August 30, 2014

Children of divorce less depressed as young adults if they grew up in socially connected times

Introductory post here.

The General Social Survey asks a question about how happy you are, and most people say pretty happy. Since that's a fence-sitting answer, I'm going to look at the "very happy" answer (the other is "not too happy"). I've separated the respondents into those who grew up with both parents, and those who did not because of divorce or separation. Since the effects might peter out over time, I'm only looking at people who were 18 to 25 years old when they were surveyed. They're also whites only, to prevent the confound of race differences in family structure.

To measure how connected their world was when they were growing up, I grouped them into birth cohorts of 10 years: 1945-'54, '55-'64, '65-'74, '75-'84, and '85-'94. Each cohort is labeled by the 0-year in the middle, from 1950 to 1990.

The first cohort is early Boomers, and their 1950s childhood was fairly atomized by historical standards, although their adolescence was more outgoing in the '60s. Those in the next cohort, the late Boomers, grew up entirely in outgoing times, although when they were little this trend was only just beginning to rise. The early Gen X cohort also spent their childhood and adolescence in outgoing times, but now at a higher level than for the late Boomers. The late X-ers grew up mostly in outgoing times, up through the '80s, but went through adolescence during the cocooning '90s. (They are the inversion of the early Boomers, who also had a mixed-up upbringing.) The early Millennials grew up entirely during the cocooning period that we're still in.

Did the environment make a difference? Here are the rates of feeling very happy among children from intact vs. broken homes, who were asked in young adulthood, separated by cohort:


Notice that the red line, showing kids from intact homes, is fairly flat across all cohorts. The happiness levels don't seem to change so much, in one direction or another, for children from intact families. Perhaps a healthy home life makes them less dependent on an outgoing climate to feel great, and they will be better buffered against the atomization of the cocooning climate.

The blue line, however, rises along with the outgoing trend and begins to fall with the cocooning trend. With not so much at home to anchor their psychological stability, they are more sensitive to changes in the wider social climate. They never reach the same levels of happiness as those from intact families, but they did manage to close the gap by quite a bit when they had grown up in the more broadly connected environments of the '70s and '80s.

We can treat happiness like a trait that follows a bell-shaped curve, and use the "very happy" response as a cut-off value for that trait. Kind of like using "can dunk a basketball" as a cut-off value for height. We can then work backwards to turn the fraction meeting the cut-off into z-scores, and from there figure out what the distance is between the average person in the two different family structure groups. This is like figuring out how much shorter, on average, one group is compared to another, looking only at how likely they are to dunk a basketball.

Here is a graph of the gap between the happiness levels of the two groups, plotted across each of the cohorts again. The units are standard deviations. For comparison, one S.D. is about 3 inches in the case of height.


Never does the gap reach zero, but it does narrow quite a bit from around 0.4 S.D. to 0.25, then to 0.15, staying around 0.15, and finally rising back to 0.25. If we treat happiness as a kind of "height," then children from broken homes were about 1.2 inches "shorter" than those from intact homes, among the early Boomer cohort. For late Boomers, the gap shortened to 0.75 inch. For both Gen X cohorts, it narrowed down to under half an inch, and then widened back to 0.75 inch among the early Millennials. I'd project that it's even wider among later Millennials, born between 1995 and 2004.

Narrowing a gap in average "height" by three-quarters of an inch in just 20 to 30 years, or one generation, is pretty good — way better than utopian attempts at social engineering. Although by the same token, the steady trend toward cocooning over the past 25 years has probably wiped out those gains. Still, it tells us that, aside from trying to keep down the rate of kids growing up in broken homes, we ought to re-evaluate the costs of cocooning.

By sealing off the nuclear household from the rest of the neighborhood and community, the unfortunate kids who have one or both parents breaking up the family won't have anywhere to turn to as they try to cope with one of the most severe disruptions a person could face, and all while they're still growing up. I know that sounds incredibly emo, but these days, it really must be that devastating.

For a reminder of how manageable and totally not-emo the problem used to be back in the '80s, look at Molly Ringwald's character from Pretty in Pink. Her mother walked out on her and her father, but since it was an anti-helicopter parent era, the father doesn't think twice about her socializing with close friends, going to dances, and bonding with the surrogate maternal figure, who manages the New Wave record store where she works. She's probably not going to turn out exactly like the normal kids at school, but y'know — normal enough.

That was fairly common back then, even for younger children. If your parents were divorced, you spent most of your time outside the home, trying to latch onto something stable and healthy. And provided that you made yourself likable, you were welcome in your friends' homes by your friends' parents.

This underscores the importance of there not being too many broken families, though. If only 10% of the kids are from broken homes, maybe the other 90% can collectively handle them as guests. But what happens when the guests out-number the hosts? Nothing good can come from that situation.

GSS variables: happy, family16, famdif16, cohort, age, race

No comments:

Post a Comment

You MUST enter a nickname with the "Name/URL" option if you're not signed in. We can't follow who is saying what if everyone is "Anonymous."