Thursday, November 30, 2006

Assortive Mating

My recent postings on tribal identification and other primitive parts of our personalities gives one reason why assortive mating would be a topic of fascination for me. The other reason, likely much more important, is the idea of legacy. I have four sons, and however much we kid them about marrying them off and how awful their own children will be to deal with, the plain fact is that 4 is a very small sample size, and individual differences and destinies will affect their future lives more than the averages of assortive mating will. Still, as parents, we try to set things up for our descendants to put the averages in their favor as much as possible. Did I marry the right sort of person myself? Have I been a parent who trained children to be parents themselves? Have I provided sufficient warning and insulation against their divorcing or having children outside of marriage? There are deeper items of legacy as well, spiritual legacy, emotional legacy, cultural legacy. I would far rather talk about those, but don’t really know what to say.

It is not that I am here conscious of my audience, which includes a goodly percentage of young friends-and-relations, including my own children. I am actually less conscious of addressing them than I perhaps should be. In discussing large abstract issues, I likely make comments which could wound those I know, because I wasn’t thinking of them specifically when I made a sweeping generalization. My audience should be what makes the discussion difficult, but it isn’t.

Legacies are transmitted bone-to-bone, and thus defy easy description. What children receive is also not quite what we thought we bequeathed. We threw in lots of extra stuff, good and bad, that we hadn’t realized, and never did pass on some things we swore were important to us. Confusing the issue further, each child in any family has different parents. The interactions are different from Day One. And with my two Romanians, who I didn’t even have until Day Five Thousand and Day Six Thousand of their lives, legacy is even more a roll of the dice.

These are close to the central facts of my life, and I haven’t a clue what to say. So I’ll go back to the grand theoretical constructs.

Kay Hymowitz’s new book Marriage and Caste is one of those sound-the-alarm books. She has looked at the numbers for education, stable marriage, unwed parenting, cohabitation, and divorce and is very worried that we are rapidly dividing into two Americas: one educated and married, the other less educated and divorced, single, or cohabiting, with disastrous consequences for the children. She summarizes this material in an essay over at City Journal. It is disquieting. The advantages of having two parents is great – greater than popular culture supposes - and one large portion of the rising generation already has advantages. Two difficulties with the Hymowitz information strike me. 1) What’s new? Even in elementary school I could have pointed out to you who had “advantages” and who didn’t. What are considered the main advantages have changed over time in our society, but they have been drawn from the same short list. These are percentages, not destinies. Perhaps I say this a bit defensively, as I was a child of divorced parents in an era (early 1960’s) when that was rare, and more stigmatized. 2) This pattern does not result in a division into castes unless it is worse each successive generation. Children of divorce are more likely to divorce themselves, but is it even more likely if their grandparents were divorced? Or does that disadvantage remain for only a generation each time? The numbers aren’t in for that, I don’t think. If some factor of deterioration – criminality, substance abuse, chronic unemployment, divorce – is 50% in the first generation, but 60% in the second and 70% in the third, then we have a more serious problem. We would then have an underclass which is not merely under, but going further under every year. I don’t think we see that. People still rise from a low place or fall from a high place with fair regularity in American society.

Hymowitz makes much of the fact that elite colleges have a hugely disproportionate percentage of children living with both parents. It worries her that this is the early evidence of a new aristocracy, which will increasingly get the good jobs and rule the rest of the country. If I thought that the graduates of elite colleges actually were destined to rule society, I would be more concerned. I think that day has long passed. Graduates of Princeton and Bryn Mawr will have higher incomes than average, and more than their share of state attorneys general. But they will also have a higher share of Gender Studies and Art History majors, who might go on to have interesting productive lives, but aren’t likely to be ruling too many of their fellows. Materials engineers from state universities are better placed for that.


Anonymous said...

Okay. Here's the thing. I'm generally okay with the idea that two-parent families offer benefits that single-parent families lack. And I see how the state actually has a legitimate interest in furthering two-parent families.

That said, I totally reject the notion that feminism is responsible for the general decline in two-parent families. And I reject the notion that women who chose to escape a marriage are uniformly, or even typically, making bad choices.

The problem is poverty, not single parenthood. The thing is, being a single parent exasterbates poverty. Women of means, who pursued careers and are beneficiaries of feminism, manage with a single income and their children don't necessarily suffer adverse outcomes. We don't really need to worry about them.

For women of less means, feminism is less of a force. Working class women have always joined the workforce. Once they married, they often stopped working and focused on family matters. In the post WWII period, economic concerns kept women from leaving work (despite the idealized image of the 50's home). Two incomes were necessary. In the 60's and 70's, working class women discovered that financial independence from men offered independence from bad husbands (whether abuserers, alchoholics, or philanderers). In an earlier age, women simply had to put up with these men or live on the streets. Now they have options. Are these women divorcing an income as well as a drunk? Certainly, but lets not just assume they are making unwise choices that are detrimental to their children.

The real issue that has changed for a net negative is children born to single mothers. That's a problem we'd all love to solve. I would promote the wide availability of birth control and the education of kids about the dangers and difficulties of early parenthood. I think teenagers benefit from abstinence. And women in their twenties benefit from pressure not to have more kids than they can manage. The problem isn't feminism, and it isn't that women are esacping bad relationships or making informed choices about being single or married. The problem is when people make a poor choice on a spring afternoon and end up with a fourth baby.

My six-and-a-half cents.

GraniteDad said...

"Women of means, who pursued careers and are beneficiaries of feminism, manage with a single income and their children don't necessarily suffer adverse outcomes. We don't really need to worry about them."

Does anyone have any source information supporting or refuting this? I'd be interested either way.

Assistant Village Idiot said...

I think mavis's opinion is the standard one you would get from Newsweek, or Oprah, or the traditional media outlets. It's the conventional wisdom. Hymowitz's essay claims otherwise, and she puts up numbers that I don't have a ready explanation around.

As I noted, I think there are questions Ms. Hymowitz doesn't fully answer. I recommend the essay that I linked to, however, for the evidence that she does provide. She would say that single parenthood is statistically so associated with multiple negative outcomes that it is the real cause, not an exacerbation, of poverty. I'll let her make her own argument.

Anonymous said...

It would be interesting to see if anyone has done new studies on the "children of divorced parents are more likely to get divorced."

I realize that it was true for a long time, but now that we are into the fourth decade of easy divorce, I wonder if there is not a reversal of sorts, where children of divorced parents can realize that marriage is something to work on and work for and try to do better than their own parents.

Of course I am falling back on anecdotal evidence - my parents divorced after 27 years, and my husband's and sister-in-law's parents both divorced in their first decade of marriage.

It has seemed for a long time that both of our marriages (my brother's and mine) have been unusually stable, given the children-of-divorced-parents thingie for all four of us (both people of both couples).

I've been somewhat uneasy about this for a long time; my parents divorced six months after I got married, shocking the @#$% out of all of us. However, my mother told me that she and my father had had a troubled marriage from the beginning, but went to great lengths to hide it.

My husband and I have been married for 24 years, and although we have definitely had our rough patches in the first, oh, 17 years or so, we have settled into quite a comfortable and happy relationship. Partly because of what my mother told me, I have had a habit of challenging things that are uncomfortable between us, rather than putting up with them or pretending that they aren't there, so I feel like I can be pretty confident in saying that our relationship is good and strong.

I realize that statistics aren't predictions, especially about individuals, but I really wonder whether our experience is unique to us. My husband's mother raised three boys alone in the sixties, when, like with your family, it was not the done thing. She (Mum) told me that she told the boys that if they couldn't stay married, then they shouldn't get married. It looks like all three of them listened to her: my husband married me when he was 21 and we have stayed married through things that frankly most people would have understood if he'd left (severely handicapped child). His elder brother waited until he was 42 (4 years ago) before he married, and his younger brother hasn't married.

My sister-in-law had an abusive father and step-father. She and my brother waited ten years to get married, and now that they have, they seem to have an exceptionally stable marriage. Staying married and dealing with things is something we talk about and strategize about.

I think it also makes a big difference that we can see the decades-later effect of our parents' divorces. My parents were quite well off; now both of them separately are having to pinch pennies to get by in their seventies. My father, who retired at 55, had to go back to work at 70 because his retirement money ran out. He used to be vice president/controller of a large conglomerate; he is now the bookkeeper for a local company.

Sister-in-law and husband's parents are similarly economically fragile.

Sorry this is going on so long, but I would really be interested in an updated exploration of this children-of-divorced-parents thing. I just would like to see some exploration of the potential for learning from other's situations rather than repeating them.

Anonymous said...

I get the Oprah swipe means you don't think I'm very serious, but that's a lame dismissal of my comments. I don't argue with Hymowitz's numbers, only her reasoning. Neither of us is concerned about educated/wealthier women. Hymowitz isn't worried because the numbers of divorced and separated educated women have declined since the 90's, and never grew too large. I'm not concerned because I'm believe that the children of single women of means do just fine, at least economically. If you think I'm wrong, please show me the numbers. Hymowitz doesn't.

The only time Hymowitz goes into the long term outcomes for children, she presents unconvincing evidence, "She and her colleague Dean Lillard examined the records of students at the nation’s top 50 schools and, much to their surprise, found a similar pattern. Children who did not grow up with their two biological parents, they concluded when they published their findings, were only half as likely to go to a selective college. As adults, they also earned less and had lower occupational status." This begs the question, did they control for poverty? We've just learned that poverty and marriage are negatively correlated. We would expect fewer poor kids at fancy colleges. So naturally more of the kids are from married (and wealthy) families.

The whole article is premised on the notion that it's family structure, not poverty, that determine outcomes for children. Yet she merely shows a correlation between poverty and family structure. The way to show that family structure is more important is to look at the children of wealthy divorced and single mothers and the offspring of poor married families. She doesn't do this.

Assistant Village Idiot said...

Yes, I guess it was a swipe, but not so much at you as the conventional wisdom. Your points are good ones. I don't have the breakdown for the outcomes for children corrected for married/income/race, though I will look for them. If you are looking for them yourself, my only caution would be that major news outlets have audiences, and they tend to highlight items that play to the audience's prejudices. The NYTimes is going to try and bury anything that suggests to wealthy career women that divorce might still be bad for their children. The data might actually go against me on this - I just preach initial caution.

To sum up Hymowitz: of the many factors that go into good parenting outcomes, marriage has been underrated and is actually the dominant factor. The conventional wisdom has been: two-parent families are just one factor out of many.

As you note, Hymowitz doesn't offer those numbers directly. I think they are implicit in the other numbers, however. Looking at the graph, the number of children in the top two quintiles who are with only one parent is fairly small. Are the outcomes for those children about the same as for the children of two-parent families in the quintile? Mavis's contention would be that they are likely quite close. That could be. Let's grant that it is for the moment.

Not all virtues are associated with increased wealth - there may even be some that are negatively associated. But the virtue of planning, looking ahead, counting the cost, is likely strongly associated with it. Tucked in the relatively small numbers of wealth + one-parent will be widows and widowers, who will resemble the marrieds in attitudes if not resources. The group will also include many who have some of the planning-ahead virtue that Hymowitz considers so valuable for bringing up children. It might well be that those parents who objectively consider the impact of marriage/no-marriage, divorce/no divorce and decide to parent singly do not show more negative outcomes.

But then we would still be in a very similar place. The operative factor then would be those who can plan ahead versus those who cannot, not married versus unmarried. Marriage would then be better understood as an indicator of that calculation, rather than the key factor itself.

Tangent: my mother, for example, divorced my father when he went to jail. That was likely the best calculation for herself and her children. That is the sort of planning ahead, counting the cost, that generally prompts people to try and create stable environments for their children, i.e. marriage.

Also, we have already noted the simply practical advantages of two-parents: More time, more extended family, more economic stability.

On the other end of the curve, we have the larger numbers of one-parents. There are still sizable numbers of two-parents in the third and fourth quintiles, however. Enough in fact, to yank down the average for all two-parent families if income itself were a stronger determinant. But the average is not pulled down by them. (I conclude this because if they did, it would mean that the top two quintiles started from an almost impssibly high level.)

What is happening in the bottom quintile two-parent group I don't know. That may deserve some consideration of its own.

mb, I am working on a longer European/American demographics piece which wshould offer some interest to you. Please stick around.

Assistant Village Idiot said...

On the drive up to the play in Concord tonight, I had a quicker and cleaner answer that I think says the same thing.

If it were only a matter of income, then children whose parent remarried would track more like two-bioparent children. According to Hymowitz, they don't; they track like children of single parents for negative outcomes.

As I wrote in the OP, these are percentages, not destinies. A single parent, or one considering single parenthood, either through intentional conception or divorce, can (and should) bring as many compensating factors to bear as possible, and living in a less-dangerous neighborhood by having a better income is certainly one of them. I believe the point is that it is not as easy to create that compensation than is commonly supposed.

I visited a website several years ago which had quite a collection of data attempting to show that fathers were not as necessary as traditional morality would claim. It did have some very interesting data, and I don't write it off lightly. There is nothing magical about marriage. It may be, as I suggested in the previous comment, that marriage is an indicator as much as a creator of advantage for children.

I commented there only briefly. Things got insulting quite quickly, and it was clear that there was only one answer that people wanted to hear. But at the time, I tried to hammer home one idea: where is the data comparing the outcomes for children of parents married before its conception and still together when the child reached 21, versus an equivalent sample whose only difference was that the parents were not married? That is, parents of same or opposite sex who were together before the child was conceived, and still together when the child reached 21?

We know why that study has not been done, I said. It is not that it would be uncomfortable or unpopular or that no one has thought of it. It is because that second population does not exist in America. There is not a sufficient number of couples fitting that together-for-those-22-years description to measure against the marrieds. They don't exist.

Interestingly, they do exist in greater numbers in Scandinavia, because cohabitating rather than married parents are the norm there. But even in that population, where you would think you would have a better chance of finding a sample size, it is still too small to be representative. Somehow, making it all the way to a child's adulthood doesn't happen quite so often.