How much of education and earnings variation is signalling? (Bryan Caplan asks)

On Twitter Bryan asks me:

Would you state your human capital/ability bias/signaling point estimates using my typology?

He refers to this blog post of his, though he does not clearly define the denominator there: is it percentage of what you spend on education or explaining what percentage of the variation in lifetime earnings?  I’ll choose the latter and also I’ll focus on signaling rather than trying to separate out which parts of human capital are from birth and which are later learned.  My calculations are thus:

1. I don’t wish to count “credentialed” occupations, where you need a degree and/or license, but you are reaping rents due to monopoly privilege.  It’s neither human capital nor signaling (it’s not about intrinsic talent), though you could argue it is a kind of human capital in rent-seeking.  In any case, let’s focus on private labor markets without such barriers.

2. Most capital and resource income is due to factors better explained by human capital theories or due to inheritance.  That is more than a third of earnings right there.  Note that the higher is inequality, the less the signaling model will end up explaining.  That is one reason why the signaling model has become less relevant.

3. Depending on job and sector, what you’ve signaled, as opposed to what you know, explains a big chunk of wages in the first three to five years of employment.  Within five years (often less), most individuals are earning based on what they can do, setting aside credentialism as discussed under #1.  Here is my earlier post on speed of employer learning.

Keep in mind that everyone’s wages change quite a bit over their lifetime and that is mostly not due to retraining (i.e., changes in the educational signal) in the formal sense, as most people stop formal retraining after some point.  The changes are due to employer estimates of skill, modified by bargaining power.  In this sense all theories are predominantly human capital theories, whether they admit it or not.

To be generous, let’s give Bryan the full first five years of income based on signaling alone, out of a forty year career.  And let’s say that on average wages rise at the rate of time discount (not true as of late, but a simplifying assumption and I think Bryan believes in a claim like this anyway.)

How much of income is explained by signaling?  I’m coming up with “1/8 of 2/3,” the latter fraction referring generously to labor’s share in national income.  That will fall clearly under ten percent, but recall I’ve inserted some generous assumptions here.

Bryan wants to call me “a signaling denialist,” yet I see signaling as still very important for understanding some aspects of the labor market.  But it’s far from the main story for the labor market as a whole, especially as you move into the out years.

That all said, this “decomposition” approach may obscure more than it illuminates.  Let’s consider two parables.

First, imagine a setting where you need the signal to be in the game at all, but after that your ingenuity and your personal connections explain all of the subsequent variation in income.  Depending what margin you choose, the contribution of signaling to later income can be seen as either zero percent or one hundred percent.  Signaling won’t explain any of the variation of income across people with the same signal, yet people will compete intensely to get the signal in the first place.

Second, in a basic signaling model there are two groups and one dimension of signaling.  That’s too simple.  A signaling model implies that a worker is paid some kind of average product throughout many years, but of course the reference class for defining this average product is changing all the time and is not, over time, based on the original reference class of contemporaneous graduating peers.  For the purposes of calculating your wage based on a signal, is your relevant peer group a) all those people who got out of bed this morning, b) all those people in the Yale class of 2012, or c) all those who have been mid-level managers at IBM for twenty years?  This will change as your life passes.

So there’s usually a signaling model nested within a human capital model, with the human capital model determining the broader parameters of pay, especially changes in pay.  The employer’s (reasonably good but not perfect) estimate of your marginal product determines which peer group you get put into, if you choose to invest in additional signals (or not).  The epiphenomena are those of a signaling model, but the peer group reshufflings over time are ruled by something else.  Everything will look like signaling but again over time signaling won’t explain much about the variation or evolution in wages.

Seeing the relevance of those “indeterminacy” and “nested” perspectives is more important than whatever decomposition you might cite to answer Bryan’s query.


This is a convincing analysis. However, the explanatory power of the signalling model is all about that first 3-5 years and 'getting in the door'. Perhaps they get sorted and paid their worth after some time, but I suspect that life time variation in earnings is too much of a "far" view to have much behavioral oomph on a recent highschool grad deciding which academy to become indentured to.

The signalling model turns post-secondary education into a positional good, leading to a mostly zero-sum arms race. Education engendering people with human capital is a rising tide that can raise all boats. Caplan is wedded to the theory in part because of his ideological proclivities viz. abolishing all education subsidies. Admitting that education is mostly about human capital introduces ambiguity, as it's not obvious that the positional externalities outweigh the positive externalities.

The more annoying thing is him calling anyone who disagrees a "denialist." That devalues a great word that should be reserved for anti-vaxers and creationists, where the truth is more forthcoming.

Someone said 'externality.' Everyone drink!

Look I think Tyler is being evasive here (also, water is wet). What this comes down to is whether the things that people do in school actually do or don't make them better at doing things that are useful. This is hard to measure, yes, but don't give up on the basic human power of observation! Almost all of us reading this went to college and we know what goes on there—mostly nonsense, with a side of networking. Every personal benefit people cite to going to college that falls outside of "gained skills or knowledge that _directly_ enhanced my ability to do things people want to pay for" counts in favor of the notion that college is bullshit. That includes networking and socialization and personal discovery and whatever else is used to justify the modern-day Mandarin system.

An old pal took a class at Harvard called 'Dinosaurs.' Things to note: (1) He knows jack about dinosaurs even after taking a Harvard class about dinosaurs; (2) even if he did, a knowledge of dinosaurs is not very useful; (3) He got an "A" because it's Harvard and that's all anyone gets; (4) After graduation McKinsey hired him for god-knows-how-much money, because he's smart and he knows how to carry water. But he was a smart water-carrier the day Harvard admitted him.

We don't need exhaustive empircal research or multi-tiered nested models or whatever the OP was about to figure this out. Just go visit a college campus, be it Harvard or Orange County Community. Resist the smell of rotting leaves in autumn and the nostalgia it brings, and look around. The business of "acquiring human capital" sounds like hard work—but very few of the students are engaged in anything that could be described that way! "Pass the bong—we have some time to kill before we collect the incomes we deserve!"

Caplan also can't deal with situations where signaling is productive and it's not possible to easily separate valuable signaling from necessary sorting for the purposes of enhancing human capital. Not everything is contained in an IQ test and some "arbitrary" hurdles are useful for both the employer and for the student who learns where he fits in the hierarchy of ability and educational capital acquisition.

I'm not sure Bryan thinks signaling plays an immense role in the labor market, but he certainly thinks it does in the higher education market. How is that possible? Well, the labor market is orders of magnitude larger than the higher education market. A small share of income attributable to signaling can still play an outsized role in willingness to pay for higher ed.

In plain English, what the hell does any of this gibberish mean?

Caveats and assumption abound, the education signal matters the first few years of ones career; not so much after that. You have to realize, for people in academia, this is their bread and butter. They're in a bubble. Thus, they think too much about it. As one prominent economist would say, just look outside your window every now and again.

Let me see if I can translate. Education signal = where you got your degree and what the degree was in; and possibly, your GPA. Academics live in a bubble, formerly known as an ivory tower. A prominent economist looking out a window = sees hundreds of students and thinks they are all marginal utility products.

There are many non-intellectual skills one might learn at a more elite college that employers find valuable. You'd be surprised the stock placed in being able to order a good bottle of wine, for example.

Indeed, the greatest skill I learned at college was how to drink like a gentleman.

I hope you're joking

Sorry to smash your hopes, but there are skills, especially social skills, you learn outside the classroom that prove important in many work settings.

Look, I'm not saying these skills are crucial or even an overriding factor. But it is possible employers are looking for a certain type of education and includes factors other than academics.

Ahh, but is knowing how to order wine just an effective signal? Or is it actual human capital?

Knowing how is human capital. Doing so is signaling.

I hear that Harvard Law School greatly increases human capital when it comes to the crease in one's pant leg.

First, I have several problems with Bryan's table. For instance how is this component of signaling different from human capital bias?

"(c) simply wasting time in ways that less productive workers find relatively painful, leading to a positive correlation between education and productivity"

In general differentiating between signaling and ability bias has to be hard. For instance do costs connected with membership in Mensa reflect ability bias or is it signaling? Anyways I think that you have the better approach. The best way to look at education and ask this question: is this level of education increasing or decreasing total social returns to education? And now this becomes very tricky. Because even having engineers studying philosophy may have some non-monetary positive effects in the same way as having an opportunity to find that 1 in a million guy among the students to select him for crucial general research.

Very interesting post, thanks.

I would add the model has a big escape hatch that needs to be figured in -- entrepreneurship. Customers respond differently than employers, in many cases they have no data at all about your education (or even who you are), they are usually focused in the quality of your product/service; there is still some significant signalling in a lot of circles but overall much less than in the employment relationship. That's how a lot of people who employers never valued highly earned incomes much larger than any employer was ever likely to provide, sometimes creating billons of dollars in personal wealth and a huge surplus to society. It seems like that ought to argue on some level that employment signalling exists and is inefficient, although of course there are other reasons for that discrepancy too (such as control).

I'm not sure that entrepreneurial billionaires are really all that common of a occurrence.

In Silicon Valley, signaling is still important when you rely on funding from venture capitalists and the like.

To the extent we have billionaires they tend to be entrepreneurs. You get some rentseekers like Gaddafi or Castro, monopolist oligarchs like Carlos Slim, but most rich economies have (almost by definition) strong institutions to prevent coercive wealth accumulation.

Silicon Valley represents a radical reduction in signalling -- in the past you didn't have twentysomethings creating hundreds of billions of dollars of wealth. They throw money at teenagers these days. Of course, it may be that value is easier to identify in software.

The customers might not care, but the creditors could.

Um, which part of your model does hysteresis go into?

"Within five years (often less), most individuals are earning based on what they can do"

How many banking top brass drove their companies into the ground, but continue to earn whatever they want because once upon a time they made a lot of money for some people (who happen to set their paychecks either directly or indirectly now)?

Your banking comment is not completely incompatible with the idea that people make what they deserved five years ago.

You underestimate the durability of the signal's effects. Your current employer may have a reasonable estimate of your performance and no longer heeds the signal, but when you change jobs the next employer does not have this information, and will have to rely on a mix of the original signal and an estimate of your current employer's assessment (mostly by asking you how much you make today). With job-hopping more prevalent than it used to be, this matters.

One other effect not discussed is the alumni networking & cliques you gain membership into. For someone undergoing, say, a Harvard MBA, it is perhaps the single most important component of the program's implied value proposition.

"Within five years (often less), most individuals are earning based on what they can do”
That sure explains why David Petraeus will be earning $150,000 to teach one course at CUNY. Just imagine what they'll be paying him in five years.

West Point + Princeton = Profit.

Terrific stuff. This deserves reiterating what Jeff Ely wrote several years ago, "Hail Tyler Cowen!"

In any case, let’s focus on private labor markets without such barriers.

The issue I would have with this, is that many occupations have soft credentials but not hard credential requirements. I work in pharma/biotech and many years ago most lab work was done by high school graduates, now you need a college degree to get hired. There is no requirement per se, but when an FDA auditor comes knocking you want to show credentials (it is the only thing they understand - seriously). I would also throw liability if you were to sued (again jurors will understand credentials). There are many other instances as well.

I take it that these discussions of signalling vs. human capital models of education are meant to apply mostly to liberal-arts education, right? I've been in post-secondary education for nine years now—four years of undergraduate Engineering Physics, and I'm now in my fifth year of graduate school in Applied Plasma Physics and Fusion Energy—and I would guess that probably 90% of the classes I've taken have increased my human capital, usually significantly. That is, taking the line from user ladderff's great post up above, these classes "directly enhanced my ability to do things people want to pay for".

Just thinking of my second-year undergraduate classes: (electrical) circuit analysis; vector calculus; solution of ordinary and partial differential equations; engineering thermodynamics. These were all things that I did not have the skills to do before I took the class, and I did have the skills to do after I took the class. Yes, I suppose there is a certain amount of intelligence or IQ required to be able to grasp the concepts, but that's a necessary requirement, not a sufficient one. Actually taking the class and mastering the material is also a necessary requirement.

To sum up, maybe this is just stating the obvious, but there is a whole world of engineering and science out there for which the education system is not simply putting a stamp on pre-existing skills or qualities in the person. It really is adding to human capital.

You, statistically speaking, are very unlikely to be doing any of that going forward unless you are lucky and love it (otherwise you are also unlucky). I wish I could find the paper, but it showed something like 100% of 30 or so biochemistry PhDs they looked at one year out of Harvard(!) were not doing anything in biochemistry, let alone a seamless extension of their education.

Bottom line, seems like we still don't know much about the underpinnings of the eye of the needle that we force everyone to squeeze through.

People all over the board are confusing terminology...when there is an existing terminology or they aren't making it up completely.

For the purposes of this discussion, are you talking about the signaling difference between having a college degree and not having one, or the signaling difference between Harvard and Michigan State?

If the former, you don't get the job in the first place. If the latter, that definitely helps you get the job, but like your high school grades, eventually fades into the ether.

Seems doubtful that earnings post 5 years into a career aren't correlated with earnings the first 3-5 years. The degree that gets your foot in the door often helps your entire career, even if you forget after 5 years that getting in the door was hard. The relevant variation is not among those with the same signal but among those with different ones. I know a lot of highly intelligent, highly foolish musicians in Austin who never got that degree and are now pushing 40. They can't suddenly wake up tomorrow and go work for a law firm, even though they might have made great attorneys. I myself worked service jobs through my 20's, finally got a BS, landed a corporate job that a monkey could do and eventually made big bucks, even though a monkey could do it. Signalling is everything.

Let's cut the Gordian knot and push for useful stiff to be taught in school. Learning useful stuff well and rapidly can be just as good a signal as learning useless stuff. Also at risk of weakening the signal let's loosen time limits on learning, so what if it take me longer to lean to do X as long as I learn, it I can do the job (that is as long as the job does not require rapid novel problem solving which many jobs do not and collaboration can help).

And if schooling is about human capital formation we should ask ourselves is there an easier cheaper way to build human capital.

Comments for this post are closed