How do Parents Matter?

causality is difficult to establish

A while back, a commenter wrote,

Many social scientists concur that parenting has negligible or no impact on children’s outcomes. However, they also acknowledge that children from two-parent households tend to perform better. This is an intriguing contradiction.

When I see data showing a correlation between parents and their children, I instinctively think in terms of genetics. Suppose that high conscientiousness of parents predicts both that they will stay married and that their children will have high conscientiousness. If high conscientiousness predicts “perform better” in school and in career, then there you can expect correlation between two-parent households and better performance of children. But the correlation does not prove causation.

A lot of parent-child correlations can work like that. High-income parents with high-income children? Well, if cognitive ability and conscientiousness strongly affect income, and if they are highly heritable, there you go.

Many of my readers probably know that Judith Rich Harris wrote a whole book about this, The Nurture Assumption. She points out that for immigrants, children are more likely to adopt the language spoken by their peers than that spoken by their parents. Now extrapolate that for other behaviors as well.

Bryan Caplan’s Selfish Reasons to Have More Kids can be boiled down to a riff on Harris. “Parenting style doesn’t matter. So relax and enjoy more kids.”

It is hard to run the experiment of giving the same child two different parenting styles. That is why studies of identical twins raised apart get so much attention.

Even there, you have to be careful about reverse causality, running from children to parenting styles. Robert VerBruggen, concerning research that looked into this form of reverse causality, wrote

“Want to be a good parent?” the paper seems to ask. “It helps if you start with conscientious and agreeable kids.”

He may have been referring to this.

What is the strongest evidence that parenting style does matter? How does this evidence get past the “also might be genetic” alternative? How does it establish causality?

This essay is part of a series on human interdependence.