Whenever the subject turns to inequality and opportunity in America, the same political conversation repeats itself.
First, liberals call for more spending to raise people out of poverty. Then conservatives point out that a major predictor of poverty is family instability, and that one of the biggest boosts to opportunity comes from having married parents.
Then liberals retort: "Well, the best way to help families is to make sure they have more money. And what's the alternative - having divorced politicians hector poor women about their choices?"
Thanks to sweeping new research on American mobility, we can have this predictable debate again. Conservatives can highlight the strong correlation in the research between two-parent families and upward mobility - stronger than for variables like racial segregation and economic inequality. And liberals can dismiss marriage promotion as a chimerical goal, and double down on redistribution.
Or both sides could be a bit more honest about the roots of marriage's decline.
Honesty from conservatives would begin by acknowledging that policies championed on the right - mass incarceration in response to the post-1960s crime wave, Bain Capital-style "creative destruction" in response to Carter-era stagnation - have often made it harder for low-income men to find steady work and stay out of prison, and made women understandably wary of marrying them.
Then this honesty would continue with a concession that certain kinds of redistribution - especially if tied to wage-earning - might help make men more marriageable, families more stable, and touch off a virtuous interaction between the financial and the personal.
Right now, I think some conservatives - though not enough Republican politicians - are willing to concede these points. But I don't see a readiness among liberals to make any concessions of their own, beyond the minimal acknowledgment that all things being equal, two parents are often better than one.
A more significant concession would be to acknowledge the ways in which liberalism itself has undercut the two-parent family - through the liberal-dominated culture industry's permissive, reductive attitudes toward sex, and through the 1970s-era revolution in divorce and abortion law.
In the first case, liberals tend to feign agnosticism about pop culture's impact on morals (even though a link is common-sensical and well supported), or to blame corporate capitalism for the entertainment industry's exploitative tendencies (as though the overwhelmingly liberal people making programming decisions had no agency of their own).
In the case of abortion and divorce, liberals expected their revolution to, if anything, stabilize the family - by reducing unwanted births and dissolving only marriages that had failed in all but name.
But these expectations were naive. As Janet Yellen and George Akerlof pointed out in a 1996 paper on the social impact of abortion and contraception, the power Roe v. Wade gave women over reproduction sometimes came at the expense of power in relationships. "By making the birth of the child the physical choice of the mother," they noted, the sexual revolution "made marriage and child support a social choice of the father."
In this new landscape, "women who wanted children, who did not want an abortion for moral or religious reasons, or who were unreliable in their use of contraception" saw their partners' incentives altered for the worse. The result was a world with plenty of unplanned pregnancies but fewer ensuing marriages, fewer involved fathers, more unstable homes.
Meanwhile, no-fault divorce probably contributed to the unexpected "social contagion" effect of the divorce revolution, in which the example of a marital split undermines marriages across a social network. And it created new reasons to delay marriage in the first place, given the risk of investing in a venture that could be unilaterally dissolved.
Many marriages, especially in the upper-middle class, were strengthened by caution and delay. But for couples with more limited resources, and more to lose from failure, no-fault divorce may have reduced the value of the institution and the sacrifices embraced on its behalf.
When liberals claim social conservatives don't have any policy ideas for marriage promotion, then, they're somewhat self-deceived. A sustained conservative shift on abortion policy and marriage law probably would, over the long term, increase the rate at which couples take vows and stay together, and improve the life prospects of their children.
So one hypothetical middle ground on marriage promotion might involve wage subsidies and modest limits on unilateral divorce, or a jobs program and a second-trimester abortion ban.
The chances of liberals embracing this hypothetical are ... nonexistent. Which is fair enough - the belief that some personal liberties are nonnegotiable whatever their policy implications is obviously sincerely held, and I can respect a worldview that privileges moral absolutes over social science evidence.
But the evidence remains. And so, for families and children, do the consequences.
Ross Douthat is a columnist for The New York Times.
By ROSS DOUTHAT