Raj Chetty’s Equality of Opportunity project has claimed that the extent of upward mobility in an area, for boys in particular, is causally linked with the predominance of single-parent families. This has led to a spat between Philip Cohen and W. Bradford Wilcox over the extent to which these findings are driven by the effects of racial inequality (ie, black families are overwhelmingly single-parent, and so the distribution of single-parent families ends up being largely a map of black America) rather than the effects of family structure itself.
While I would agree with Cohen that Chetty’s methods do not allow him definitively to distinguish between the effects of racial inequality and the effects of family structure, I won’t pretend to be undecided or unbiased here. My own view follows Sara Mclanahan, that family structure is not merely a confound of racial inequality, but rather an important contributor to it. Mclanahan and Christine Percheski write, in a well-known essay, Family Structure and the Reproduction of Inequalities:
Over the past four decades, income inequality has increased and family structures have diversified. We argue that family structure has become an important mechanism for the reproduction of class, race, and gender inequalities. We review studies of income inequality and family structure changes and find a wide range of estimates of the correlation. We discuss how increases in income inequality may lead to increases in single motherhood, particularly among less educated women. Single motherhood in turn decreases intergenerational economic mobility by affecting children’s material resources and the parenting they experience. Because of the unequal distribution of family structure by race and the negative effects of single motherhood, family structure changes exacerbate racial inequalities.
At this point, it might be useful to step back and ask why marriage rates have declined in the first place, putting aside the effects of that decline.
The technology of birth control and the cultural and political changes associated with feminism and women’s mass entry into the workforce are obviously hugely important. But given that marriage rates have stayed high (and divorce rates have decreased) selectively for high-income and well-educated people in liberal as well as conservative states, cultural explanations seem insufficient in themselves.
Another, more plausible explanation is the very high marginal tax rates that lower income people face when considering marriage, because of the presence of means-tested welfare programs.* There is little doubt that the huge expansion of means-tested welfare programs in the late 60s pushed lower-education women away from marriage. Even a study that concludes that the attempts at marriage promotion of the 90s welfare reform were largely unsuccessful notes that the effects of previous expansions in the welfare state were less ambiguous:
Many studies have concluded that more generous welfare programs are associated with higher rates of female household headship and lower rates of marriage (e.g., Hoynes 1997; Grogger and Bronars 2001; Lichter, LeClere, and McLaughlin 1991; Lichter, McLaughlin, and Ribar 1997, 2002; McLanahan and Casper 1995; Moffitt 1994; Schultz 1994; and the references therein).
That’s one side of the story. The other, less congenial to many conservatives, is the relentless downward trend in men’s median wages, the even more relentless downward trend in labor force participation, and the massive increase in incarceration particularly among black men. There are a lot of stories conspiring to produce those changes, but undoubtedly the net effect is to make men less appealing as marriage partners almost across the board, and particularly in low-income black neighborhoods. This may be even more true if it is relative wages that are most important. As discussed in a recent study by Marianne Bertrand and Emir Kamenica, “Within marriage markets, when a randomly chosen woman becomes more likely to earn more than a randomly chosen man, marriage rates decline. ” The general upward trend of black women’s earnings and educational attainment might therefore compound the downward trend of black men’s situations in its impacts on marriage rates.
The meta-analytic challenge in discussing these issues is that decreases in marriage rates are both effect and cause of more liberal politics (or if you prefer, maintaining high marriage rates within a group are both an effect and a cause of more conservative politics.) There is an incentive for participants to dissemble for this reason, or to ignore inconvenient findings, beyond their individual research agenda or intellectual commitments.
So, let’s at least agree on a few stylized facts:
1) The portion of people married at each income and education level is steadily dropping, with lower education and lower income individuals dropping much faster.
2) It’s not actually surprising– for this very reason– that many blue state liberals are in stable marriages; the most liberal states are the wealthiest ones
3) In spite of this income gap, it’s unclear whether marriages are more stable in liberal or conservative areas:
4) More unambiguously, growing up in more conservative areas tend to increase the likelihood of marriage later in life.
5) Married people tend to become politically conservative, even controlling for income, race, and age:
For this reason, and given the strong political orientation of most academics and journalists concerned with family inequality, it’s not shocking that each release of evidence that there are social costs to marital instability is accompanied by angry denunciations of those who release the evidence.
*Sometimes these incentive effects of means-tested programs are shrugged off as too abstract and recondite to influence the behavior of low-income people on a subject as personal as marriage. My own anecdotal experience would contradict this. I know well one couple with kids who elected not to get married because the woman would lose Section 8 housing vouchers, and be forced to leave their NYC neighborhood where the father worked and where all their friends and family resided. At the time, they thought of themselves as married “in all but name” (in fact, I was surprised when the father told me, after coming to my wedding, that they weren’t married) but the truth is that marriage is a well designed institution for enforcing commitment (the sheer embarrassment of telling everyone you know that you’re getting divorced is helpful in weathering many a marital rough patch) and, like most unmarried lower income couples with children, they ended up splitting up. The interior of a marriage is a bit like Schrodinger’s Cat– those of us on the outside of the box never know what’s going on inside or why– but I can’t help but feel that the formality and ritual of marriage makes a difference.