Marital instability, as we’ve often noted in this space, is associated with higher rates of poverty among women and children. But the decline of marriage also includes people replacing marriage with cohabitation, sometimes with multiple partners over time. Like divorce, the breakdown of a cohabiting relationship can plunge a woman and her kids (deeper) into poverty. But which would be worse? That is, when a couple with children splits up, does the woman’s household income take a bigger hit if the couple was married or if they were cohabiting?

Even thinking about that question—let alone generating the answer through appropriate research methods—is surprisingly complex. Less income pooling and less traditional gender-role ideologies among cohabiting couples together with more instability in informal unions could mean that cohabiters are less interdependent and therefore better-prepared to split. On the other hand, economic vulnerability contributes to cohabitation: those with unstable incomes are more likely to cohabit than to marry, and the savings associated with sharing housing costs are more important at lower income levels. Add to that the fact that cohabiting women are more likely to qualify for means-tested government programs than married women because many programs do not count cohabiting partners’ incomes, and it is far from clear which group would see the greater drop in income after their union dissolved.

Despite all the theoretical complexity, in a recent Demography article Laura Tach and Alicia Eads were able to demonstrate that from 2000 to 2007, legal status didn’t matter much: for (formerly) married women with children, household income was 33 percent lower a year after a divorce than the year before it; the drop for cohabiting mothers following a break-up was 30 percent. But the two situations weren’t always so similar, Tach and Eads found. Trends for the formerly married and the formerly cohabiting diverged sharply between the 1980s and the 2000s. Back in the 1980s, women with children had household incomes fall 42 percent in the wake of a divorce, but “only” 20 percent when a cohabiting union dissolved. Over time, the monetary cost of splitting has decreased among the married and increased among the cohabiting.

The temptation for those of us familiar with rhetoric about marriage becoming a luxury good—increasingly something that only the elite can reach—is to interpret these changes through that lens. It seems perfectly reasonable to assume that dissolving a marriage has come to cost less over time because the ranks of the married have come to be dominated by elite women who can do relatively well financially even without a second income, and to interpret the trend among the cohabiting as evidence of their increasingly disadvantaged status. After all, cohabitation has been rising and especially so among the poorer, so wouldn’t a break-up hurt the poorer more?

That logic does not stand up to the evidence. Tach and Eads found that even among the least educated, the monetary cost of divorce dropped over time while the monetary cost of dissolving a cohabiting union increased. The trends can’t be fully explained by class differences in access to marriage, for the trends pertain regardless of class.

Trends in cohabitation now contribute more to rising inequality than trends in marriage.

So why has divorce come to cost less while breaking up a cohabiting union has come to cost more? Tach and Eads find it’s mostly attributable to married women’s earnings: over the time period they cover, married women’s earnings rose more than married men’s earnings. In other words, married women’s own earnings became a bigger share of their household income. Cohabiting women’s own earnings did not. This gave married women a relative advantage after a break-up.

Married women also benefit more from child support after a break-up than cohabiting women do. Why? Because cohabiting women were more likely to be receiving child support before their unions dissolved—they were more likely to have children from former unions. Child support helps buffer the economic loss of a break-up for women with children only if the children came from the newly broken union. That’s hardly rocket science, but it translates into a relative advantage for formerly married women because a smaller share of their children came from prior relationships. Since the 1980s, enforcement of child support has been ratcheted up, and more so for women who have given birth outside of marriage—fathers were more likely to be held accountable for the support of children born during marriage even before the increased enforcement. Nonetheless, child support still provides a more important buffer after marriage than after cohabitation.

The story with respect to government transfers runs the same way: cohabiting women were more likely than married women to be receiving transfers during their unions and therefore less likely to see increases after their break-ups. In contrast, married women were more likely to become eligible for transfers like the earned income tax credit after they divorced. (Tach and Eads’s overall findings did not depend on whether the EITC and housing subsidies were included in household income, however.)

Social class clearly plays a role here: cohabiting women have seen relatively poor earnings growth over time, plus they have already fallen onto safety nets that married women might fall onto after a divorce. Union instability in prior relationships also clearly matters in that the costs of dissolution are higher for mothers when the buffer provided by child support is already in place—when the woman’s latest partner has no financial obligation to her child. Plus social class and union instability both depend on each other: break-ups lead to lower household incomes among women with children, and unions are less stable at lower income levels.

Tachs and Eads are among many who argue that union instability contributes to rising inequality in the United States, but this recent article indicates that the costs associated with trends in cohabitation now contribute more to rising inequality than trends in marriage. Marital break-up still costs heavily, but cohabitation has become more common and disproportionately affects low-income children. These are the children for whom the costs of a split have increased across the past three decades. So anyone who thinks that breaking off a cohabitation is less consequential than divorce should consider the monetary costs to women with children: they are great for the individuals involved, and they contribute to rising income inequality nationwide.