Reading some of Inductivist’s work on fertility and political orientation among wealthy white women, I was struck by the following:
Education’s effect is not the influence of IQ. Among this demographic, IQ is unrelated to fertility. The correlation between IQ and number of kids is .02–basically zero. Education’s influence appears to be cultural.
I previously did my best to portray the strong, inverse relationship between educational attainment and fecundity in the US. The correlation is indisputable, but that, of course, does not necessarily inform us about causation. The fear of dysgenic dystopia in the future is usually premised on the notion that IQ, not schooling, and fertility run in opposite directions.
Inductivist’s statement made me wonder how much IQ matters once education is taken into account. It’s easy to conceptualize why contemporary education and procreation don’t mesh well. Young men and women spend their most fertile years accumulating debt and squandering assets (if they have any to begin with). By the time they’ve graduated, established themselves in their careers, and fixed their personal balance sheets (if they ever manage to do these things, and increasing numbers of them do not), their biological clocks, especially women’s, are past noon.
It’s more difficult to do the same with intelligence. Popular culture does not particularly celebrate childbearing, but procreation isn’t public enemy number one, either. The world-shrinking consequences of technological progress means there are more things than ever for intelligent people to pursue, of which having and raising kids is just one of an ineluctably growing many, and not a seemingly attractive or intellectually ‘actualizing’ one at that. But total fertility rates in the West were well above replacement for several decades after children had clearly become economic, time-consuming ‘burdens’ rather than an extra pair of hands to help on the farm. Intelligence and education are linked, often confusedly, in the public mind, but a plausible explanation is that more intelligent people tend to pursue more education, and end up having fewer children as a result.
Using the GSS, let’s consider just that. In the following graphs, respondents are categorized into five groups* based on educational attainment and into five groups based on IQ as measured by wordsum scores. For contemporary relevance, only data from the last two decades are considered, and all respondents included are at least 35 years age to allow time for family formation to have occurred.
In the first graph, we see how educational attainment and fecundity relate among those of similar IQ:
So, among those who scored in the 0-3 range on the wordsum test (the real dumbs), those with less than a high school education (black bar), average just over three kids. Those real dumbs who graduated high school (brown bar) and those who spent some time in college without graduating (white bar), average about 2.5 each, the real dumbs with a bachelor’s (yellow bar) about 1.75 kids, and those with post-graduate degrees (baby blue bar) 1.5.
For all IQ groupings, education is a strong predictor of fecundity. Even among the really smarts, obtaining a college degree reduces total fertility by 0.5. Among those of modest intelligence, staying in school means staying out of the delivery room. I’m sure there’s a cute phrase like “Yale or jail” applicable here, but it’s not coming to me at the moment (soliciting suggestions in the comments).
In the second graph, we see how IQ and fecundity relate among those of similar educational attainment:
The relationship here is considerably weaker. The effect of education on average total fertility rates among those of similar IQ is five times as large^ as the effect of IQ on average TFRs is among those of similar educational attainment. This is detectable in the latter graph by noticing how there is a clear decline in fecundity as we move up from one level of educational attainment to the next, but there is not much difference across the IQ spectrum among those of similar educational attainment.
While aware that it is an erroneous oversimplification to argue that the demographic transition in the West has been driven primarily by the increasing duration of formal educational attainment, the latter is strongly associated with the former. The more I look into it, the more difficult it becomes to think about the developed world’s inverting age pyramid without considering educational romanticism.
We need methods to speed up the educational process, like self-paced coursework and subject-specific standardized testing (think AP tests for those in college) that allows autodidacts to receive credit as soon as they’ve demonstrated proficiency in a subject rather than after four inefficient months of spending three hours per week having it delivered to them at varying levels of effectiveness. Ideally, passing the bar would be the only requirement for practicing law and passing the CPA exam the only requirement for becoming a CPA. If this results in a perceived glut of lawyers and accountants, the respective tests can simply be made more difficult. While the relative value of high parental socioeconomic status will decrease and higher conscientiousness might as well, higher intelligence would be rewarded with more precision and young professionals would be able to get to work years earlier and with tens of thousands or hundreds of thousands of dollars less debt on their shoulders.
Necessity being the mother of innovation, here’s to hoping the impending popping of the education bubble ushers in changes along these lines.
GSS variables used: AGE(35-89), BORN(1), EDUC(0-11)(12)(13-15)(16-17)(18-20), WORDSUM(0-3)(4-5)(6)(7-8)(9-10), YEAR(1990-2010), CHILDS
* Groups are mutually exclusive, so those who are listed as having graduated high school went exactly that far but no further. Obviously, those who have doctorate degrees also graduated from high school, but they are only represented in the “post-graduate” grouping here.
For IQ, respondents are broken up into five categories that come to roughly resemble a normal distribution; Really Smarts (wordsum score of 9-10, comprising 13% of the population), Pretty Smarts (7-8, 26%), Normals (6, 22%), Pretty Dumbs (4-5, 27%), and Real Dumbs (0-3, 12%).
^ Ordinary least squares standardized regression coefficients are (.16) for educational attainment and (.03) for intelligence as measured by wordsum score. My thanks to Inductivist for walking me through the process of how to determine these.