A quick note on the genes-SES gradient

I’ve previously pointed out that rich and poor people differ on more polygenic scores than just for educational attainment:

Here’s some more data to back that up. Again I’m using the Health and Retirement Study (and, for simplicity, only white people). The picture below shows the relationship between 58 polygenic scores (PGS) and income.

I’m looking at things an unusual way round: asking how much someone’s income predicts their score, rather than vice versa. So the x axis shows how many standard deviations the PGS is predicted to change, for a one standard deviation change in income. I’ve corrected for multiple testing using the false discovery rate, which is a bit more lenient than the usual way. I think that’s appropriate here.

This is a picture of the genes-SES (socio-economic status) gradient. Obviously, we can’t be certain exactly which associations are real and which are just noise. But you can see a lot of cases where income has a large association with the score. Some of them are social phenotypes like education or age at first birth; others relate to mental health or cognition; others pick up physical health or health-related behaviours like BMI or smoking.

There are different possible causes for these relationships.

• Pure confounding. Richer people smoke less, you did a genome-wide association study on smoking, you captured some genetic variants that correlate with being rich and smoking less, but have no actual effect on either.

• Social mobility. The genetic variants cause the phenotype; the phenotype makes you richer or poorer. Obvious candidates would be educational attainment and cognition. We have good stories for why these may lead to higher income in a modern society.

• True causality, but going the other way. The genetic variants actually cause you to have higher income, and then that makes you less likely to smoke/be depressed/etc. There are many stories for why this might be true also! Money gives you a lot of advantages.

• Marriage markets. This relates to our working paper. Smart, slim, or healthy people marry rich people. Their kids inherit their genes and (some of) their income, via wealth, via getting a good education, or whatever. So now the genes and the income are associated.

I find all of these quite plausible, and of course any or all of them can be present together. Anyway, the picture shows that many PGS are associated with income — not just the obvious ones like education. So, it’s important to understand why. In particular, in all but the case of pure confounding, the genetics might be making an important contribution to inequality, whether that’s income inequality or specific health inequalities.

Here’s a picture of the raw data for the significant PGS, just to give you a sense of what that looks like. Note that the y axis cuts off a lot of the variation in PGS around the smoothing line.

---

If you enjoyed this, you might like my book Wyclif’s Dust: Western Cultures from the Printing Press to the Present. It’s available from Amazon, and you can read more about it here.

You can also subscribe to this newsletter (it’s free):