Bryan Caplan  

The Case Against Education on Who You Know

Independence and Growth... ELECTRICITY!...
The noble Vipul Naik has been prodding me to address the social networking benefits of education.  Here's my first take on the subject from the current draft of The Case Against Education.
Who You Know

About half of all workers used contacts - relatives, friends, acquaintances - to get their current job.[1]  You could argue that education pays despite "low measured learning" because the measures focus on what you know instead who you know.  If you want to get ahead, studying is overrated.  Instead, the upwardly mobile student should strive to win friends and influence people.  The better your school, the more your friends and people can do for you after graduation.

This story has a kernel of truth, and may occasionally be dead right.  Overall, though, the case is weak.  The modern economy is vast and diverse.  Few of the students you meet will end up in your line of work - even if they share your major.  As a result, they'll probably never be in a position to help you.  If you're looking for a good job, you don't want generic contacts.  You want relevant contacts.[2]

Empirically, friends in your narrowly-defined occupation seem quite lucrative.[3]  So are older male relatives (father, uncle, grandfather) who know the boss or vouch for you.[4]  When researchers estimate the average benefit of "contacts" or "social networks," though, some find a positive effect on employment and wages, some no effect, and others a negative effect.[5]  If this seems implausible, bear in mind: Even if your cousin or college roommate plainly "got you your job," you might soon have found as good or better a job on your own.

Who does meet useful contacts in school?  If you want a job in education, school is the ideal place to network.  Once I decided to become an economics professor, I strove to meet other economics professors.  One of them, Tyler Cowen, got me my job.  (I also met many philosophy, history, and law professors.  Career payoff so far: zero).  If you're earning a professional degree in law or medicine, or majoring in relatively vocational subjects like engineering, you and your classmates could plausibly trade career favors down the line.  Stanford's computer science program could be a great entree to Silicon Valley.  At some elite schools, fraternities funnel their brothers into finance and consulting.[6]  Hell Week really could land you on Wall Street.  For most students, however, lucrative networking begins only after they graduate and find their niche in the sprawling modern economy.

Final point: Suppose making connections in school were the path to career success.  Would this rescue human capital extremism?  Not really.  The key claim of the human capital model is that school makes students more productive from a social point of view.  But much of the career benefit of "knowing the right people" - often known as "nepotism" - is redistributive.  If every applicant for a dream job has a contact, this doesn't mean that every applicant gets his dream job.  What it means, rather, is that applicants need better contacts than the competition.

You could object that friends work better together.  Yet every teacher knows that friends are more likely to cover for and distract each other; that's the rationale for seating kids boy-girl.  A cordial workplace is probably more productive than a friendly one.  You could argue that contacts provide valuable information, improving the match between workers and jobs.  But to make this argument, you must abandon human capital extremism and admit the central premise of the signaling model: worker productivity is hard to discover.  Once you admit that informal information from school peers pays off, how can you deny that formal information from school teachers pays off?  The "who you know" story ends up as a version of the signaling model of education - not an alternative.

[1] See e.g. Mouw, Ted.  2003.  "Social Capital and Finding a Job: Do Contacts Matter?" American Sociological Review 68: 868-98; Ioannides, Yannis, and Linda Loury.  2004.  "Job Information Networks, Neighborhood Effects, and Inequality."  Journal of Economic Literature 42: 1056-93.

[2] See e.g. Obukhova, Elena.  2012.  "Motivation vs. Relevance: Using Strong Ties to Find a Job in Urban China." Social Science Research 41: 570-80.

[3] Mouw 2003, "Social Capital and Finding a Job," pp.882-8.

[4] Loury, Linda.  2006.  "Some Contacts Are More Equal Than Others: Informal Networks, Job Tenure, and Wages."  Journal of Labor Economics 24: 299-318, esp. 308-12.

[5] See e.g. Mouw 2003; Ioannides and Loury, 2004; Pellizzari, Michele.  2010.  "Do Friends and Relatives Really Help in Getting a Good Job?"  Industrial and Labor Relations Review 63: 494-510; Loury 2006; Obukhova, Elena. forthcoming.  "Do Job-Seekers Benefit from Contacts?  A Within-individual Study with Contemporaneous Searches."  Management Science.

[6] Marmaros, David, and Bruce Sacerdote.  2002.  "Peer and Social Networks in Job Search."  European Economic Review 46: 870-9.

Comments and Sharing

COMMENTS (12 to date)
AS writes:

Wait a minute.

Why isn't networking treated as a market failure of education? It, like signaling, has private benefits for the individual that don't carry over to benefits for society. This should be part of the argument *against* education subsidies, not for it.

So why does Bryan talk about addressing the "social networking benefits of education"?

K Lakshmi writes:

I agree entirely.

Although at times the school you go to can really help land your job, it really depends on many other factors. To assert that education can be a place to accelerate your career path through networking is a weak argument. If people were getting hired solely on the standard of their school and connections, [basically if the economy valued networking through education] it would boil down to who has the better connections, as Caplan points out... It is quite bizarre to visualize education's primary focus as a network builder.

It is interesting to speculate, however, how much impact social media has helped accelerate the process... What percentage of graduates land their first job by the help of somebody else and through what means? As a college student myself, I know several students who have gotten jobs through Linkedin networking. Is frictional unemployment decreasing because of networking? These statistics would help either side of the argument...

Peter H writes:

Few of the students you meet will end up in your line of work - even if they share your major. As a result, they'll probably never be in a position to help you. If you're looking for a good job, you don't want generic contacts. You want relevant contacts.

When you enroll in a traditional 4 year college at 18, you generally lack a line of work. So contacts in a large variety of fields may give you a large number of places to seek initial out-of-school employment, in a variety of fields. Initial employment out of school is an incredibly important determinant of lifetime earnings, which is why students who graduate in recessions show persistently lower lifetime earnings.

I'm not arguing this is socially desirable in the aggregate, but from the perspective of an individual student, more contacts means a greater menu of plausible choices out of school.

Lakshmi K,

On social networking websites, people who have similar interests often network, and I think this would bring down frictional employment a lot. :)

Ghislain writes:

I think networking to find a job was a subject of an econtalk podcast.

The summary was: better have lots of loosely related and varied connections than really close friends.

The reason is simple:
The goal of the network is to be aware of opportunities, not helping you get a specific position.

Let's imagine I have a friend and a position is open at my company. This friend has the right "education" but is not really good. Should I risk my credibility at my company to promote him? I would not. And if he is really good, he does not need my help.

William Bruce writes:


The subject was addressed in the podcast with Newman on low-wage workers, although that may not be the podcast you have in mind.

Glen S. McGhee writes:

Steve McDonald studies this at Chapel Hill. Somewhere I learned that after 50 yrs, the importance of social ties increases, gobbling up the majority of successful initial job contacts. So, there is a temporal effect.

As for social ties for new grads amounting to anything, however, I have my doubts. For one thing, if there are no jobs, then it doesn't matter who knows someone.

My niece graduated a prestigious school, but we started to worry when the grad update newsletter stopped listing the jobs the kids had -- in fact, she ended up back in school. And this was after spending a mint at boarding schools to get her in.

Long story short, I used to believe in the effect -- once I learned about the importance of structural holes (Burt) and all that, but now I'm not so sure.


Many employers in my country in the industry I work in (The media) will not even look into the resumes of people if someone they know does not refer them. This is especially true for college dropouts like me. I got my first two jobs without any networking, but I got my third job because I knew an ex-employee. But, I knew this ex-employee only because he had read my work, and liked it.

Glen S. McGhee writes:

Steve McDonald's work on this: Vita and Publications.

Bostonian writes:

Here is an anecdote showing that graduates of certain colleges raise the credibility of their employers, from "Whitney Tilson: Human qualities in management matter for long-term success" in the Financial Times :

“[Hedge-fund legend] Julian Robertson was maniacal on the importance of management: ‘Have you done your work on management?’ Yes, sir. ‘Where did the CFO go to college?’ Umm, umm. ‘I thought you did your work?’ He wanted you to know everything there was to know about the people running the companies you invested in.” Lee Ainslie, Maverick Capital


Since admission to top schools is somewhat arbitrary, I wish school prestige mattered less.

James Wilson writes:

"A cordial workplace is probably more productive than a friendly one."

Two anecdotes: Penn Gillette said he and Teller don't socialize much away from work. And Mike and Mike of the ESPN talk show have said the same. Penn & Teller have been around some 35 years; Mike and Mike for 13.

I suspect both duos stayed together for so long for the very reason that they're not close friends. Their disagreements, then, don't get personal and fracture the professional relationship.

Steve Sailer writes:

Higher education serves more as a filter in the mating market, which is extremely important.

Comments for this entry have been closed
Return to top