Grey Swan

I used data from the National Longitudinal Survey of Youth to examine how IQ affects income. The survey takes a set of young people in 1979 (late teens to early twenties) and interviews them on a broad range of issues every few years. While surveys have continued beyond 1996, I only have data up to that year.

In the simple linear regression of Income vs. IQ a clear trend emerges – income diverges by IQ as we age*. This may seem counterintuitive – doesn’t raw intellect matter less and less as we get older? Apparently the answer is no. Perhaps this is because the smartest people continue learning at a faster rate as they age, or perhaps this is because it takes a while for there salary to catch up to the value of work produced. E.G., A McDonalds employee is unlikely to be paid below the value of their production. However, many financial companies use strongly hierarchal structures that effectually underpay recent college graduates. Adoption studies confirm a similar trend: Adopted children’s financial outcome decreasingly correlates with their adoptive parents and increasingly correlates with tested aptitude.

What about education? The problem with controlling for education in the data is that smart people tend to get educated more. Anyone with an IQ over 130 who fails to graduate college likely has other uncontrolled variables that decrease their income outlook. Consequently, we would expect a systematic downward bias in our results. Running the regressions shows just this – IQ decreases in importance but continues to be important.

The unfortunate aspect of the above analysis is that education matters less than many people think. Spending more and more on education meets decreasing returns because the raw material – the people who are being educated – stays the same.

*Note that IQ only explains about 15 percent of all variation in income – other factors continue to be very important.