Marriages can be a volatile affair, especially when the spouses have little incentives of staying together. Traditionally, marriage has offered one big economic advantage over celibacy: specialization. And once spouses are locked into their specific tasks, divorce becomes very costly as the husband was not good at household tasks while the wife would not have fared well on the labor market.
But with the dramatic increase in female labor market participation in the past century, this has changed. As women have better opportunities on the market, they do not feel tied to marriage as much. The conjecture is then that the higher the income share of the bread-winner, the more stable the marriage is. Kornelius Kraft and Stefanie Neimann study this using the ever-giving data-set, the German Socio-Economic Panel, and confirm these conjectures except for one. The income proportion theory is not gender-neutral: female bread-winners are more likely to divorce than male ones.
I find this particularly interesting because it gives a good explanation for the increase in divorce rates across all countries. And do not tell me it is because divorce laws changed. These laws were modified under pressure.