Bryan Caplan  

IQ and Hiring: Does the Law Matter?

PRINT
Worstall on Cyprus and Russian... Krugman Gets Hoover Wrong...
While investigating the legality of hiring based on educational credentials and IQ tests, I came across an interesting old post by Half Sigma.  Why credentialism is safe but IQ tests aren't:
[I]t's OK to use college degrees as a hiring requirement... because everyone knows that college is really important and the reason it's not OK to use IQ tests is because everyone knows that IQ tests measure only the ability to take the IQ test.
The italics are Half Sigma's, and signify sarcasm.

Half Sigma goes on to argue, though, that the law makes little difference:

I agree that hiring based on degree does have the effect of making it more likely that those with higher IQs get into the better career tracks. But is this the reason why employers look at degrees? Because they would rather use IQ tests but aren't allowed to?

The average educated person, trained in their beliefs by the liberal media and liberal college professors, believe that IQ measures nothing but the ability to take an IQ test. Employers' beliefs are no different than those of the average educated person. If employers are not using standardized tests, it's not because they are legally prohibited, but because they don't believe in them. The laws usually reflect the beliefs of the majority. If there was a huge groundswell of support for using IQ testing, the laws would be changed.
On the surface, this seems totally reasonable.  The key problem, though, is that on the market, unpopular true views naturally spread via selection and imitation.  If IQ tests are really better employment screens than education, then the few employers who hire based on IQ gain a big competitive advantage.  They survive and grow, their flexible competitors copy them, and their rigid competitors shrink and die.  That's one of the reasons why markets are better than democracy.

If this sounds like mere econo-dogma, consider the rise of index funds.  In the 60s, a few academics noticed that dart-throwing chimps could match the performance of fancy investment managers.  Most people thought these academics were crazy.  Since the academics were largely correct, however, the few people who listened to them got rich and revolutionized their industry.  If index funds were illegal, the majority would never have gotten its richly deserved comeuppance.


Comments and Sharing





COMMENTS (36 to date)
JKB writes:

I think you'd do better thinking of IQ disliked by the "elite" rather than the majority's adverse opinion. Much like Peter Hitchen's recent confession regarding his Revolutionary Marxist support for immigration. I would suspect something around the 'elite' stereotyping minorities as being unable to do well on IQ tests. Since most 'elite' don't run businesses that use low skill workers but they don't feel the impact. But many do run enterprises where the college degree, even if not required by level of tasks, it does add prestige to the enterprise.

Testing attracts state employment agencies, lawsuits, the EEOC, etc. All driven by 'elite' operators pushing the agenda. Even if the employer wins and justifies the use, he loses, since defense is costly, distracts from running the business and can destroy any profits.

Steve Sailer writes:

How discrimination laws are applied differs wildly from industry to industry. For example, the movie and television business in Los Angeles engages in disparate impact discrimination against Latinos to a palpable degree. The blue collar jobs on movie crews are overwhelmingly filled by white men even though Latinos fill most other blue collar jobs in Los Angeles County.

HH writes:

The key problem, though, is that on the market, unpopular true views naturally spread via selection and imitation. If IQ tests are really better employment screens than education, then the few employers who hire based on IQ gain a big competitive advantage.

A few point:
1. IQ tests are probably better at testing for pure ability, but education is far better at measuring conformity and conscientiousness, which are both traits employers value.

2. Assuming the differences in #1 offset (unlikely, but for the sake of argument), education is roughly equal as an employment screen to the employer. Education, however, is free to the employer (IQ testing costs are probably low as well). The employer has no real incentive to make the first move to IQ testing. However, education is vastly more expensive and wasteful to the employee, but the employee doesn't have the ability to convince employers to use the IQ test as a proxy.

Martin writes:

"If IQ tests are really better employment screens than education, then the few employers who hire based on IQ gain a big competitive advantage."

Employers might believe this, but as long as their customers do not, then employers will hire people with a college degree as opposed to people with an impressive IQ.

For as long as customers do not believe this to be true, then no employer will risk it. The lower price charged by the employer to the customers can reflect either cost-savings or quality differences. What do customers believe?

If you look at the websites of professionals providing services you will often find listed the degree and the school. This holds for independent professionals and those that have organized themselves in a firm or are hired by one.

How many students would go to GMU if the website listed "Bryan Caplan has an impressive IQ" versus "Bryan Caplan PhD econ from Princeton"? Similarly if you would be an economic consultant?

Patrick writes:

The military, which is practically immune to prosecution, administers IQ tests to all recruits. They seem to know something.

Glen Smith writes:

Patrick,

The military also uses several other items that are indicators of and conformity. In fact, in the military, indications of conformity then conscientiousness are the more important variables. The IQ testing is more "tie-breaker" than anything else.

Finch writes:

@HH

> Education, however, is free to the employer

Education is not free to the employer. Try hiring someone with an MIT degree and someone with a UCSB degree and see if there's a difference. Try hiring someone with an MBA and someone with a GED. Part of the cost is born by the employee, but nowhere near all of it. If it was entirely born by the employee, noone would ever bother with education.

A firm hiring on the basis of IQ tests could hire people with less education, and pay them less. However, your first point seems correct that IQ is not the only thing schooling provides evidence for.

Mike W writes:

But...IQ tests only a single attribute, why would employers rely on so narrow a source of information about an applicant. Education...the school attended, major, class standing, grades, extracurricular activities...provides much more information for the hiring decision. IQ says nothing about motivation...which is probably more important (except in the hiring of college professors) to an employer.

@ Patrick

The military, which is practically immune to prosecution, administers IQ tests to all recruits. They seem to know something.

Apparently you were not in the military. (Sorry, couldn't resist)

ChrisA writes:

The very large multinational that I worked for used to require IQ testing as part of the approval process for promotion to senior levels. They had to abandon the idea. I had some of my team go through this process and to my surprise there was little correlation between capability and the test results. I remember one guy especially, PhD, well known in his field, author of many technical papers, testing as below average intelligence. How do you deliver that message that we can't promote you because you tested as below average IQ?

I would say that a college degree and demonstrated work experience is a much better judge of capability than a ten minute raven matrix test.

HH writes:

@Finch

I'm afraid I don't follow anything you said.

Education is not free to the employer... Part of the cost is born by the employee, but nowhere near all of it.

By the time a college graduate is applying for a job, he's paid his tuition, room, and board and invested four years into that education. The employer thinking about hiring him didn't pay for any of that. They bear no cost of an applicant/employee's previous education.

Try hiring someone with an MIT degree and someone with a UCSB degree and see if there's a difference. Try hiring someone with an MBA and someone with a GED.

I'm not sure anyone has claimed that these would be equivalent. I certainly haven't. One of the main points of Bryan's approach to education is that is mostly signals that people are smart, rather than adding human capital. It is perfectly consistent that different schools signal different things. This is undisputed, and I'm not sure what it has to do with my original point.

If it was entirely born by the employee, noone would ever bother with education.

This is just empirically wrong. In the present day, the cost of education is largely born by the (future) employee. (Ignoring scholarships and gifts and the like.) People spend time and money to go to school, with the expectation that this will raise their earnings potential enough to offset the cost of the education.

Clearly the voters have not been requiring IQ testing of those for whom they vote.

DrC writes:

It's not very difficult to gauge someone's intelligence from their educational credentials. Moreover, employers who require a high IQ can just ask for SAT, GRE, GMAT, etc. scores. Those tests correlate with IQ tests almost as well as IQ tests correlate with each other.

The question then becomes, is the extra information provided by IQ tests greater than or equal to the costs of obtaining IQ scores from applicants? I think the answer to this question is no for the vast majority of employers.

Finch writes:

People get paid more when they are more educated. That's how employers share the cost of education. It doesn't matter that they do so after the education has occurred.

An employer who could measure the qualities education supposedly measures via some other less costly way could hire people with those qualities but less education and pay them less because those attributes are not bid up in the market. This would be a profitable strategy until a substantial fraction of the market figured it out and started measuring those attributes in a more cost-effective way and paying more for them rather than for education.

Finch writes:

Put another way, when a student pays for education, he's buying something of value, human capital improvement, the Harvard admissions office stamp of approval, whatever. He then sells what he bought in the labor market. It has a positive price, in fact the price is much greater than what he paid for it in the average case. So, who bore the cost of the input? It's only true that it was the student in the simplest possible sense that the student physically paid. Economically, though, some fraction of that cost is passed through to the labor purchaser.

Dan Carroll writes:

Most jobs do not require a high level of intelligence to complete. And, yes, IQ tests only measure a very narrow slice of intelligence. Social skills ("Emotional IQ") are very important to success in management and sales, for instance. However, most jobs do require diligence and conscientiousness. A college education doesn't teach many specific skills of value, but students do undergo four years of constant testing on various levels. However, in some fields, IQ tests (SAT, CPA, CFA) are valuable. I don't believe the 4-year college degree is efficient, but I also don't believe that simple IQ testing is sufficient for most employers. However, 4-year college degrees are heavily subsidized, enjoyable for most participants, and important status markers for our culture. For it to be replaced, better testing, measurement and filtering would need to be developed. Which is why I think, over the long run, online education is viable - it provides the ability for constant measurement and feedback.

andy writes:

Economically, though, some fraction of that cost is passed through to the labor purchaser.

How does an employer pays a cost of a useless degree?

Swimmy writes:

I've applied to several jobs that require both an IQ test and a test on conscientiousness. They just don't call it that.

"Personality tests" are filled with measures of conscientiousness: Do you answer similar questions worded differently the same? How attentive are you to double negatives and other odd grammatical structures? Do you have the patience to answer the same question multiple times? Do you answer all questions?

These are frequently followed up by math tests. A math test may not be as robust as a full IQ test, but it is a weak IQ test nonetheless. It can be justified if the job requires significant math work.

I've even gotten word tests, with questions similar to SAT standards, like "X is to Y as Z is to ___." These can be justified if the job requires writing. Combine with a math test and you can't much tell the difference between a job application and the SAT, save for length.

If you want to see them for yourself, apply for a financial analyst position at just about any large corporation. Dish Network requires these tests, as did Frontier airlines and several others.

College is a better measure of many desirable behaviors, of course, but the fundamental assumption of this discussion is wrong. Whatever the law may be, many businesses DO make job applicants take IQ tests.

Finch writes:

> How does an employer pays a cost of a useless
> degree?

It depends on the employer's elasticity of demand for the useless degree and the employee's willingness to accept certain wages. I've worked places that required the receptionist to have a college degree, so they paid a little extra and got somebody with a degree in dance, thereby bearing some of the cost. The pass-through rate surely varies by major, with much of the cost of useless degrees born by the student and much of the cost of useful degrees born by employers. I haven't gone looking, but this is kind of obvious; there must be a literature.

If MIT started charging a dollar more for an electrical engineering degree, salaries for MIT electrical engineering grads would immediately jump by a large fraction of a dollar.

Mad Mat writes:

I think this post makes a good point, but I am unconvinced that the market reward for using IQ tests is large enough to have a significant effect on the survival of big companies.

Suppose Big Co. stops requiring college degrees and instead requires IQ tests. The IQ tests successfully select all the people who had good college degrees anyway. Who else does it select? Well, people who are smart and don't have college degrees. How many of those are there? I'd guess the vast majority of them are clever people with autism spectrum disorders, anxiety problems, etc. which prevent them interacting socially or make them unconcerned that they havent followed the enormous social pressures on them to go to college. And companies do not want these people for other reasons.

One other possibility is that Big Co. could try to attract people who intend to go to college at age 18 before they actually start the degree. But anyone clever enough to pass the IQ test is clever enough to know that accepting this offer puts a huge asterisk on their resume if they ever want to go and work for another company later in life. Not a good deal.

The ideal of course is that they pick up people who are clever and wanted to go to college but couldn't because they were born into a poor family &c. But in an age of extensive government subsidy for college, it's more typical these people go to "bad" local colleges rather than that they go nowhere at all.

I do not see it as a surprise or a mystery that companies do not adopt controversial IQ tests rather than nice-and-fluffy college degrees, given that we live in a world where almost everyone worth hiring has a college degree anyway. The trillion dollar question is, would companies adopt IQ tests if the government stopped subsidising college, ballooning the size of the intelligent-but-degreeless population.

Eric writes:

The military uses IQ testing to qualify people for specific jobs. Success after training will depend on social stuff.

The Army used IQ until it became illegal, then they just used the same tests and score but changed the name to GT. Now recruits are offered positions based on General Technical instead of IQ.

Bryan Willman writes:

I once had a manager from a consulting firm visiting the college I was at flat out say to an ACM meeting:

"We prefer people with master's degrees, because they are generally older and more mature."

It was thirty years ago, so that quote isn't exact, but it's bloody close.

This is part of the answer to the general question:
Why require a college degree?
Because:
(a) it's legal,
(b) there will be no 19 year olds and if there are they will be exceptional,
(c) this person has proved a level of "carry through to the end",
(d) it filters out many applicants and there are far too many anyway,
(e) the BEST graduates from GREAT programs ARE provably better than everybody else with comparable experience. it helps focus on them.

James F writes:

A college degree tells you a different mixture of things about a person, some combined metric involving their conscientiousness, intelligence, conformity, and probably other things. It's generally much more important to have employees who are ready and able to fit into arbitrary hierarchical task-based environments than to find the people with the highest IQs.

When employers do want to select strongly for intelligence, it's easy to do so while interviewing, in a way that also ensures they'll be able to have a coherent conversation about technical subjects (which IQ tests don't measure and is also critical to such jobs). Surely you've heard of what the interview process is like at Google and other such tech companies.

The Anonymouse writes:

Re: the military.

The military does IQ test, even if they call it something else (the relevant metric on the ASVAB is, as a previous commenter noted, called your GT score). But it is not my experience that they test for diligence and conscientiousness and use IQ as a tie-breaker. Rather, at least from the narrow field I was in (the infantry), the primary (and not inconsiderable) effect of that score is in job-selection: you either are selected for a particular job field (MOS) based on your score, or you are given a range of available job fields for which your score qualifies and in which they have a need. If you can select, a higher score opens more doors--someone with a low score is not going to be allowed into, say, cryptoanalysis or avionics repair--but it does not close any (a high score can still choose to go into the infantry* or field artillery, which negates some of the negative effects of those fields being available for the lowest scorers).

Naturally, the military values diligence and conscientiousness above all. They weakly test for these things through minimum entrance standards (were you able to procure a diploma or GED? do you have the self-control to avoid a felony?). More importantly, however, they have a time-tested and quite effective process of inculcating those qualities, rather than testing for them beforehand: boot camp. I gather this is because, at least among 18-21 recruits, it is much easier to teach self-discipline and teamwork than it is to raise IQ.

*Very possibly apocryphal, but I have heard many times that the infantry is a special case: they accept, as minimum entry standards, among the lowest ASVAB scores qualifying for enlistment, but also contain a substantial contingent of the highest-scorers. This is typically explained by the fact that those who ace the entrance testing probably have pretty good job prospects outside the military, and joined not to work a desk but to hump a ruck and fire a rifle and do the "adventurous" stuff you can't get in civilian life. This comports with my reasoning for choosing the infantry.

Bostonian writes:

Mike W wrote, " IQ says nothing about motivation...which is probably more important (except in the hiring of college professors) to an employer."

Professor Angela Lee Duckworth of U Penn has found that IQ scores are affected by motivation:

http://www.upenn.edu/pennnews/news/penn-research-demonstrates-motivation-plays-critical-role-determining-iq-test-scores

Penn Research Demonstrates Motivation Plays a Critical Role in Determining IQ Test Scores

Steve Sailer writes:

The military won't let you enlist unless you score at the 31st percentile or higher on the AFQT, a four part test that correlates so highly with tests designated as IQ tests that Herrnstein and Murray used it for much of The Bell Curve. The AFQT is a subset of the ten part ASVAB. If you score high enough on the AFQT subset, then the military looks at your ASVAB scores to see what you'd be good at. (For example, on ASVAB test is concerned with knowledge of vehicle repair.)

Mark V Anderson writes:

Many commenters have made a pretty good case that an IQ test would not substitute very well for a college degree, because such a test does not measure attributes like conformity, tenaciousness, etc. But I sure hope there is some kind of filter that does measure all these traits, but isn't so expensive in time and money as a college degree. Our economy currently wastes many billions of dollars every year on education for students who are only there to pass through the filter to a better job.

And yes it would be economically beneficial for any employer that discovered such a filter that could be used legally. It is true that most desirable recruits already have four year degrees, but there are still millions in this country that have the requisite attributes that haven't been able to get a degree because of their circumstances. Hiring such employees would result in much cheaper labor, plus labor that is probably very happy to have a professional job despite no degree. Such a firm would make a mint.

Himanshu Sanguri writes:

The author has made a decision that can not be generalized. Employer- Employee- Education is a very conditional framework. There are many factors like nature of job, ideology of employer, type of education degree like vocational or traditional, which plays a vital role in deciding the match. IQ tests are a sort of a filter machine to surface out the best among the worst :-)

Mad Mat writes:

"Many commenters have made a pretty good case that an IQ test would not substitute very well for a college degree, because such a test does not measure attributes like conformity, tenaciousness, etc. But I sure hope there is some kind of filter that does measure all these traits, but isn't so expensive in time and money as a college degree."


Yes, but this is a social benefit, not a private benefit to companies. The overwhelming majority of education expenses are borne by government.

Another problem is that most people are "sold" degrees by their teachers, parents, the media, etc. - not by employers. They don't conduct a thorough cost-benefit analysis and decide on a degree level education as a profit-making exercise, they instead believe (falsely?) that degrees will pay back no matter what.

Bryan might rejoinder that this gives people who do an advantage in the marketplace - and it does. But only in the sense that they end up with more savings or a higher income after loan repayments. It doesn't mean that they gain control of other peoples' decision making processes, like in his example of a more competitive company supplanting a less competitive one.

Finn writes:

Finland: IQ tests are not that rare when hiring educated people. Education premium is moderate, because college and university fees are zero and the students even get money for living from the state. Moreover, the law stipulates that a wage decided by a union with employers' organization is binding for nonunionized workers and companies too, on the same industry, if 50 % of workers are within the agreement. Therefore, part of the education premium shows in the lower employment rate of the unskilled.

In other aspects than labor market and tax rates, Nordic countries are relatively free, in the average.

Mark V Anderson writes:

Mad Mat --

First of all, I don't agree that the overwhelming majority of expenses are borne by the government. Haven't you heard of the enormous debts incurred by students? I think most of these debts are paid back, which is a cost to the student.

I also don't see why that even matters. Lots of folks pay for this cost, whether as taxes or payments to colleges. It would be a good thing to lower these costs, since much of these costs only benefit the economy as a very expensive filter of who gets what jobs.

As I pointed out in my previous post, an alternative filter would greatly benefit an employer who used this filter, who would get a cheaper and better workforce. And if many employers used this alternative filter, the economy as a whole would greatly benefit, as the college filter would be de-emphasized.

So now it is just a matter of creating this alternative filter, which of course is not a trivial exercise. No test can determine attributes such as motivation and conformity, because the test-takers will tailor their answers to get the job.

Richard writes:

Seems to me that when
a) The culture highly prizes educational credentials; and
b) It is easier to receive educational credentials the smarter you are, then

Those who have really high IQs and poor educational credentials are signaling something very undesirable about themselves. These include

1) Lack of interest in social standing, or so lacking in social skills that the person cannot figure out what society considers prestigious
2) Lack of conscientiousness or emotional stability. The higher the IQ, if the person didn't finish college the more likely that something else was seriously wrong or abnormal.

These beliefs are not unreasonable. A couple of years ago, ABC interviewed a guy who had an IQ of 200 and he was working at a bar in Wyoming. In the interview, he went on about how he couldn't get along with his step-father, he couldn't get along with others in college, etc. If an employer gives this guy an IQ test and he gets a score of 200, the employer will rationally wonder whether something else is really wrong with him.

You all seem to be going about this backwards. To quote Earl Hunt: "In economic terms it appears that the IQ score measures something with decreasing marginal value. It is important to have enough of it, but having lots and lots does not buy you that much. " (http://www.americanscientist.org/issues/issue.aspx?id=878&y=0&no=&content=true&page=8&css=..)

That is, IQ testing is not particularly helpful to employers who hire college graduates, except in finding diamonds in the rough. For example, most of the high ability, low income people are white men and women. It would be very useful to employers looking for smart people to find them, because "signaling" wouldn't necessarily apply--these are people who don't have thousands of non-profit organizations waiting to help them (http://educationrealist.wordpress.com/2013/03/18/why-most-of-the-low-income-strivers-are-white/)

Where IQ testing would be extremely helpful is in service industries: fast food, retail workers, and other job categories with high turnover that can no longer count on the high school diploma as a meaningful differentiator.

Michael Keenan writes:

I think the index fund argument opposes your point. Most average investors should be mostly invested in index funds, but they aren't. Most average HR managers should care about IQ tests, but they don't.

And of course there's the old problem of the incentives of the HR manager being different from the company. Suppose you're choosing between a Harvard graduate and a smarter, but uncredentialed, candidate. If you hire the Harvard graduate and she's terrible, you don't get in much trouble for that. If you hire the uncredentialed person and she's terrible, then you're an idiot for hiring her and passing up the Harvard person.

Ben Nader writes:
If MIT started charging a dollar more for an electrical engineering degree, salaries for MIT electrical engineering grads would immediately jump by a large fraction of a dollar.

Almost certainly wrong.

If the grads could negotiate for that extra dollar after the change in their education costs, they could get it before; they haven't gained any bargaining power or anything extra to offer employers.

The firms pay more because of the better signal & the better human capital investment of a more expensive school, not because it's more expensive.

Finch writes:

Come on Ben, they still have the right to study something else. Who bears the costs depends entirely on the relative elasticities. In this case, there's no (or very limited) substitute for the employer and plenty of options for the employee, so employers have little choice but to bear the increased cost. Leaving out nuances like taxes, financial aid, and that it's NPV of salary that matters.

Think about taxes. Who benefits from the mortgage interest deduction? Do you think fat-cat CEOs are punished by tax on corporate profits? Or do you think maybe, just maybe, the guy writing the check might not be the economic point where the buck stops?

It would be interesting to study if this was an important part of college price inflation. It's a market kind of like healthcare with third-party payers being very important.

James A. Donald writes:

Selecting candidates by IQ might be almost legal - but every business knows that when a sexism racism etc case comes up, everything is deemed relevant. The entire business, and everyone in it, is reviewed for signs of political incorrectness.

So even if you could get away with an IQ test, when you get sued, and you will get sued, you will not get away with it.

Comments for this entry have been closed
Return to top