Not just dysgenic outcomes, but dysgenic intentions too

From the ever-relevant OKCupid dataset

I have a new study out with Ed Dutton:

• Kirkegaard, E. O. W. & Dutton, E. Intelligence and Anti-Natalist Intentions on Dating Sites: An Analysis of the OKCupid Dataset. In book: Intelligence, Race, and Sex: tribute to Helmuth Nyborg at 85. 2023.

A body of research indicates that people who are more intelligent tend to have fewer children than do those who are less intelligent, at least since around 1900 (Lynn, 2011). Nyborg (2012) has predicted that the consequent IQ decline will lead to the eventual decay of Western civilization. However, there is little research on fertility intentions and intelligence. Do smarter people end up with fewer children because they ideally desire fewer, or is it due to competing interests, such as a desire for money and status combined with more efficient use of contraception, as Nyborg (2012) observes? We analysed the OKCupid dataset of predominantly Western, English-speaking online users. Employing an ad hoc intelligence test composed of 14 160 Intelligence, Race and Sex questions on the dating service, we find that intelligence does indeed negatively relate to fertility intentions (β = -0.15, ordinal regression), even adjusting for age, sex, and race/ethnicity (β = -0.14). We also replicate the usual pattern of a negative association between intelligence and actual fertility, though the dataset was suboptimal for this analysis as fertility was only a binary outcome.

We used the ever relevant OKCupid dating site dataset. This is a dataset we obtained by scraping dating profiles on a dating site that used to have thousands of questions that people could answer in order to obtain better matches (applied assortative mating). Because of its large size -- up to 67k cases for an analysis -- it has sufficient power to look for small trends. One such case is looking into the origins of our current dysgenic fertility pattern. Dysgenic that fertility correlates in the wrong direction with a socially desired trait or phenotype, whether this is negatively with intelligence, or positively with antisocial behavior or mental illness. In this particular case, we looked at intelligence. The OKCupid dataset has a weird intelligence test. It's not a test someone consciously designed, but rather the result of users submitting questions on the site and 14 of these happen to be decent measures of intelligence. The questions are:

1. Which is bigger, the earth or the sun?

2. STALE is to STEAL as 89475 is to what?

3. What is next in this series? 1, 4, 10, 19, 31, __

4. If you turn a left-handed glove inside out, it fits on your left or right hand?

5. In the line ‘’Wherefore art thou Romeo?’’ what does ‘’wherefore’’ mean?

6. How many fortnights are in a year?

7. Half of all policemen are thieves and half of all policemen are murderers Does it follow logically that all policemen are criminals?

8. Which is longer, a mile or a kilometer?

9. When birds stand on power lines and don’t get hurt, it’s most likely because of what?

10. Etymology is?

11. If some men are doctors and some doctors are tall, does it follow that some men are tall?

12. A little grade 10 science: what is the Ideal Gas Law?

13. If you flipped three pennies, what would be the odds that they all came out the same?

14. Which is the day before the day after yesterday?

You can download the data yourself (supplementary materials) and analyze the 2500+ questions too.

Scoring the above 14-item IQ test is of course not as good as a longer, better designed test, but it's good enough for many practical purposes. We have seen this by correlating intelligence with things it should correlate with, whether belief in paranormal or pseudoscience (e.g. astrology), how many books people read, and even their own antisocial behavior (self-reported). In this new study, we looked at one particular question of interest:

• “How many children would you ideally like to have?” with options being: “None”, “1-2”, “3-4”, “5 or more!”

• The profile also showed an option for whether one already had children or not (yes/no), meaning that it measures 0 vs. 1+ fertility. Not optimal, but better than nothing.

We then fit a model to predict this ideal fertility from the usual variables:

With or without controls, intelligence has a beta of about -0.14. How large is this? Recall since this isn't a linear model, the value is hard to interpret as it's a logit. We can plot the model predictions to see what it means:

Here we see that the effect of intelligence -- controlling for race, sex, and age -- is an increase in the share of people want 0 children, and fewer who want 3-4 children. There's little effect on whether people want 1-2 children, and a slight decrease in the 5+ children. If we think of the IQ values in the typical range, from -2 to +2 standard deviations, the effect size is from about 18% who want no children to about 26%. The effect isn't large but it's there (p < .001).

The model also shows that various non-White groups (Asian, Hispanic, Black, maybe MENA) desire more children than the White comparison group. Visually:

Whites and Indians seem to have the largest share of people who want to be childness, and the lowest share of people who want 5+ children. The sample is largely left-wing, so probably not too much can be made of this finding without looking at political ideology. To be honest, we probably should have done that, but we forgot.

We can also look at whether we can see the usual effect on actual fertility, given the limitations of the data. Fertility is binary, 0 or 1+, and the age of the sample is fairly low because it's a dating site. Still, we get this result:

Again, the beta of -0.35 doesn't mean a correlation-like metric, it's logits. We see that controlling for the other variables, intelligence is associated with less chance of having any children. Some of the races categories also predict realized fertility, sometimes in opposite direction than for the desired fertility. Asian is negative (-1.17, p < .001), while their desired fertility is higher than Whites (0.39)! The sex interaction wasn't seen -- that female dysgenics is stronger -- but probably just due to low power (standard error is 0.046).

Conclusions

• Intelligence dysgenics can be seen "in the wild" on a dating site. It shows up for both desired fertility and for (so far) realized fertility.

• This result holds adjusted for age, sex, and race.

• Sometimes desired fertility and realized fertility don't appear to match up, suggesting people fail to realize their desires (so far).

• Our results replicate the study of Meisenberg & Kaul (2010), but we find a dysgenics on desired children, whereas he apparently doesn't (NLSY79 data).