[A] woman's employment status has no effect on the likelihood that her husband will opt to leave the marriage. An employed woman is more likely to initiate a divorce than a woman who is not employed, but only when she reports being highly unsatisfied with the marriage. . . .
For a man, not being employed not only increases the chances that his wife will initiate divorce, but also that he will be the one who opts to leave. Even men who are relatively happy in their marriages are more likely to leave if they are not employed[.]
The finding seems to reflect data that I've cited in the past on the subject that appear to show that economics is still probably the strongest tie that holds marriages together and the biggest influence on recent trends in divorce rates, but that the dynamic is complex. In a nutshell, the key economic factor seems to be the economic dependence of women on men in a marriage. When it is present, divorce is rare. When it is absent, marriages tends to survive in good times, but not in bad ones.
I've hypothesized in previous posts that the driving factor behind the divide in divorce rates between college educated and non-college educated couples is that college educated women relative to non-college educated women (1) have more to gain economically from their husbands (who have better incomes and often benefits like favorably priced health insurance and defined benefit pension plans), (2) sacrifice more of their earning potential by being married and having kids (first, because they actually stay at home as a result, and second, because experience matters much more in professional careers than in blue collar jobs), and (3) have lower expectations that they will be able to secure a match or new job that puts them in at least as good a position as they are in when married (because they have more room to do worse and less room to improve, and because the better men tend to stay married while the less desirable ones do not, on average).
Thus, even though college educated women (at least when they have children and are married to college educated men) generally have more earning capacity than non-college educated women (at least when they have children and are married to non-college educated men), also tend to have a greater economic dependency upon their husbands.
Other data have tended to show in multiple countries that an important factor in divorce rates of couples relative to each other within countries is the ratio of a husband's income to a wife's income. Marriages in which wives have more earning power relative to their husbands are more likely to end in divorce.
A Father's Day column by Vincent Carrol in the Denver Post questioned this economic interpretation for a rise in divorce rates, noting that job prospects for men aren't that much worse than they were in the 1960s and that in absolute terms couples are strong married in hard economic times because they can pool income and reduce expenses than they are when they are not together. But, his analysis paints with too broad a brush.
The men who are most likely to get divorced, those who lack college educations, have faced four straight decades of near total economic stagnation, sharing in almost none of the economic growth of that overall prosperous era, while better educated professional men have shared in that economic growth. Meanwhile, the economic prospects of women have improved dramatically in those four decades in all social classes.
But, while much better than they were in the early 1960s (when women made up single digit percentages of most professions and often remained unmarried and childless if they chose to do so at all), the economic gains for higher level women professionals who marry and have kids have consistently shown the greatest lag behind men in the same professions. The "mommy track" career pattern, of choosing to take time off from work to raise young children and to take positions within a profession that provide a better work-family balance to a greater degree than men, as much as more blatant discrimination in compensation between men and women, accounts for a very large share of the enter gender gap in compensation today. Women who don't marry and have kids and have college educations have incomes very similar to men with college educations (even if they have married and/or had kids).
The "mommy track" gap is very top heavy. Low skilled jobs of women with no college education tend not to have very seniority and experienced based pay scales. The seniority and experience factor for women who are lawyers, and doctors and business chiefs, in contrast, as illustrated with data like the University of Michigan Law School alumni survey, can be intense.
Considering policy responses to this situation is beyond the scope of this post. Obviously, nobody is seriously proposing that we should return to the 1950s regime when women were all but barred from almost all professions by school teaching, nursing and secretarial work (and professional such as laundry work and child care that constituted outsourced forms of homemaking). Neither do we want to return to the Baby Boom era barefoot and pregnant model in which economic dependence that strengthened marriages flowed from having children early and often.
It may be that marriages that are too strong aren't even desirable. The stronger a marriage is, as a result of economic dependence, the more an economically dependent wife will tend to tolerate mistreatment in that marriage, up to and including domestic violence, a husband that engaged in incest with her children, affairs, and so on.
But, it is also the case that we either have to learn to find ways to make a reality with less stable marriages work better, or rethink the economic incentives within marriage and the way that women choose to be involved in the workforce, if we wish to improve the status quo.
Addressing issues that aren't driving the problems won't produce solutions. Even changing the standards under which marriages are entered into, or divorce is available wouldn't necessarily work. In the era well before no fault divorce, when economic trends weakened marriages arose, divorces and de facto divorces became more common and changes in the law regarding the availability of divorce, followed rather than preceded this trend. Likewise, parallel to the trend of high divorce rates in recent decades (although they have started to fall again) is the increasingly widespread trend of couples never getting married in the first place.