What makes really smart people tick? Why do some end up earning so much more than others?
And how much do these disparate outcomes have to do with their personalities? A new study by Miriam Gensowski, at the University of Copenhagen, sheds fascinating light on these and other questions.
Gensowski revisits a data set from schools in California in 1921-1922, based on students who scored in the top 0.5 per cent of IQ distribution.
The data then covers how well these students did up until 1991. The students were rated on five personality traits and behaviours: openness to experience, conscientiousness, extraversion, agreeableness and neuroticism.
One striking result is how much conscientiousness matters.
Men who measure one standard deviation higher on conscientiousness earn, on average, an extra US$567,000 (NZ$865,340) over their lifetimes, or 16.7 per cent of average lifetime earnings.
Measuring as extroverted, by one standard deviation higher than average, is worth almost as much, US$490,100. These results were most pronounced for the most highly educated of the men.
For women, the magnitude of these effects is smaller. For one thing, women earned less because of restricted opportunities.
But also, extroversion is more strongly correlated with higher earnings than is conscientiousness, unlike for the men.
It may surprise you to learn that more "agreeable" men earn significantly less. Being one standard deviation higher on agreeableness reduces lifetime earnings by about 8 per cent, or US$267,600.
In this context, you can think of agreeableness as meaning a person is less antagonistic and more likely to consider the interests of others.
You might have thought agreeableness would be correlated with higher earnings, but alas not. That said, this result is confirmed only for high IQ individuals in California for this span of the 20th century. It may or may not be true more generally.
And as always, correlation does not prove causality.
One possibility is that more agreeable men self-select into lower-earning, more-subordinate professions, in which case acting like a jerk at work won't automatically bring you that higher paycheque.
Higher IQ is also correlated with higher earnings, by about 5 per cent or US$184,100 for a one standard deviation boost.
That's a bit surprising, because the sample is already within a band of very high-IQ individuals. But apparently being "even smarter yet," at least as measured by IQ, is correlated with additional pay.
Most generally, these personality traits start to correlate more strongly with income when workers are in their early 30s, and their influence peaks between the ages of 40 and 60, dwindling thereafter.
The data also lead to some interesting ways of rethinking education.
There is no evidence, for instance, that individuals who received more higher education ended up with greater conscientiousness.
That said, once we adjust for personality traits, higher education does seem to yield pretty high financial returns, more than 12 per cent a year for getting a university degree, again within this sample only.
A doctorate is worth about 1.5 times that, or about US$1.7 million.
These correlations are suggestive only, but they are consistent with a model where going to university doesn't teach the very smart how to buckle down, but it does impart some concrete skills and maybe give them the connections to earn a lot more money.
The correlations are evidence against the view that the well-educated earn more, primarily because they are intrinsically better workers.
There is no special reward for being head of the class!
Another interesting result is that IQ and conscientiousness are not very well correlated. That implies that finding ideal workers isn't so easy.
The quality of openness, however, is moderately positively correlated with IQ, so you might expect that smarter workers are more willing to experiment and try new things.
We still don't know how generalisable these results might be, partly because some regions don't offer the opportunities of California, and partly because the results measure how things worked in an earlier time.
Perhaps they've changed for women more than for men?
Most of all, I am struck by just how little we know about what determines earnings, or how much human personality matters for life outcomes, or how well we can measure personality in the first place.
You should think of this investigation as the first word, not the last. In the meantime, maybe it's not such a bad idea to buckle down and work hard, don't be shy, and don't freak out if you are just a wee bit grumpy at times. Your pay packet may depend on it.