We’ve been testing “intelligence” for about a century. What does an IQ test “test”? It tests what is called “abstract intelligence,” basically solving logic problems. There is a correlation—not a cause-and-effect linkage—between high IQ scores and both good grades and good job performance. In contrast, many of the standardized tests (SAT, GRE, MCAT, LSAT, GMAT) that hold the keys to opportunity in life measure verbal and quantitative ability. Apparently the two types of tests measure different things because IQ scores have been rising steadily for decades, while measures like the SAT have stalled or even fallen. (Then there’s the third kind of test called “personal judgment” used by employers, teachers, and voters. I don’t see scholars doing much work on what I conjecture to be a key factor in individual success.)

Raw IQ scores rose steadily in all developed countries throughout the 20^th Century and continue to do so today. Three points per decade is the normal increase. That means that my older son should have an IQ about 20 points higher than my father and about a 12 points higher than myself.[1] They are rising across all groups tested, rather than in just part of the population. The dumb are getting smart and the smart are getting smarterer.

It’s hard to tell why they are rising. Improved nutrition and health may play a role by allowing most children to develop more rapidly. IQ tests are usually given to captive school or draft age populations, so a fast first step in life might contribute to rising scores. Also, more education became available to the lower income groups in the course of the 20^th Century. This could move the “left tail” of a standard distribution to the right. On the other end of the spectrum, smart kids increasingly hang with smart kids. A “social multiplier effect” may explain why the IQ scores of this group continue to rise when improved nutrition and education cannot be factors. There is an intriguing third possibility. Both computer games and popular television shows have become increasingly complex. The narratives in each art form stress complicated plots and lots of characters. Perhaps this fosters a constant analysis of abstract relationships.

Then there are Torrance Tests of Creative Thinking. These aren’t discussed much, if at all, in the media. What one analysis of the scores since 1968 shows is that the Torrance scores for American children have been dropping since 1990.

While something is causing IQ to rise, it probably doesn’t have anything to do with what is going on in the classroom. Better nutrition, more years in school, watching “Lost,” and playing “Call of Duty” might be the most reasonable explanations. In contrast, the stagnation (at best) in the SATs and the decline in the Torrance Test scores might be related to what is happening in classrooms since they test things that schools claim to teach and foster.

There are implications for social policy. First, improving both childhood nutrition and education help poor children. That could get tangled up with questions about the quality of parenting supplied by poor people. FDA guidelines on nutrition and good advice on stimulating the intellectual development of children has been available for decades. Not everyone uses them. Second, schools and colleges might want to think about incorporating complex television series and electronic games into their tool-kit. If William Shakespeare were alive today, would he prefer to write “Kinky Boots” or “Breaking Bad”? Third, mixing higher IQ students with lower IQ students in the classroom may be good for the lower IQ students. It won’t be good for higher IQ students. I can hear the Republican outcry against Democrats’ “redistribution of grey cells” already. I might be part of it. (“Smarter than ever,” The Week, 16 September 2011, p. 11.)

[1] This may be an argument for having children as early in life as possible. It cuts down on the IQ gap between yourself and your children. Otherwise the little bastards will be just insufferable.