Bryan Caplan  

Why Isn't the Military a Stronger Signal?

The Posner and Weyl Historical... A Recession Soon?...
Interesting question from EconLog reader Jonas Graham, reprinted with his permission.

Dear Mr. Caplan,

I read your book "The Case Against Education," well, as much as I could read for free off Amazon and Google Books, anyway. :P  I might buy it, eventually...

You say that college is the best way to signal to an employer that you are intelligent, conscientious and conform to social norms.  Other ways such as maintaining a Science Fiction blog, copying the dictionary or eating Kosher, despite being a goy, would just come off as "weird".

However, there is one way to signal that you are conscientious and conform to social norms, in fact *WAAAAY* more than college: enlisting in the military.  The military is very respected and very much not "weird", especially in the US.  Even in the bluest of "blue states", you will see USO lounges at airports, military discounts, etc.  Soldiers work hard, are very disciplined (for example, they have to make their bed every morning, even though it's stupid and there's really no point!) and do they ever conform to social norms!

Alas, it doesn't really signal intelligence, but, of the three, that's the easiest to measure: you could just have them do an IQ test.  The military, itself, administers an IQ test, of sorts, called the ASVAB.  Although it would be "weird" to put one's ASVAB score on one's resumé, certain military jobs are reserved for people with high ASVAB scores, for example: Military "Intelligence" and Psy Ops.  Wouldn't somebody with military experience in one of those fields also signal intelligence?

Have you ever come across any data comparing career success of college graduates vs. military vets?  I have found the following anecdotal account about how many employers still prefer a Bachelor's degree to military experience and the author talks about how military experience should signal the same things as a Bachelor's degree, but employers don't seem to "pick up" on these signals.

I have seen your interview on the Rubin Report, by the way, and I know you are a "pacifist" but, considering how big the US military is and how many vets there are, I wonder if you ever looked into this or just what's your take?

Perhaps you talk about this in the parts of the book I missed? :P

Anyway, huge fan!  Thanks for the enlightening read!


Jonas Graham

COMMENTS (22 to date)
Santiago Tórtora writes:

In my country people do say that military service is a good way to keep people off the streets, away from drugs, instill moral values and so on. This is used as justification for compulsory military service in a way similar to the justifications given in the First World for universal education.

It might be that first world colleges are just too good, and too much has been invested in them so there is sunk cost bias. Since colleges are terrible in my country, we don't have that bias and politicians seriously propose military service as a signal.

I think a look at education and labor markets in the Third World could be very enlightening in many ways. For example the way employers recruit from college students instead of college graduates could mean that admission exams are a powerful signal while graduation is not.

Jonathan S writes:

What percent of "high intelligence" military personnel never leave the military or public sector?

I also imagine the military is a strong for law enforcement and manual labor.

3rdMoment writes:

I followed your link under "anecdotal account," and didn't find much evidence of this claim. In fact the link contains this quote:

"Curious, he contacted the company’s HR department and learned that, other than an exception for military experience, all people hired into this position had to have a minimum of a bachelor’s."

So in at least some cases it would seem the military IS providing a signal.

jc writes:

@Johnathan makes an important point, imho.

The signal must appropriately match the receiver.

The message of the signal may, rationally, be a perfect fit for the attributes an employer claims to be looking for. But if the receiver can't/won't receive this signal in its undiluted form, it will be more noise (or even counter-signal) than signal.

Cultural considerations probably matter most. If society has, for whatever reason, chosen Signal X as its most socially accepted signal, an even better signal may be rejected. Algorithms that screen out online applicants based on certain criteria and/or rules that validate and thus only accept specific types of signals (laws/regulations/policies/etc., embedded within these algorithms or entirely separate) also matter. But these will, over the long term at least, generally simply reflect cultural norms.

The culture of law enforcement, though, probably welcomes the military signal. Thus, the contents of the signal - conformity, conscientiousness, etc. - are received as signal rather than noise, in its undiluted form.

With the proper receiver, for example, the evidence of "conformity" embedded within the signal is received and accepted. With a receiver programmed/calibrated to look for/desire a different signal (or signal source, e.g., a higher status source like a college rather than a source that is lower status in their eyes), the evidence of conformity is ignored, and the signal that is actually received is that the sender does not conform.

The polarity is effectively reversed. I send a message that says "X", but they hear "Not X".

David Sager writes:

Consistent with Bryan's thesis, a preference for degrees despite military service may be a function of the GI Bill, ie further subsidization. The military forces one to conform, it does not confirm voluntary conformity (merely signing up is a very weak signal in the noise of conformity signaling). The GI Bill, which not only pays for college but pays veterans to go to college as well, stands available at the end of service as a subsidized available voluntary conformity signal. Not conforming when being paid to conform stands as a stronger signal of voluntary nonconformity than the military signal of involuntary conformity. However, military service *with* a college education is a magnified signal of conformity.

This is the perspective of a veteran in a doctoral program.

RohanV writes:

In my experience, most (non-technical) people concern themselves mostly with the most common case, and don't really think of the uncommon or rare cases. If presented with the uncommon case, they'd think about it and agree.

College degrees are the common case. Military service is the uncommon case. If presented with a resume with military experience and no college degree, I think most people hiring would be willing to consider the military person just as much as the degree.

But the problem is that people writing rules or specifications for computer systems often don't think about exceptions, and just put down the common, "happy" path. That creates a problem with people/systems downstream who aren't allowed to deviate from those rules or specifications.

tOKEN writes:

Joining the military requires zero up-front cost, unlike college. For that reason, some people join the military as a way to get away from a bad home life as a child in which they have little in the way of parental support. That is not true for most people in the military but it is true of some. Hence, John Kerry’s famous “botched joke” that if you don’t do your homework and aren’t smart, you get “stuck in Iraq.” This makes joining the military instead of college a very bad signal.

Matthew Moore writes:

I think military service would be a stronger signal than high school completion for "blue collar jobs" - i.e. those that require a tonne of discipline, hard work and practical thinking.

It's not a good substitute for a college degree as a signal, because it doesn't really signal the ability to sit at a desk and write readable prose / meaningful spreadsheets / learn to code, which is most of what "college jobs" entail.

This doesn't contradict Bryan's thesis at all, which is that college degrees signal both conformity AND general intelligence

JVM writes:

Naive question but is it possible that military service is not a strong signal on its own because there is no filter? My impression is that there is a low bar for enlisting. Once you do you need to survive boot camp but that has an 85–95% pass rate, which is much higher than university admissions. After that you're not even allowed to leave for four years! So I assume the dropout rate is quite low. Really the only signal comes from the person being willing to enlist, which I'd say is highly ambiguous in its interpretation.

Compare to elite college admissions with <10% acceptance rate and students who are free to leave at any time, and it seems like there's no comparison in terms of filter (and by extension, signal).

As an employer, I would say that doing something in the military that I know is highly filtered such as Navy Seals actually would be a strong positive signal to me, probably stronger than an elite university. Advancement and promotion might also be a signal, one I wish I understood better. But time spent as a low-ranking enlisted person would be a weak signal for a typical high-paying job.

Yotam Gafni writes:

Military is like the anti signal for the open job market. It shows you gained all the wrong ideas about life - to give orders instead of gaining subordinates trust. To make plans in a system without competitors or good feedback mechanisms (aka money). In the army its all about power politics - you can justify the wrong decisions as inevitable as long as youre able to get your commanders in line. Almost all army high officials that go out to be CEOs fail miserably or get their position in a crony way, to use their army connections for contracts. I can elaborate much more but all of this is highly accessible and can be easily researched with studying army in my country of Israel.

MT writes:

Do we know that being in the military isn't a strong positive signal to employers?

gwern writes:

All the studies I know of which have decent controls for endogeneity like the Vietnam draft lottery show that military enlistment has no benefits or harms for career and lifetime earnings. One interpretation is that the military takes up too many years, and you never make back the lost compounding. Another is that it's not as good a credential in the first place - even when recruiting is easy, military standards tend to be pretty low. So if there's any signaling equivalent, you would have to be in a well-known elite specialty... but stuff like Seals or Rangers take *even longer*, are easy to flunk out of for random reasons (far more so than college), and may do a lot of damage themselves. (It seems like every in-depth profile of a special forces 'operator' I read mentions the legacy of body and health problems. You'd think they were football players.)

Liam McDonald writes:

A pacifist? I thought you were anti-war and not a pacifist per se.

Peter Gerdes writes:

I suspect the problem might be that military service doesn't genuinely signal diligence/conscientiousness. Since the military has such strong tools to enforce compliance not available to the average employer it's not clear doing well in the military will translate into the kind of conscientiousness necessary to do well in the private sector.

Jeff G. writes:

I think this is more evidence in favor of the signaling theory of education. As JVM pointed out, the military is easy to get in to and hard to get out of. So completing an enlistment doesn't work as a very good signal of some innate ability. On the other hand, it does develop skills that an employer should value (teamwork, work ethic, punctuality, etc.). And yet, relative to a college degree, it isn't valued as much by potential employers. That sounds like a point for the signaling theory and one point against the human capital theory.

Miguel Madeira writes:

"That sounds like a point for the signaling theory and one point against the human capital theory."

Only against the "soft skills" version of the human capital theory.

But, about the signal theory, perhaps could be useful to divide it in two sub-theories:

a) Intentional signalling - employees hiring people with college degrees because they think "if he has a college degree, these mean that he probably is intelligent, hard-working and conformists"

b) Unintentional signalling - People with college degrees being more productive because they are intelligent, hard-working and conformists, and employers hiring people with college degrees because they observe that they are usually more productive, but without loosing time to think if this is because some innate ability or acquired traits.

The "b" could be an explanation to why don't appear alternative signals - because employers don't explicitly think of college degree as a signal.

Jay writes:

I will note that the U.S. military pretty much requires officers to have attended college. Going to a service academy (West Point, Annapolis, etc) and serving as an officer is usually considered a strong signal of quality. Service academies don't charge tuition (though graduates are required to serve) so competition is intense. Going to college in ROTC* is also a much stronger signal of quality than enlisting.

*the Reserve Officers' Training Corps (or the Retarded OverTrained Chimps, according to grunts)

Pat writes:

Skipping college to enlist signals low intelligence to many people (I'm not one of them) and any attempts to offset that signal like including an IQ measure signal weirdness. They can't win.

Joe writes:

This is only somewhat related, but I always thought it interesting that you generally seem to be able amass responsibility younger and faster via the military than through the private sector.

19 year olds in charge of super expensive army helicopters without a college degree through the high school to flight school program? 22 year olds in charge of whole platoons?

Very unusual for private sector employees to be put in charge of so many expensive assets so early in their career, at least that is my impression.

Jeff writes:

I had seven years of enlisted service from 1975 to 1982. I saw no evidence that enlisted people are more intelligent than civilians, nor did they seem to be any more conscientious than the college students I met when I went to school later. Some were, and some were not.

Thomas Strenge writes:

Despite all the "thank a vet" propaganda, military service is NOT a benefit in the job market. People hire for a specific skill set and pass that on to HR in order to avoid discrimination. HR, often without fully understanding the job requisition, then looks for someone who matches the specific skill set as written. If you don't have direct experience or match the exact buzzwords, you're done. Gone are the days when people could look at your resume and say, "Gee, she can do this job." Combine that with general ignorance of what a military job is like, and vets are unhireable. Now, if you do have direct experience and veteran status, then it's different. But by and large, military service on your resume means nothing. I haven't been brave enough to replace yet with "joined the circus" for the relevant years, but I wouldn't be surprised if that actually improved my job search.

Phil writes:

Since about 75% of the population aged 18-25 are ineligible for military service (criminal records, overweight, no high school diploma, or physical disability), military service should provide a pretty decent signal that an applicant has already passed a screening test. So long as the veteran has an honorable discharge, they have demonstrated they can conform to the rules and regulations and adapted fairly well to a strong culture.


Return to top