The Bell Curve, by Richard J. Herrnstein and Charles Murray

I seem to recall a while back that a president of Harvard University got into some trouble by quoting from this book, but I could be mistaken. The ideas presented by its authors would certainly be regarded as controversial, if not downright inflammatory in some circles. Hernnstein and Murray performed a statistical analysis of a long term study following thousands of young people through their lives, called the National Longitudinal Study of Youth, NLSY, trying to determine the effects of IQ, or cognitive ability, on life outcomes.

Intelligence among the general population when placed on a graph, forms a statistical distribution called a bell curve, with the bulk of people in the middle, and a very small percentage at both the lowest and highest ends, hence the title of the book. When the authors talk of "low" or "high" intelligence, they're talking about the ends of the bell curve, not the great mass of "average" people in the middle.

The authors claim that in nearly all measures of success or failure in society, cognitive ability has an influence. It's obvious that those who test higher in intelligence have a better chance at getting a good education, and also rising to higher levels of education. But it's perhaps little known that students at the top end of the bell curve have been increasingly skimmed off the top to attend ten of the most elite universities in the country, and the people who graduate from those universities go on to run our industries and government, mostly.
Job performance in nearly all fields is affected by intelligence, as well. "It may be said conservatively that for most jobs, based on most measures of productivity, the difference in productivity associated with differences in intelligence diminishes only slowly...The cost of hiring less intelligent workings may last as long as they stay on the job."

While the authors make the point that "you cannot determine what a given person will do from his IQ score", they also say that high cognitive ability is "generally associated with socially desirable behaviors, low cognitive ability with socially undesirable ones." and "...the brighter young people in the NLSY are also the ones whose lives most resemble a sometimes disdained stereotype: They stick with school, are plugging away in the workforce, and are loyal to their spouse."

By the way, for the first half of the book, in order to take away the subject of race from the discussion, the authors used only the data from the white youths studied. So all of the results of low intelligence they discuss are with respect to caucasians.

Poverty is also strongly correlated with intelligence. They say it's actually better to be born smart than to be born rich. Interestingly, in 1939, half of the US population lived below the poverty line, but increases in productivity after WWII began to lower that rate, up until the mid 60s, when it leveled off around 10% and has remained ever since. Poverty actually began to be systematically studied in the mid 1800s, and it was believe that the poor fell into two categories, "deserving" and "undeserving". Some were poor because of circumstances beyond their control, and others were poor as a result of their own behavior. Whatever happened to that line of thought?

Hernnstein and Murray also link criminal behavior with intelligence. In contrast to theorists who offer a sociological explanation of crime's causes, such as poverty and unemployment, they say that more serious or chronic offenders generally have lower IQ scores than casual offenders, and that high IQ scores provide some protection against lapsing into criminality for people who have lived in environments regarded as risk factors for criminal behavior. One good quote, "Many people tend to think of criminals as coming from the wrong side of the tracsk. They are correct, insofar as that is where people of low cognitive ability disproportionally live."

In the second half of the book, the authors begin to look at the issue of race or ethnicity and intelligence. I believe this is where that Harvard president got into trouble. Although it's extremely unpopular to say it these days, there are - in the aggregate - differences between ethnic groups. The authors talk a bit about the whole issue of  cultural bias that is assumed to make it nearly impossible to test a person's IQ accurately if they've come from a different cultural background than the population the IQ test is allegedly written for.

It turns out that there are some tests available that have a strong correlation with cognitive ability scores, which appear to have no way to be culturally biased. One of these, called digit span, involves repeating a string of numbers spoken to you, either forwards or backwards. Whites, blacks, and asians perform differently on these tests. Another one involves placing your hand on a button, depressing it. Lights of different colors are arranged around the button in a wheel. When certain lights turn on, certain actions are required, such as touching the light to turn it off, or perhaps a more complex action with multiple lights. Again, different ethnic groups score differently. It's difficult to imagine that either numeric values or flashing lights can be culturally biased.

One standard test that's been used for many years to test cognitive abilities is the ASVAB. It's taken by all people who enter any of our armed forces. Strangely, scores on the ASVAB actually do a very good job of predicting success within various MOS's. Even more strangely, given assumptions of cultural bias implicit in intelligence testing, the ASVAB does a good job on predicting outcomes for whites, blacks and asians alike.

The authors go on to make some predictions about what the future may hold, and offer some ideas for helping to solve today's pressing social problems, based on their understanding of the underlying causes. I think they're on a lot more shaky ground with predictions than on their data-based conclusions about IQ and outcomes, but this is really an interesting and thought-provoking read.