It is no longer news to report on the discouraging trends in non-marital childbearing in the United States since the early 1960s.

So compelling were these trends already in 1984 that sociologist Charles Murray took that single statistic as the foundation for a larger argument about the failure of the social programs known collectively as the war on poverty. Ten years after that, Murray’s Losing Ground was reprinted, and prominent economists began paying attention to the trend as a matter of seriousness and importance. Congress and the Clinton administration took up the challenge of a major overhaul of the welfare system with non-marital childbearing as one of the primary foci of that reform. Why all the fuss?

Here are the endpoints. In 1960, the percent of all births to unmarried mothers was around five percent nationwide. Today, that proportion stands at approximately 41 percent. Growth in non-marital births has been equally striking across racial groups—with the largest gains in percentage terms among white Americans, and the largest current proportion of non-marital births occurring among black American women—at just under 70 percent.

While these figures have been well reported, what is less well known is that the growth in non-marital births has been a phenomenon concentrated much more heavily among poor and less-educated women. A recent report by the U.S. Census Bureau summarizing 2011 American Community Survey data reports that a staggering 68.9 percent of women in the very lowest income bracket (household income less than $10,000) experienced non-marital births, compared with 9.0 percent in the highest bracket (household income more than $200k). Between these endpoints, the percentages of unmarried births decrease sequentially with each higher income level. Correspondingly, in the same data 57.0 percent of births to women with a high school degree or less were non-marital, compared with only 8.8 percent among women with a four-year college degree.

Understanding the trend and its covariates matters for a variety of well-documented reasons. Children born outside of marriage are more likely to grow up in poverty, exhibit poor developmental outcomes, and experience unstable family arrangements. At the same time, women and men who give birth to children outside of marriage are more likely to stay poor. Together these facts are related to the striking finding, uncovered only recently, that the single largest correlative predictor of economic mobility at the individual and community level is the fraction of children being raised by single moms. Chetty et al. state, the strongest predictors of upward mobility are measures of family structure such as the fraction of single parents in the area. "As with race, parents' marital status does not matter purely through its effects at the individual level. Children of married parents also have higher rates of upward mobility if they live in communities with fewer single parents."

These findings link the trends in non-marital births with something more than the sum of individual outcomes: the near-term health of the American dream. They also hint at a further possible concern, which is the future of egalitarianism itself. Indeed, lagging neatly behind the trend in non-marital births has been another sizable shift: rising income inequality. The income share of the top one percent moved from about 33 percent in 1970 to nearly 50 percent in 2005. This is the widest spread in income in the U.S. since 1917—except for the year prior to the stock market crash of 1929.  Establishing causal relationships at this level of social analysis is econometrically formidable. One would not want to rush into an off-hand claim about causality that is really a story about simultaneous effects.  Nevertheless, the correlations suggest that huge shares of non-marital births are at the very least an index of brave challenges to the realization of a just social order.

So what do we know about the causes of rising non-marital births? Sociologists Kathryn Edin and Maria Kefalas put it well: “Social science currently tells us much more about what doesn’t explain the trend [the dramatic break between marriage and child-rearing] than what does …” After a landmark five-year study, Edin and Kefalas found that urban non-marital childbearing is not easily explained by the simple story of contraceptive failure nor by the oft-cited proposition that the desirability of marriage has declined within low-income communities. Edin and Kefalas report that they find “surprisingly little evidence of the much-touted rejection of the institution of marriage among the poor … Marriage was a dream that most still longed for, a luxury they hoped to indulge in someday when the time was right, but generally not something they saw happening in the near, or even foreseeable, future.” Non-marital childbearing, the researchers argue, is a more complex phenomenon, one that is linked to the quest for meaningful relationships in the lives of poor women. As the researchers put it, “While the poor women we interviewed saw marriage as a luxury, something they aspired to but feared they might never achieve, they judged children to be a necessity, an absolutely essential part of a young woman’s life, the chief source of identity and meaning.” These findings have provoked a renewed effort to understand the causes and consequences of non-marital childbearing. The Fragile Families and Child Wellbeing Study, led by a team of researchers from Princeton University and Columbia University, is one example of a longitudinal study designed exactly for this purpose.

From the perspective of demographics and economics, the story of non-marital childbearing is equally complicated. The most important fertility-related change of mid-20th century was arguably the introduction of the Pill, and yet it would seem that the Pill ought to have exactly reduced non-marital fertility if it had any causal effect at all. The revolutionary advance in reproductive control was expected to enable women to time births optimally with respect to education and career investments, as well as with respect to relationships and marriage. And in fact, available empirical evidence suggests that the Pill did just this—at least for certain groups of women. Seminal work by Harvard economists Claudia Goldin and Larry Katz has shown that the Pill facilitated widespread increases in education and career preparation, induced a later age for first marriage, and delayed—or re-timed—childbearing. In more recent work, Martha Bailey provides evidence that the Pill may be causally connected to decreases in marital childbearing related to labor market participation. So there is good empirical support for the expected relationship between the Pill and consequent education, fertility, marriage, and labor market outcomes for women. For the most part, these outcomes have been unqualifiedly positive: more higher education, more advanced degrees, more women entering professions, marriages optimally timed, and children planned more effectively.

Theoretically, however, these findings do not close the case. Economists have also pointed out that the introduction of a new technology normally creates winners and losers. An important paper by Nobel prize-winning economist George Akerlof (together with Janet Yellen and Michael Katz) first made this case regarding reproductive technologies. Akerlof et al. demonstrated that male behavior could generate more non-marital childbearing following the introduction of improved fertility-control through a mechanism in which sexual norms shift regarding pre-marital sex, and the culture of “shotgun” marriage is simultaneously undermined (since the probability of conceiving a child is so low). In the theoretical narrative, some women who conceive a child under the new regime might want to keep that child and yet be unable to find a spouse. Goldin and Katz reinforce this same point in a different way. They provide a model of marriage markets in which a re-ordering occurs following any technological change that facilitates female employment. In the model, the change essentially improves the marriage prospects of some women who enter the labor market, and lowers the average match quality of women who do not—women formerly evaluated solely on non-market terms (e.g. home production). Thus, Goldin and Katz make a plausible case that the Pill may have reduced the marriage prospects of some women, potentially increasing non-marital childbearing among women with the lowest labor-market potential.

Taken together then, these theoretical and empirical arguments would suggest that the Pill has had not one but two distinct effects on American women,treating two groups substantially differently. The first effect was to improve the educational and career outcomes of women in the middle- to upper-level socio-economic brackets. This correspondingly increased the opportunity cost of having children, leading to decreased marital fertility and smaller family sizes. The second effect was to weaken the marriage prospects of women in the lowest socio-economic brackets. Without facing the same opportunity costs to childbearing as their well-educated counterparts, the weakened marriage market led to an increase in non-marital childbearing in this group. Strikingly, these two effects appear to match the observed income/education split in non-marital childbearing highlighted above, as well as the reduced family sizes of post-1960s America. Further, this tale-of-two-effects suggests that the Pill might have introduced a destructive modification into the lives of the poorest women in society. Notably, such a modification would arise from social forces beyond the control of poor women themselves: namely, strategic male behavior in private relationships, and the blind operation of unregulated marriage markets that became, perhaps imperceptibly at first, stratified on the basis of economic bargaining power.

Careful statistical work is very much needed to identify the causal role of the Pill on the profound changes in non-marital births observed in the second half of the 20th century. At the same time, additional work on the whole story of fragile families should be pursued vigorously for two important reasons: first, because unstable family patterns are not a mere statistic. They are the genesis of a tremendous weight of human suffering at the individual and community levels. And second, we have yet to understand how patterns of family formation give rise to, and are consequent upon, the economic health of a nation. We need to get the role that the family plays in matters of social justice, such as economic mobility and the tempering of gross economic inequality, right.

Historically various strains of thought on the left have located a major source of inequality in the family itself, through channels of social status, wealth accumulation, and the transmission of human capital. But the above discussion points to a competing or counterbalancing relationship between the family and inequality in which the family might act as a natural check on extremes at both ends of the economic spectrum. On the bottom end, marriage among the poorest classes may facilitate economic mobility for children as well as for adults. On the upper end, larger family sizes might have a tempering effect on the tremendous concentration of wealth in the upper one percent. These considerations, offered only speculatively, hint at new avenues for illuminating the left-right debates on inequality that continue to plague public discourse.