Now that the data from the 2014 General Social Survey are online, we can look into some other trends that were too hazy to study before. The sample size of Millennials used to be too small, but now with another wave of the survey including them, they can be better investigated.
One thing I've been wondering about for awhile is if the children of divorce are sticking together when they themselves became parents, or if they're just going to perpetuate a climate of broken homes.
In an earlier post, I looked at the trend by birth cohort. Growing up without both parents became more common starting with people born in the late 1950s, and only grew more and more prevalent with each cohort afterward, right up through those born in the late 1980s. The sample size was too small for '90s births to see whether it continued or reversed.
Now that there is a large enough sample size, though, it looks like the trend did reverse. I grouped respondents into five-year age cohorts, and no matter how you move that five-year window around, those born in the early-mid-'90s were more likely be living with both parents at age 16. The difference is only a few percentage points, but that's still remarkable considering that every previous cohort since the late Boomers showed a steady and notable decline of growing up in an intact family.
These results showed through for whites, blacks, and "other" races. Race could not have anything to do with the overall results anyway, since more recent cohorts are blacker and Mexican-er, and those groups have higher rates of broken homes than whites do. Simple demographic projections would have predicted a steady decline, but it looks like the return toward the bi-parental household succeeded in spite of racial demographic trends against it. It is therefore like the falling rates of violent and property crimes over the past 20-25 years, despite a blacker and browner population.
The reversal also held for upper, middle, and lower classes (I used number of years of education as a proxy), although it was stronger and came somewhat earlier for those higher up on the social pyramid. This is unlike the pattern that Charles Murray details in Coming Apart, where for example the lower class continues to get divorced at higher rates over time, while the upper class has returned to marital normalcy after the initial perturbation back in the '70s.
This makes me doubt the earlier explanation I gave that linked broken homes to the status-striving and inequality cycle -- things that have steadily gotten worse since sometime in the '70s or early '80s.
If the first cohort to be struck by the broken homes epidemic was born in the late '50s, and the direction began to reverse with the cohort born in the early '90s, that suggests a link to the cocooning-and-crime cycle. Being a child of divorce became more common among those who were small children during the rising-crime period of circa 1960 to 1990. If you were a little kid during a falling-crime period -- most Silents and Boomers, who grew up in the Midcentury, and later the '90s babies -- you were more and more likely to grow up in an intact family.
One effect of a rising-crime climate is giving less weight to the future and living more in the now -- not surprising when rising crime rates make a safe and secure future look less and less likely. This is a "facultative" response, one that responds to current conditions. If those conditions are present long enough over time, people will evolve an "obligate" response: their discounting of the future becomes more wired-in where the environment is violently unstable.
So perhaps the parents splitting up and telling their kids good luck was part of the greater pattern of impulsiveness or discounting of the future. Abortion rates took off until circa 1990 as well -- hard to think of a more callous attitude toward your child's future than that.
Once the crime rate started falling in the '90s, parents projected a safer and more secure future, and began weighing the future more heavily. As one sign, they became less likely to opt for abortion or divorce as a solution to the "don't feel like raising kids" problem.
GSS variables: family16, cohort, educ, race