February 26, 2016
From the ambitious original research website Human Varieties:
Posted by Dalliard
In his recent book Hive Mind, economist Garett Jones argues that the direct effect of IQ on personal income is modest, and that most of the benefits of higher IQ flow from various spillover effects that make societies more productive, boosting everyone’s income. This, he says, explains the “IQ paradox” whereby IQ differences appear to explain a lot more of the economic differences between nations than within them.
Jones does not say in his book what he thinks the exact effect of IQ on personal income is, but on Twitter he has asserted that “Fans of g would do well to look at the labor lit: 1 IQ point predicts just 0.5% to 1.2% higher wages.” He has also said that, in terms of standardized effect sizes, IQ accounts for only about 10% of variance in personal income (a correlation of ~0.32).
While I don’t doubt Jones’s overall thesis that the effect of IQ on productivity is broader than its effect on personal productivity or income, I think he understates the importance of IQ in explaining income differences between individuals. I analyzed a large American population sample and found a substantially larger effect of IQ on permanent income than previous investigations. It appears that the literature Jones refers to has failed to pay sufficient attention to various measurement issues. …
I used data from the NLSY79 which is an ongoing longitudinal study that follows the lives of a large sample of Americans born in 1957-64. Specifically, I used the nationally representative subsample comprising more than 6000 individuals.
This is the massive governmental study first made famous by The Bell Curve. It’s still going on. The Pentagon paid to have the sample take the military’s AFQT enlistment exam in 1980 to solve the disastrous misnorming problem of the Stripes era. The AFQT is an SAT-like test that correlates closely with IQ tests.
The NLSY79 contains income data collected from individuals annually until 1994 and every two years thereafter. Using these data, I calculated a measure of permanent income for each individual, following Mazumder’s recommendations to the extent that the data allowed it.
My permanent income measure is the average income calculated from up to nine biennial income reports from age 32 or 33 to age 47 or 48.
The slope coefficient for white men (the reference group) is 2.5% (95% CI: 2.2%-2.7%).
The Hispanic parameter estimates are not significantly different from the white ones, so the same equation applies to white and Hispanic men, although given the low sample size for Hispanics, this cannot be asserted very firmly. Black men, in contrast, have a significantly lower intercept and a significantly higher slope coefficient: each additional IQ point predicts 3.6% (95% CI: 2.6%-4.5%) more income for black men.
The R squared for the entire model is 23%.
That’s about an r = 0.48 for a model of AFQT, race, and sex.
When analyzed separately, IQ explains 20% of income differences in whites, 18% in blacks, and 16% in Hispanics. The correlations are 0.45 (white), 0.42 (black), and 0.39 (Hispanic).
In the social sciences, a correlation of 0.2 is often described as “low,” 0.4 as “moderate,” and 0.6 as “high.” So these correlations are moderate by social science standards, especially when modeling something like income, which people bring a lot of different factors to bear upon, such as smarts, education, work ethic, unique talents, skills, nepotism, personal relationships, sales talent, looks, regional differences in pay, health, etc. etc.
The slopes and correlations are lower for women, especially white women. (These are measures of individual income, not of family income — more white women are housewives without individual incomes.)