I've written a bit about the premium accrued to men in the US after they get married. For women, it's the opposite story. After controlling for education, experience, IQ, ethnicity, and number of children, the average woman is hit with a 10% penalty. As a couple, this still leaves a substantial family premium when subtracted from the man's 44% salary increase. However, what explains the loss in earnings?
It can't be ability bias because married women are, on average, more conscientious, ambitious, and cooperative than their single counterparts. Therefore, if we account for this selection bias, the penalty is probably more than 10% - perhaps double.
Human capital likely plays a role as the married woman probably deprioritises her career and strives for a better work-life balance.
Then we're left with signalling, and the hard truth is that employers almost certainly see a married woman and assume, often correctly, their job won't be their foremost concern. They'll also see a younger woman with a ring on her finger and think, "maternity leave".
I'm unsure which is the more significant causal factor between human capital and signalling because they work together. I'll lazily call it half and half until I see better data.