Tags
CEOs, IQ, IQ floors, IQ thresholds, linear relationships, Lion of the Blogosphere, U.S. presidents
The wildly popular Lion of the Blogoshere is the latest blogger I’ve seen commenting about a study showing that Swedish elite CEOs have a mean IQ of 115. He writes:
These findings are consistent with my previous speculation that most career tracks have cognitive floors, and once you are above the cognitive floor, having additional higher IQ isn’t of much use. The cognitive floor for management positions in corporations seems to be around 110 or so. After that, getting promoted requires personality traits that are correlated with “fitness for the military” rather than higher IQ.
This idea that IQ is quite important below a certain threshold, but quite unimportant above that same threshold is very popular and potentially true, but I’ve always preferred a simpler model that splits the difference: IQ is mildly important across virtually all levels.
I don’t know if my simple model is true, but the paper seems to show that minor CEOS have a mean IQ of 108 and major CEOs have a mean IQ of 115. Now the paper defines minor CEOs as CEOs running 8 figure companies ($15 million or less) and major CEOS as running a 10 figure company (more than $1.5 billion USD). From here, we can speculate, that for every tenfold increase in company size, the average IQ of the CEO increases by 4 points. So:
Million dollar companies: Average CEO IQ 104
Ten million dollar companies: Average CEO IQ 108
Hundred million dollar companies: Average CEO IQ 112
Billion dollar companies: Average CEO IQ 116
Ten billion dollar companies: Average CEO IQ 120
Hundred billion dollar companies: Average CEO IQ 124
Trillion dollar companies: Average CEO IQ 128
Ten trillion dollar companies: Average CEO IQ 132
Of course there’s no such thing as a ten trillion dollar company, however the sitting president of the United States is often described as America’s CEO. The United States has a GDP of about 17 trillion and preliminary data suggests that U.S. presidents have an average IQ around 130, about what you’d expect from the CEO of a company the economic size of America.
So I don’t necessarily agree that the importance of IQ suddenly drops beyond a certain threshold; instead I see IQ as mildly but equally important across virtually the full range, but I could be wrong because comparing presidents to CEOs is not exactly an apples to apples comparison, and an economist could probably come along and rip this analysis to shreds.
The Lion of the Blogosphere also writes:
The problem here is that while an IQ of 110 to 115 is high enough to get promoted through the corporate ranks, it’s not smart enough to make smart decisions about the direction of the business. This explains why most corporations are so poorly run and continually make costly business mistakes.
I agree with this 100% but I would go a lot further and say that an IQ of 130 is not smart enough to be president of the United States. That doesn’t mean there haven’t been terrible presidents with IQ’s way above 130 and excellent presidents way below 130, but the lower IQ great presidents were probably great in spite of their lower IQ’s, not because of them; or because of certain personality traits correlated with lower IQ, rather than the lower IQ itself being beneficial. In fact I think the job of an American president is so complex that none of them have been smart enough for the job, and that’s probably why the correlation between IQ and job performance among presidents is so low. When everyone is in over their heads, success and failure will be largely determined by luck.
I guess not not sure how lion is qualified to say businesses are poorly run. Kinda a vague statement on his part since. I wish he were right because I would make a killing buying put options on these poorly run, soon-to be bankrupt companies lol. Companies fail not so much because of management, but due to market conditions.
Well, look at all the companies that had to be bailed out by the government. That’s evidence that they’re poorly run. When you reach a certain level, you become “too big too fail” because too many people are relying on and invested in your success, so you can make a lot of dumb mistakes and still thrive.
hmmm… but that’s just a dozen or so companies that were bailed out out of tens of thousands. Many companies were forced to take bailout money when it wasn’t necessary. Jamie Dimon, CEO of JPMorgan Chase said he was forced to take bailout money.
test post by pumpkinperson
It would be interesting to see how well the actual competence of the CEO correlates with IQ.
Correct.
As Peter Lynch put it, “Go for a business that any idiot can run — because sooner or later, any idiot probably is going to run it.”
I am quite surprised with the results, as my guess would be higher for average IQ for a CEO. The income vs IQ correlation does not fit these results, probably, as CEOs are the highest paid people.. Anyway, the test seems robust.
And your method of dividing by correlation for height. Height difference was 0.5 SD. Divide by 0.2 height correlation, it gives 2.5 SD, which is IQ 137.5 , which seems more accurate.
The IKEA owner, Ingvar Kamprad (an alcoholic), picks Peter Agnefjäll as CEO instead of one of his three sons. Peter Agnefjäll graduated from Linköping University, which is only the 6th best university in Sweden. I am just surprised, and I try to understand why that is so.
Increasing IQ with net worth of the company is already a good sign, but the numbers are still very low…The paper seems to show 2 SD high for a billion dollar company. I am not sure…
It might be better to correct the height-IQ correlation for reliability before dividing,raising it to 0.23.
But CEOs, like U.S. presidents, are probably taller than other people of the same IQ, because height is a valued political and business leadership trait independently of IQ. For example, doctors tested in the same study had the same IQ’s as CEOs but were shorter.
It might be easier to become a CEO in Sweden because it’s a small country and CEO pay is probably a lot less, reducing competition. The IQ’s of American Fortune 500 CEOs are probably at least 10 points higher:
http://brainsize.wordpress.com/2014/07/18/the-iqs-of-fortune-500-ceos/
As I said on Leon’s blog, 115 sounds a bit low but let’s go with it. Supposing the average CEO has a 115 IQ. That doesn’t mean anyone with an IQ of 115 is capable of running a company. For that matter, not everyone with an IQ of 160 could either. There are other qualities required to run a company as well. It just so happens that the average IQ of those possessing the best combination is approximately 115. Of course, there will be some who possess an abundance of all necessary qualities and they’ll be the cream of the crop.
I personally think psychology and personality play a major role. It’s not necessary to be able to do everything. One need only be able to see the big picture and delegate and manage those who can do what you can’t. I think being too smart gets in the way of that because it’s maddening when you can do everything better than everyone else. Sometimes you just have to let less capable people do their job and accept that they’ll make stupid mistakes as part of the price of not doing everything yourself.
As for president, politics is so dynamic and uncertain that you can’t manage everything. So it’s better to have a first rate crew with an even tempered captain even if he isn’t the smartest person in the room. If you’ll notice, the real blunders are never caused by a lack of intelligence but emotional immaturity and psychological bias. Having said that, it’s a real shame that the qualities that help one reach positions of influence (either CEO or politician) aren’t necessarily the ones that help them do a good job once they get there.
destructure,
There was a study that tried to estimate the IQ’s of U.S. presidents. I haven’t read it all so I could be wrong, but if I understand table 4 correctly, it looks like about a 0.3 correlation between IQ and leadership performance.
http://www.acsu.buffalo.edu/~jcampbel/documents/SimontonPresIQ2006.pdf
0.3 is not that low, but keep in mind that squaring the correlation suggests that only 9% of leadership performance is explained by IQ, and 91% is explained by other variables (luck, and probably the very personality traits you suggest).
I suspect this applies not just to presidents, but CEOs or just managers of any kind.
‘I agree with this 100% but I would go a lot further and say that an IQ of 130 is not smart enough to be president of the United States.’
lets just assume that IQ is real and meaningful for this one. i dont know why any leader needs an IQ above the bare minimum. it doesn’t take a colossal intellect to pick genius lessers and then referee the arguments made by those genius lessers.
When two geniuses argue, it’s very hard to know who is right unless you’re a genius yourself; and the problems a leader faces are so complex, that likely BOTH genius advisers are wrong.
i didnt say you had to know who’s objectively right. you just have to know who made the better argument. most of genius lies in assembling the non-obvious premises to then make an argument, not the argument itself. if you’re smart enough to grok the rules of argument, you’re smart enough to referee.
I’m not arguing that IQ plays an important role in being president, I’m just arguing that whatever role it plays, it plays at all levels. Most people think IQ is really important below 120, and then virtually irrelevant above that level. I on the other hand, think IQ is only a bit important at ALL levels. So an IQ 170 will will be just a tiny bit better at a job than an IQ 160, but an IQ 70 will also be just a tiny bit better at a job than an IQ 60. But tiny differences in job performance at every 10 point interval add up, so an IQ 170 is a vastly better worker than an IQ 60. The only time where I’d expect a threshold beyond which IQ became irrelevant to job performance is if the job were so simple, that virtually everyone above a certain IQ could do it as efficiently as possible. But when it comes to the U.S. presidency, there is no such threshold, and even the smartest presidents have regrets about the huge miscalculations and mistakes they made on the job.
i wanna reiterate that im just assuming in arguendo that IQ is real and important.
when they say virtually irrelevant they are referring to a diminishing returns horizontal parabola curve where x is ‘higher IQ’ and y is ‘better performance.’ Those “gained performance bits’ are smaller with every single IQ gain. So IQ 70 wlll be a LOT better than IQ 60. IQ 170 will be a smidgen better than IQ 160.
intelligence allows you to make sense of available data. it does not make data more available. it does not create data where there is none. it does not make the random determined. there is a data limitation that bottlenecks raw intelligence’s (or as you say IQ’s) usefulness in the context of the Presidency — and many other jobs.
o shit *determinable. my b.
when they say virtually irrelevant they are referring to a diminishing returns horizontal parabola curve where x is ‘higher IQ’ and y is ‘better performance.’ Those “gained performance bits’ are smaller with every single IQ gain. So IQ 70 wlll be a LOT better than IQ 60. IQ 170 will be a smidgen better than IQ 160.
I’m impressed! Yes, that’s exactly what proponents of the threshold model would need to show to make their case, but such graphs are almost never provided; & the few times that they are, no threshold effect is dedectable unless the job is so ridiculously simple that there is a low maximum level of efficiency with which it can be performed.
It’s possible that IQ is far more important for dumb leaders than smart ones, but without the kind of graph you describe; it shouldn’t be assumed.
In one of my few psych classes I think I remember that clever people might do worse at certain jobs, because they were bored by them whereas their cognitive inferiors weren’t.
And swanknasty, of course, is pointing out that pearson’s correlation is only good for a bivariate normal distribution.
The regression line, the curve of conditional expectations, is only a straight line when the bivariate distribution is normal or elliptical.
dat spearman tho.
…
anyway, http://arxiv.org/pdf/1011.0663v1.pdf
SAT-M to G.P.A fig. 2 may be of interest.
tho IQ/g are assumptions of a sort anyway, so…
I would never assume that an increase in IQ does not make a performance difference at all levels, if only in speed and less errors in any profession. When you speak to someone who is 10 IQ points or more lower than your IQ, it is easy to discern the performance difference.
130 should theoretically be plenty for a President. I’d venture that 125 would even pass, especially given the support that modern presidents receive. I think that you might be underestimating what a 130 IQ is. It’s nothing to sneeze at, and only 1/50 Caucasians will have such an IQ. To put in in perspective, just a little bit north of that IQ would sit your average university physicist. Though, the physicist comparison isn’t precisely appropriate to presidential decision making as IQ is not always reflected in quantitative skill I’d be willing to bet that the majority of Presidents have not been much over 130-135, when their IQs have been that high.
My educated guesses on recent presidents, based on observation:
Obama – Probably 130. A high verbal intelligence but I get the impression that it is over-expressed perhaps in comparison to quantitative ability. A lot of these individuals will sit around 130 in my observation, and black men tend to statistically max out at around 130.
GWB – I think that a lot of his stupidity was an act, and I say that as a non-fan. 125-130. Though it could be a little higher depending on how much of an act the gaffes were.
Clinton – 140
Bush Sr – 140
Reagan – 130-135
I’m not old enough to have a significant impression before Reagan.
I’m going to be estimating Reagan’s IQ on President’s Day.
I think that you might be underestimating what a 130 IQ is. It’s nothing to sneeze at, and only 1/50 Caucasians will have such an IQ.
I realize it’s very high, but the job of the President is unbelievably complex and important, so no matter how many IQ points you have, you never get to a point where you don’t need more.
Look, the president is surrounded by an enormous advisory body round the clock that supports and often totally replaces him in matters of decision making.
You shouldn’t assume it’s as difficult as you are assuming it to be.
And I’m not denying that it’s a fairly difficult job but I doubt it requires stronger quant skills/abilities than a job as a city trader.
And I’m not denying that it’s a fairly difficult job but I doubt it requires stronger quant skills/abilities than a job as a city trader.
The key word is “requires”. I don’t think being president “requires” a high IQ, in fact with a lot of luck and the right personality, even someone with an IQ of 90 might be a good president.
When I do think is that the probability of being a great president increases very slightly with each IQ point, in a linear way, all the way up to well beyond 150.
And that’s not to say most great presidents have been anywhere near 150
By contrast there might be other jobs that people with IQs below 115 have almost zero chance of being competent in, let alone great, but once you’re above 140, extra IQ does not increase performance.
“increases very slightly with each IQ point, in a linear way, all the way up to well beyond 150”
Why is there any reason to assume that at all when the policies of a given president will be decided by the various committees and advisory bodies that constitute a government of the day?
Even people criticising Reagan for that era’s policies (Be those criticisms misguided or not in the economic and foreign sense) are definitely misguided in criticising the man for it was the government of the DAY that selected those policies.
Since it’s that government that makes the much larger and in fact almost total weight of the important decisions why assume through all that interference we can somehow deduce president from policy and thereby his IQ?
The only really significant method we have to deduce the intelligence of the presidents is their writings (Many of which are intentionally doctored and written for them by others in themselves) their own positions (Which may simply be the recieved wisdom of their parties also, saying nothing perhaps of the presidents themselves) and last and most importantly their academic records.
I think that being a “Good” president is defined by being a rallying point for the people of a nation, and the large part of that is the power they have to inspire through speech, and though that may be related to intelligence by some measure I suspect it has more to do with being in touch with the Zeitgeist at the right time when the right thought strikes you, and your character and the impression it leaves on your writings, than it does IQ.
Why is there any reason to assume that at all when the policies of a given president will be decided by the various committees and advisory bodies that constitute a government of the day?
Because it’s the president who picks his advisers, and it’s the president who ultimately decides when his advisers disagree.