Arnold Kling  

The Status-Good Model of College

The Future of Online Education... Kahneman on the Crash...

Bryan correctly characterized my views. However, to me they imply a tipping equilibrium. Something that is a status good in one era can be the opposite in another. Think of smoking, for example.

At some point, there may emerge a generation for whom having their children attend an elite college is not a status good. Maybe today's twenty-somethings will turn out to be that generation. In any case, once that happens, the whole industry will collapse and not recover.

Predicting when that will happen is difficult. It's like predicting any tipping point, such as the tipping point at which an asset bubble pops or the tipping point at which people no longer believe that a government will pay its debts. Once the equilibrium tips, the employer prejudice might tip away from those who attend brick-and-mortar colleges, so that their wage premium could turn negative.

Comments and Sharing

COMMENTS (9 to date)
Joe Cushing writes:

I think once society gets sick of this notion that a young person has to start off life with 10s of thousands of dollars of debt, and very little to show for it, people will begin to hate college. I think college hurts productivity. It delays entry into the workforce by 4 to 10 years--more and more often by 6 years--and does little to make people more productive. It does very little compared to a corporate training program. So few people do what they went to school for. This tells us college is mostly useless. I have 7 years of college, a 2 year, a 4 year, and a masters degree, and none of it ever got me anywhere that I couldn't have gone with a high school diploma and a little sense from the company I was working for. I gave up 7 years of my time for nothing but a signal. I spent 4 years doing a military career I never wanted in order to be able to afford this college. So that's 10 years worth of effort (some of it overlapped) and a lot of tax money to achieve nothing special. To put icing on the cake, I'm unemployed and trying to break into the field I graduated in 3 years ago.

It turns out that businesses value one thing more than an education, they all want to hire somebody who has already done the job for 2 or 3 years. They don't want to invest in anybody.

Wallace Forman writes:

"However, to me they imply a tipping equilibrium. Something that is a status good in one era can be the opposite in another."

I agree. I am not sold on Bryan's signalling story of online education.

Once upon a time Greek and Latin used to be an important part of the curriculum. In Bryan's story, they could never have dropped out of the curriculum, because students who failed to take these courses would not have been signalling their intelligence, conscientiousness, and conformity. Yet here we are, in a world where the highest paying degrees appear to be engineering, math, and science degrees with probably very little classics instruction.

If Harvard decided tonight that all of its classes would henceforth be online, would it make sense for businesses to infer that its identical student body had failed to signal "a combination of intelligence, conscientiousness, and conformity"? What if Harvard (more reasonably) gradually transitioned to an increased reliance on online education over the course of three decades, while maintaining the same admissions/tuitions/grading structure?

Wouldn't the proper inference be that the Harvard students were still the same group of highly intelligent, conscientious, and conforming people they were before the switch/transition, while students at low-end online schools were not? Signalling is currently institution specific. There doesn't seem to be any clear reason to think that online education ought to be any different.

GU writes:
"in a world where the highest paying degrees appear to be engineering, math, and science degrees with probably very little classics instruction."

On average, STEM pays off the most in the short run, but not necessarily in the long run. Saying that STEM pays the most totally ignores Ibanking, private equity, hedge funds, and management consulting, which pay more in the short run and long run than STEM jobs, but are only open to students from elite universities. But even public accounting firms (the big ones) start newly graduated kids at $50k-$60k, with lots of room for advancement and raises, and accounting cares little about academic prestige.

My STEM friends constantly complain about hitting an earnings plateau in their 30s, usually in the low $100s. Not a bad living, but if you're really ambitious and driven to earn more money, it's not ideal. Making a lot of money while using one's STEM training requires becoming an entrepreneur (which most people don't want to do), or earning an MBA and going on more of a business management track.

There are a lot of philosophy/classics/English/etc. majors that go to law school and make more money than STEM folks. Or they get Ph.D.'s and become professors and receive huge non-pecuniary benefits (no boss, 4 months of vacation a year, etc.)

The point is that just because holders of engineering B.A.'s earn the most *on average* right out of college does not mean they earn the most over a lifetime. Moreover, students from elite schools, or even strong regional schools, can often make more than STEM grads without a STEM degree.

And I hope I don't need to say that many people do not attend college purely or even mainly to maximize their monetary earnings.

Wallace Forman writes:

My actual point is that people don't study Latin much anymore.

RPLong writes:

For all the education modelling here and on Marginal Revolution, I'm still not sure I understand what the core disagreement is. Why does it have to be all one thing?

Y = b1x1 + b2x2 + b3x3 + ... + bnxn

Where xn is benefit a particular individual derives from his/her educational experience, bn is the proportion of particular benefit out of the total benefit of education for any individual (Y).

The same good may mean different things at the same time to different people. What am I missing?

Becky Hargrove writes:

It's been hard for me to contribute to these discussions of education, signalling and digital learning because of my concerns for the big picture. A couple of points:

The existing model of education is what we utilize for what is also a consumer good in its own right: the desired knowledge based work or career. Now, we can make the model more efficient overall with digital learning, but this also shrinks our prospects of gaining paid knowledge work, for efficiency means pulling directly paid labor away from that same pool. To compensate, the (formerly) paid labor moves toward creating more knowledge product to sell. That's easy, although we seem to be overwhelmed with product we can choose the best of the best.

What's missing? A market was changed and partially removed with no new corresponding marketplace to capture knowledge. This matters because knowledge markets are the nuts and bolts of social organization. In other words, digital learning is but the next step of having no reason to get together at local levels. For all the flaws of public school, they were practically the last bastion of social organization in countless places. How do we get back to the direct use of knowledge at all levels for social organization? Let knowledge flower beyond formal schooling: now's the chance to make knowledge use real in ways it never enjoyed in the 20th century.

TimG writes:


I agree its an empirical question as to which majors pays more, and obviously in principal its possible for majors to pay off differently in the short term or long term. That said I completely disagree with most of your assertions.

Its safe to ignore ibanking, management consulting etc... because so few grads do those that it doesn't affect the average any non stem major. Its kind of like saying being an all-star basketball player is a good career.

There are a lot of philosophy/classics/English/etc. majors that go to law school and make more money than STEM folks.

Define a lot? I think would describe this group as very small. The Law school business model has been uneconomic for a long time and is in the process of disintegrating.

The advantage of stem degrees is that they lead mostly to careers that follow a normal income distribution (ie the 25% and 75% earners aren't that far apart). If you included the medical profession with stem degrees they look even better. Most of the careers you described (accounting excepted) have a long tail and you are thinking in terms of the extreme "winners" which is the disagreement - not short term long term.

Tom E. Snyder writes:

Wallace Forman writes:

My actual point is that people don't study Latin much anymore.

Et tu, Brute?

Anonymous writes:

The status value of college relates fundamentally to the connotation of being a scholar. The wage premium of knowledge is a distinct thing that is often acquired via college, but doesn’t have to be (think trade school, apprenticeships, self-taught, work experience, etc.). The status and wage premium values were traditionally inter-related because social status has, since antiquity, been the basis for assigning people to fill inherently scarce positions of power, prestige, and high income. That status plays such a role is not by design, but is a result of well documented assortativity property of social networking.


Being a scholar requires full time dedication to study that benefits no one else in particular, and therefore one cannot earn an income studying. Going back to antiquity this has consistently meant that being a scholar requires one to draw on an existing stock of wealth for a prolonged period of time, and therefore very much a function of social class.

Two things have changed this in modern times: The first is the notion of merit scholarship - which provides that such expense will be paid for by benefactors or the institution itself when the person in question has exceptional aptitude. The second is access to credit to pay for scholastic expenses.

Until the recent past, any higher education has been uniformly an NPV positive proposition. This is not because that is how college is supposed to operate, but rather was a consequence of circumstances. Because access to college was historically so severely restricted, the presence of a university education had been a key, if not dominant, factor in the transcending or affirming one's social standing (especially post-feudalism). Since it is axiomatic that the prestigious and remunerative positions in society define the elite, the university education gradually became kingmaker.

With the rise of merit scholarships and borrowing, the reliability of a university education in signaling social status has waned. Since merely having attended university no longer provides a strong proxy for class, determining pedigree is now best done by assessing what kind of university one attended. Universities can purify their link to the elite and purge impostors by 1) charging prohibitive tuition and 2) being unreasonably selective about what exactly constitutes merit. The reasons for the former are obvious, the later less so.

Being so gratuitously selective ensures admission requires a lifetime full of unrenumerated preparation on the part of a student and their family. These soft costs include such things as fees for and shuttling to and from extracurricular activities, monitoring grades, hiring tutors, forgoing (teenage) income to minimize non-academic distractions, and policing a child’s social influences to ensure all are positive reinforcements for the desired behaviors. Taken together these soft costs become a significant barrier over which wage-earning (hereafter referred to as peasant) families must hurdle. The more competitive and selective the institution, the less likely a peasant is to be adequately ‘qualified’.


Peasants can ill afford to fritter away precious little time and money they have for something that has negligible value. The result then, with the rise of merit scholarships and debt-funding, is a shift in focus on to whether the educational attainment pays for itself – whether the costs are justified by the wage premium value of the knowledge. This shift in focus did not parallel the rise in access, but rather far lagged it. Recall that perhaps as little as a decade ago obtaining a college degree – any degree – was uniformly presumed to be a good proposition. What was overlooked was that the strong link to social class, and therefore prestigious and remunerative positions in society, was what had made this true. There are not more of these coveted positions simply because more people went to college.

Realistically, the opposite has probably been true. Recent innovation in business, law, academia, and even medicine have commoditized and functionally leveraged large parts of what were historically non-tradable, upper-class professions. Finance, for its part, had seemed rather exceptional in that innovation had tended to create jobs. That property seems likely transient – a mirage spawned of leverage.

Absent a substantial wage premium, there is little reason for a peasant to consider college. After all, with 57% of those 25+ having attended college and fully 40% having a degree, college now asserts nothing about one’s membership in the elite, provides only a modest dose of social mobility, and in some cases may not even provide a wage premium. Questioning whether the costs are justified by the realizable benefits is a welcome breath of fresh air against the backdrop of industry self-congratulation, nauseating platitudes, and unsettling displays of blind, uncritical faith.


The effect exorbitant tuition and hyper-selectivity together create is to make the college proposition largely NPV ambiguous or negative (especially if hurdle rates reflect money utility, and are higher the poorer one is). Those who are admitted in spite of the apparent lunacy, then, tend towards those for whom time has little opportunity cost and money little marginal value. The social class signal-to-noise ratio for these types of universities remains strong.

But the most important thing to realize is that the status value of higher education is not the result of some grand design. It is predominantly the result of a phenomenon known as assortivity: the preference for a network's nodes to attach to others that are similar in some way.

To see this, first observe that it is axiomatic that people in power are the elite. Also, vacancies in positions of power are filled by those currently power. Finally, people prefer the company of others who are like themselves. Ergo, the elite naturally desire to fill vacancies in power with other elite. The trouble is that the ID badges for being a member of the elite (think country club membership) do not seem particularly relevant as qualifications for holding the exceptionally influential or remunerative positions in society. Pedigree of education, though, fits the bill perfectly.

Having graduated of a prestigious university is ostensibly a holistic measure of intellectual merit so incredibly distinctive that its graduates are often simply presumed to be qualified for any kind of position, however wildly tangential it may be to their formal training or actual capabilities. It is the perfect device for enabling the elite to exercise their assortative preferences, while feeling a genuine sense of propriety about it.

Comments for this entry have been closed
Return to top