http://www.economist.com/blogs/economist-explains/2016/02/economist-explains-8

EVERYONE knows that labour markets are not fair. Whether it is skin colour, gender, or some other characteristic, minority groups tend to fare worse than the one group that, at least on average, seems to live life with the "easy" setting on: white, well-educated, men. For every dollar earned by a white non-hispanic man in full-time work, the average white woman in America earns 78 cents, and an average Latina woman only 56 cents. Gay men are no exception to this: even knocking out the influence of factors like education and experience they earn less on average than straight men, by around 5% in France and the UK, and 12-16% in Canada and America. But one minority group seems to do better than others: lesbians. Why?

Research into this area is tricky; getting decent data is hard, and asking people to reveal their sexual orientation can be even harder. But studies across the world (in Canada, America, Germany, Britain and the Netherlands) tend to uncover the same phenomenon; while gay men suffer an earnings penalty, gay women seem to earn more than straight women. In a survey of 29 studies published in January 2015, Marieka Klawitter of the University of Washington found an average earnings premium of 9% for lesbians over heterosexual women, compared to a penalty of 11% for gay men.

Establishing with certainty why this premium exists may be an impossible task, but various theories have emerged. One possibility is that lesbians might face positive discrimination, perhaps if employers expect them to be more competitive and more committed to work than their straight female colleagues. One study did find that in the (less heavily regulated) private sector the penalty for gay men was heavier and the premium for lesbians was larger, which is consistent with this theory. Another idea is that lesbians are responding to the gender of their likely partner. They might have to work harder to plump up household income in the absence of a male partner. Or, it could be that in same-sex couples women find it easier to shrug off expectations that they will take on the bulk of childcare or household chores. Same-sex couples do seem more likely to be dual-earners, even when there are children, and they also appear to share chores more equally than different-sex ones.

If the last mechanism is the one at play, then it could be that lesbians do face discrimination in the labour market, but just not as much as heterosexual women, and shows up as a wage premium. But lesbians are not a privileged group. Qualitative studies have found that they face discrimination in hiring processes relative to heterosexual women, and although they might earn more than straight women, they still earn less than men. Poverty rates among lesbian couples are 7.9%, compared to 6.6% among different-sex couples. Work is still predominantly a (straight, white) man’s world.