It estimates that, among heterosexual men working over 30 hours a week, married ones earn almost 20% more than singletons; this controls for age and education among other things
However, among gay men, the marriage premium - well, partnership premium - is just 4%.
This poses a challenge for those who think the marriage premium is due to selection effects. After all, if it were the case that the same things that make a man attractive to an employer also make him marriageable, shouldn’t this effect be as powerful for gays as straights?*
Instead, this might suggest that much of the marriage premium is causal. Maybe marrying a woman causes a man to earn more, perhaps because he expects to have to provide for children, or because he can focus more on his careers as his wife saves him time by doing housework. For gay men, marriage doesn’t have these boosts to earnings.
Does this suggest that we should encourage marriage, because it raises productivity?
No. The opposite is true for women. Married women earn almost 4% less than single ones - even controlling for having children. Maybe looking after hubby distracts a woman from her work.
What’s more, whereas being gay is bad for a man’s earnings, it’s great for a woman’s. Lesbians earn more than straight women; this seems true in the UK as well as Canada. And whereas a woman loses money by marrying a man, she gains by shacking up with another woman; lesbians in couples earn 10% more than married women - again, controlling for having children.
If we judge policies purely by their labour market impact, it makes as much sense to encourage women to become lesbians as it does to promote heterosexual marriage - though such a policy has other drawbacks.
However, what stands out about this research is that bisexuals, men or women, earn much less than gays or straights: it‘s not clear why.
All this raises a question. Why do we hear so much about the gender pay gap, when much of this is due simply to women having children, and much less about sexuality pay gaps (gay vs straight men, bis vs others) when these might arguably reflect deeper injustices?
* Maybe not. It could be that what employers value is a conventional mindset, and married men have this whilst married gays don’t.