Drawing Connections Where There Are No Connections

Thomas Edsall’s latest New York Times offering, “We Are Leaving ‘Lost Einsteins’ Behind”, was pretty much what I expected it to be but I thought he raised some interesting points nonetheless. Let’s start with the title: what does he mean by “Einsteins”? Colloquially, it generally means a person with a high IQ. Let’s consider that for a moment.

As far as we can tell Einstein never took an IQ test but it’s estimated to have been four standard deviations above normal or more. .37%, a little more than a third of one percent, of the U. S. population might have IQs that high assuming normal distribution. That’s not a particularly good assumption as it turns out but it will do as a first order approximation. That means that there are a few more than 1 million people with IQs that high in the U. S. So, yes, there could be hundreds of thousands of undiscovered high IQ individuals in the U. S.

Does it matter? I think the answer is not very much. It helps a lot to have higher than median intelligence. It opens a lot of gates—it’s very difficult to become a medical doctor, for example, unless your IQ is more than one standard deviation above normal. But not all geniuses are successful and there is a positive correlation between a number of mental disorders including depression and very high intelligence.

Additionally, IQ tests only test a narrow band of skills and there are other skills, e.g. social-emotional development, which have been found to be more important than a high score on an IQ test.

As it turns out that’s what Mr. Edsall is pointing out but he focuses on another very narrow measure:

“Current talent search procedures focus on the assessment of mathematical and verbal ability,” wrote David Lubinski of Vanderbilt and Harrison J. Kell, a senior researcher at the Educational Testing Service, in “Spatial Ability: A Neglected Talent in Educational and Occupational Settings.” Lubinski and Kell stress the failure of many of such searches to test for the cognitive skill known as spatial ability.

What’s “spatial ability”? It’s the capacity “for mentally generating, rotating, and transforming visual images”. I know of no evidence that Mark Zuckerberg or Bill Gates excelled in that particular area but there is considerable evidence that they do have high IQs as measured by IQ tests. Today seven of the ten largest STEM areas are computer-related. I have rarely observed notable spatial ability among people in information technology other than in the rather narrow area of computer graphics. Quite the opposite in fact.

I think Mr. Edsall is cherrypicking. However, let’s assume that he’s correct and that we’re allowing a substantial source of ability to go to waste (also cf. here). The areas of engineering in which spatial ability would be most useful, e.g. mechanical engineering and civil engineering, have seen sharp declines in the native-born population over the last couple of decades. Why? The explanations are pretty simple: wages in those areas peaked a long time ago, demand for mechanical and civil engineers has not been increasing, and we can import as many as we might want from India which will ensure that wages don’t increase. Additionally, the steep decline of U. S. manufacturing employment which followed China’s gaining most favored nation trading status greatly reduced the need for people with spatial ability.

All of which point to the solution to the problem: resume making more of what we consume here and stop importing engineers from abroad. I suspect that was not Mr. Edsall’s point but it’s the direction in which the incentives point.

Before someone mentions it first, median IQ in China is higher than median IQ here. There are more people in China with IQs one standard deviation above normal than there are people here.