (producing the macro evidence). The second possibility is that expansion of the supply of educated labor when demand is stagnant could cause the rate of return to education to fall rapidly. In this case, the average Mincer returns (Mincer 1974) estimated in the 1960s and 1970s overstated the actual marginal contribution to output from educational expansion in those instances where the demand for educated labor did not expand rapidly enough. Third, schooling quality may be so low that it does not raise cognitive skills or productivity. This could even be consistent with higher private wages if education serves as a signal to employers of some positive characteristics, such as ambition or innate ability.
Conventional labor economists predictably gravitate toward the straightforward "low-quality schooling" explanation.  It's the easiest to reconcile with their human capital extremism: "Sure, schooling is great for building human capital, but it's got to be done right."  The whole problem is incompetent, corrupt, and/or under-funded schools - the very kind that so often dismay development economists when they visit the Third World.

Before you buy the conventional take on Pritchett's results, however, you should ask yourself, "Does the low-quality schooling theory have any other predictions?"  Indeed it does.  If the problem is low-quality schools in the Third World, the private return to education in the Third World should be low as well.  For human capital extremists, schooling increases income - national and private - by teaching useful skills.  So if low-quality schooling fails to boost national income, it should also fail to boost private income.

This prediction is the opposite of the truth.  In low-income countries, the private return to education is unusually high.  As Hanushek and Woessmann explain in "The Role of Cognitive Skills in Economic Development":
The [private] rate of return to education is centered at about 10 percent with variations in expected ways based largely on scarcity: returns appear higher for low income countries, for lower levels of schooling, and frequently, for women... (Journal of Economic Literature, 2008)
This finding also contradicts Pritchett's second explanation - oversupply of education.  If there's too much education on the market, it should raise neither national nor private income.  This leaves just one and a half of Pritchett's stories standing:

1. The rent-seeking story: Education successfully teaches socially wasteful job skills.

2. The signaling story: While education teaches few useful job skills, strong academic performance convinces the labor market that you've got the Right Stuff.  (Later in the article, Pritchett points to evidence against the Pure Signaling Model.  Unlike most labor economists, though, he's careful to distinguish this polar case from empirically sensible signaling stories).

I've obviously spent a lot of intellectual energy defending #2, but I'm open to the possibility that #1 is a big deal, too.   My main doubt: Rent-seeking is intensely practical - and most of the curriculum is far too impractical to train students to do anything in the real world. How on earth do history, higher mathematics, foreign languages, or Shakespeare make you a more effective lobbyist or bureaucrat?   I just don't see it.  The same goes for "soft" college majors.  While carefully selected political science classes might enhance your rent-seeking ability, most of their coursework is, as usual, purely academic.

Bottom line: The signaling model is by far the most promising explanation for Pritchett's results.  Signaling fits every students' first-hand experience.  Signaling fits the global data.  The only thing signaling doesn't fit is conventional labor economists' love affair with education.

COMMENTS (15 to date)
Doug writes:

Will we know education credentials certainly are quite helpful for at least one type of rent extraction, being a professor or administrator of the state-funded university system.

Kiran Pai writes:

That is because conventional labor economics do not use data available from India. I would say nearly all Economic errors of thinking is because they only think in terms of America. Of course that is why I am hoping to sell my India dataset for millions of dollars some day. People who are interested in winning Nobel prizes should contact me.

India has two supposedly awesome institutes. As in the only reason people are so hugely successful from its two Prima Dona colleges IIT and IIM is because they only admit students with high IQs and in case of IIMs students who are also better in communication skills. There are more than 10 colleges which have better Professors, better facilities etc etc. But not a single soul in India would even bother going to them instead of the IITs and IIMs. Remember such data is not available in US. Because nearly all top ranked colleges in India are indeed better in quality too, broadly speaking.

I mean such evidence from a country of 1.2 billion human beings! Not enough to prove that education today is mainly about signaling ? Only a fraud wanting to "signal" his liberalness would actually value education for its non signaling effects.

Steve Sailer writes:

There is probably some intriguing longitudinal data from various advanced countries comparing the ability and performance of conscripts in World War I and World War II. America elites, for instance, were shocked by how uneducated, even illiterate so many draftees were in 1917-18. The conscripts of 1941-45 proved more satisfactory, and a lot of that improvement was attributed to the steady increase in average years of education between the Wars.

Steve Sailer writes:

Thinking about WWII raises the point that a lot of education can be private and informal. Early in the war in the Pacific, the Japanese enjoyed at least rough technical sophistication in the high-tech aircraft carrier combat. But the Japanese never recovered from their losses at Midway and subsequent attrition. In particular, they couldn't train enough pilots and mechanics to make up for their losses, while America could churn out vast numbers of technically proficient warriors.

A major reason was that only part of Japan had been much mechanized before the war. They non-mechanized parts of the Japanese workforce tended to be good craftsman, but skilled with medieval technology and medieval materials such as wood. The Japanese couldn't figure out how to train enough adults to maintain modern machinery. There's a certain age (28? 32?) after which learning new technologies is increasingly hard and the Japanese couldn't convert people to the high tech of the day fast enough.

In contrast, the U.S. on 12/7/41 had tens of millions of "shade-tree mechanics" who had experience working on internal combustion engines in Model Ts and the like. They'd educated themselves in the basics of modern machinery. You could rapidly train guys with experience with cars to work on planes or similar technology.

Arthur writes:

In third world countries the theory that the university is an elite club where you build contacts that you'll use to rent seek in the future seems especially plausible.

Pave Low John writes:

Odd that you would mention history or foreign languages as useless skills for lobbyists or bureaucrats. I can think of quite a few ways knowing some history (economic, military, political, etc...) would help a person in those positions.

Same goes for foreign languages. Who is positioned to be the better lobbyist, the one who speaks English or the one who speaks English and Spanish fluently? Then compare those two with someone that speaks English, Mandarin, Japanese and some Arabic. And yes, those people do exist, I've worked with them, they come in very handy when you deal with international stuff. Extremely handy, in some specialized circumstances....

Glen Smith writes:

How does the prediction work? I would expect that low-quality education would have a very high individual return with low national return moving to less individual return as the quality of education got better given a human capital model.

MG writes:

I think Brian may be understating the degree to which a puff education can be rewarded by rent seeking. I also think we could safely lump it with signalling. I see signalling and rent seeking as Plan A and Plan B for those inclined, or only able, to get more fluff than education. Plan A is when you were (already) smart, went to an elite school, and wanted the additional signalling to get the most productive (read, lucrative) jobs in the most unproductice sectors: acedemia, government and its lobbying, some forms of elite finance and law, etc. Plan B is for those without the certificates and correlates of those completing Plan A. It includes all the myriad functions that the thought leaders, culture setters, and policy makers among the Plan A'ers have convinced/coerced the rest of society that it needs: from more bad academics to middle level bureacrats to medirocre artists to diversity trainers to sustainable this and that...

Chris Stucchio writes:

Bryan, I think you should stop using the term "human capital extremism".

If this term becomes popular, far too many people who agree with you will use it to dismiss in favor of the human capital hypothesis without actually addressing them. Additionally, the people you describe as "human capital extremists" are likely to dismiss you in much the same way.

Why not use a more neutral term which doesn't divide people into "us and them"? For example:

"It's the easiest to reconcile with the purely/primarily human capital theory..."

Brian writes:

Is not the rent-seeking story what classical education proponents would claim? My understand of the argument of learning the classics, and classical languages is it trains one to a correct way of thinking and exposes you to a better culture (your more likely to change culturally) when you study these areas. It is this way of thinking and or culture that improves productivity.

When one studies and understands classical cultures and how these cultures understood human nature and understand classic history, it enables one to deal with the biggest problem in business "people" much more effectively. If ones fundamental viewpoint and though process of human society and it nature is more accurate then your solutions and work will have a much higher chance of successes. It is direction not intent that determines destination.

So if a educational program is teaching a fundamentally flawed concept of human nature, purpose, and intent/understanding of human society, then this education will become counter productivity.

Many communication specialist claim on their own individual languages influence thought process and therefore cultures themselves. There are even arguments that most of the world using using U.S. software changes third world thinking to a more American thought process which in turns has lead to increased productive.

If you don’t believe some cultures are better than others then off course you will most likely not see any value in classical education.

Another argument is because of rapid changes in the modern world flexibility and adapting to change are critical skill sets. A classical education allows one to be more rounded therefore is much better equipped to deal with a more broader set of situations, process, institutions, ect. (I personally don’t think this is right but it is an argument).

I am really just repeating the arugments I have heard Victor Davis Hanson make so you might want to read some of his arugments for classical education.

How on earth do history, higher mathematics, foreign languages, or Shakespeare make you a more effective lobbyist or bureaucrat?

Rather easily I'd think. Better educated people are usually more pleasant to be around. They're more charming. I've always thought a lot of Bill Clinton's effectiveness in politics came from his earlier successes in the back seat of a Chevrolet.

Also, David Friedman is a proponent of the idea that a better educated populace is one that is better at rent-seeking, and thus less productive.

Floccina writes:

I saw a PBS show about education in the 3rd world and I was struck by the mismatch between what the students were taught in school and what would be useful in their societies. I am not surprised that the schooling does not help growth. If a child's mostly likely profession is goat herder why not try to teach them how to improve the output of the goat operation and to be a smarter consumer and maybe some basic healthcare? Rather it seems they only attempt to prepare the student for further education.

Mounik Lahiri writes:

I think the majority of these are abstractions of how the real world operates. There is as much merit in education promoting a more eclectic and democratic understanding of the natural and social world, as there is for education to be geared towards an emphasis on those skills that are required to succeed in the talent economy.

Societal outcomes of education are not just those which are apparent in the short run but those that shapes society in the long run from the returns accruing from investments in human capital in the long run. What an economist should never get into is trying to decide possible social outcomes of seemingly selfish actions. Economics is for an understanding of disjointed reality in a one dimensional "cetiris paribus" condition whereas in reality there are thousands of variables playing havoc to such uni-dimensional theorizations.

Having said that one needs to use the tools provided by economics to better analyse possible side effects of decision and then take a normative decision on where they stand on the possible impacts of policy or investment, to the extent possible by a well researched perspective.

daubery writes:

"There is as much merit in education promoting a more eclectic and democratic understanding of the natural and social world"

This is an argument that is almost never made by politicians when they are trying to justify education subsidy and yet is almost always made by 'intellectuals' themselves. Granting that this still gives it some importance, I really think it's an anti-argument.

1. How to teach people to be better or "more informed" voters is surely a matter of opinion. What would the average gender studies professor say people should be taught to make them better democratic citizens vs a sociology professor vs a masonomics professor? What we're really talking about here is indoctrinating people in what their holders believe at better. Even starting from the premise that democracy is a good thing and we should try to make it work better is an ideological assumption of sorts.

At best, this is allowing the government to pick and choose what sort of ideological indoctrination to subsidise. At worst, it guarantees that the views of a fairly narrow interest group - university staff - gets to indoctrinate people in whatever they happen to like.

2. Beyond what is explicitly taught, there is the 'culture' of education. Does it really engender individual reasoning and critical thinking? Or does it teach hoop-jumping to please an authority and get a reward? Even in the West, schooling more closely resembles a conscript army camp than it does the society it is training people to live in, and that only partially diminishes at university. I don't want society to look more like school, and I think that that is precisely what insisting on more school does.

Bill Conerly writes:

Wouldn't a good signalling system raise macro productivity? Getting the right people in the right jobs should be a valuable activity.

If signalling is not successful in getting the right people in the right jobs, then firms that eschew traditional signalling (hiring drop-outs for managerial positions) could cut their costs with no reduction in firm productivity.

Or am I missing something?

Comments for this entry have been closed
Return to top