Families, Freedom and Responsibility

Ta-Nehisi Coates did a post that described his non-traditional family that somehow produced kids without an abnormal level of dysfunction, and in fact, a whole lot of high-function. Ross Douthat had a very nuanced and balanced take on it, with which I broadly agree. There is a difference in shading, however, that I think would tend to lead in somewhat different policy direction (not that either of them were getting into policy, but somebody has to be the dork here). Ross made the point that pointing to your non-traditional family that seemed to do pretty darn well is, in effect, like pointing to your 97 year-old uncle who smokes three packs a day and saying that this calls into question all this stuff about smoking causing cancer:

But on the other hand, the generalizations matter too. The “artifice” of the traditional family isn’t just an artifice, and the values that social conservatives hold so dear – monogamy, marriage vows, the idea that every kid deserves a mother and a father in his life – don’t just exist to make people in non-traditional families feel bad about themselves. In the aggregate, Dan Quayle was right. In the aggregate, marriage is better for kids than single parenthood. In the aggregate, marriage is better for men and women than long-term cohabitation. In the aggregate, divorce is bad news – for your finances, your health, and your children’s long-term prospects. And in the aggregate, if you’re concerned about income inequality or social mobility or the crime rate or just about any area of socioeconomic concern, then you should be at least moderately fretful about the long, slow decline of the American two-parent family – among blacks, whites, and Hispanics alike.

But I think what is in dispute here is whether these statements of cause-and-effect are correct. I remember reading the famous “Dan Quayle Was Right” article when it came out in 1993. I had the same issue with it then as now. It asserted scientific knowledge about these questions:

According to a growing body of social-scientific evidence, children in families disrupted by divorce and out-of-wedlock birth do worse than children in intact families on several measures of well-being. Children in single-parent families are six times as likely to be poor. They are also likely to stay poor longer. Twenty-two percent of children in one-parent families will experience poverty during childhood for seven years or more, as compared with only two percent of children in two parent families. A 1988 survey by the National Center for Health Statistics found that children in single-parent families are two to three times as likely as children in two-parent families to have emotional and behavioral problems. They are also more likely to drop out of high school, to get pregnant as teenagers, to abuse drugs, and to be in trouble with the law. Compared with children in intact families, children from disrupted families are at a much higher risk for physical or sexual abuse.

All this says, of course, is that these problems are more prevalent among children of divorce or illegitimacy than intact families. What we don’t know are the counterfactuals: how would these kids have fared if their parents had stayed married or never been married? Despite the “social scientific evidence”, it’s hard to see, at least in a non-totalitarian society, how we could construct random assignment trials to tease this out.

This doesn’t mean that I think Ross’s view on the best family structure on average is wrong. But it should lead us to a whole lot of humility about this belief, and especially that (1) it may not be right for everybody (which Ross doesn’t assert), and (2) that changes in the economy, technology or lots of other things might even change this for the average person over long periods of time. Especially if (2) is correct, the future is likely to arrive unevenly and imperceptibly, and so we would want less central control over this. We would want lots of different arrangements attempted in order to see which ones work, and should expect a variety of different family types to be present. Even those that are not widespread are “reserve options”, like limited species in ecology, that can take off more rapidly when and if conditions change.

The obvious approach in this situation is the libertarian formula: “To each his own”. But there are a couple of limitations to this.

First, if you’re going to live in a way that those around you find immoral, or inadvisable or, at minimum, highly risky, then don’t expect them to provide a whole lot of help when and if you need it. It would not only feel unfair to do this, but more importantly, would prevent forces of cultural evolution from eliminating those family forms which can’t compete. It would stop adaptation. A hothouse effect that simultaneously permits freedom and decouples performance from reward is a breeding ground for decadence. Symmetrically, if you assert that I have a moral requirement to help those in extreme need, then be prepared for strings to be attached to that assistance. If you don’t want the strings, don’t take the assistance.

Second, there is the free-rider problem. As Heather Mac Donald has pointed out many times, it’s one thing to be the one non-traditional family in a neighborhood that can rely on decent public schools, safe streets, cultural trust and so on that (by this theory) require either other traditional families and/or coercive restraints on behavior, but your non-traditional family would have a much tougher time doing well if you had to live in a free society composed predominantly of other non-traditional families. This is unproven, in the same way that the so-called social-scientific findings about families are unproven, but is a theory that, it seems to me, people should be able to test through experience. This observation leads to, as I’ve often argued, allowing experimentation not only at the level of individuals, but also for various neighborhoods, towns, states and so on the have wide flexibility in the kinds of coercive community laws they establish.

In other words, we’re back to my old hobby-horse when it comes to social issues: subsidiarity.