May 26, 2015

Broken homes more likely for children of transplants

An earlier post showed that transplants are less connected with their extended blood relations than natives. Does that hold even for their closer relations in the nuclear family? Let’s look now at whether transplants are more likely to be raising their children in a broken home, i.e. without both birth parents.

The General Social Survey allows us to study transplants in a purer form than simply “raised in the ‘burbs, moved to the nearest city”. The GSS only asks for the respondent’s Census region -- New England, Pacific, West North Central, etc. “Transplant” here implies a greater degree of deracination.

I’ve limited to focus to whites since family structures vary wildly among whites, blacks, Asians, and Hispanics. I’ve also made the comparisons between natives and transplants within three separate class levels, based on number of years of education (0-12, 13-16, and 17-20).

Looking at respondents with children, the marital status of regional transplants is identical to natives.

1. Percent married among those with kids

Class: % natives ___ % transplants

Lower: 75 ___ 75

Middle: 78 ___ 79

Upper: 85 ___ 85

At first glance, then, children appear equally likely to grow up with a set of married parents, regardless of their parents being transplants or not.

However, the GSS also asks respondents who say they’re married, if they’ve ever been divorced. Now the differences show up.

2. Percent divorced among those married with kids

Class: % natives ___ % transplants

Lower: 20 ___ 26

Middle: 19 ___ 25

Upper: 16 ___ 17

Combining these two tables into one, we see that transplants are more likely to be married with children yet previously divorced (although not for the upper class).

3. Percent married, never divorced among those with kids

Class: % natives ___ % transplants

Lower: 60 ___ 56

Middle: 63 ___ 59

Upper: 71 ___ 71

So despite the initial impression of children of transplants growing up with married parents, it turns out that their household is more likely to include a step-parent than the households where parents are natives.

Were the transplants themselves more likely to have grown up in a broken home, and perhaps they’re just passing along a genetic predisposition? Somewhat, but not entirely. The next table shows the likelihood of having grown up in a broken home for natives vs. transplants, where the classes are now based on the education level of the respondent's father.

4. Percent growing up in broken home

Class: % natives ___ % transplants

Lower: 12 ___ 14

Middle: 11 ___ 13

Upper:  9  ___ 13

Transplants are more likely to have grown up in a broken home, but the differences are only half as big as in the previous table. Partly, the transplants are passing on a tendency toward raising children in broken homes that would have happened whether they stayed put in their region or not. But just as much of their kids growing up in broken homes is an effect of the parents being transplants.

Transplants with advanced degrees (17-20 years of education) are an exception here, as they were more likely to grow up in a broken home yet are just as likely to be raising their own kids in an intact home.

Overall, though, having moved to a different region than the one you grew up in increases the risk of your children growing up in a broken home. Thus the destabilizing effects of migration on the bonds of kinship are not limited only to the more distant, extended family ties, but even to those between parents and children, albeit to a lesser degree than the damage done to extended family ties.

GSS variables: family16, marital, divorce, childs, regtrans (region, reg16), race, paeduc, educ


  1. Transplants beget transplants.

  2. Makes sense, without a baked-in support network of family and old friends, the normal stresses of marriage and parenthood are amplified.

  3. I have a hard time believing that 2% points are going to be statistically significant.

    Though 4-6% ones probably are.

  4. In huge samples, just about any difference is significant. How "meaningful" they are is a separate question.

    But given that we're talking about stewardship over your own flesh-and-blood children, rather than distant relatives, even a minor difference is damning evidence of footloose migrants, in my eyes.

  5. One aspect of this I remember is that transplants under the regtransplant variable are older, with average age 43.45 for Native vs 47.44 for Transplants

    Being older would give you more time to be divorced and then remarry in.

    Plus if you're a bit older, your kids are a bit less likely to be part of a broken home exactly if you do divorce.

    Average ages by marital and transplant status

    0: Native 46.10 68.70 46.03 41.79 28.94
    1: Transplant 48.11 69.23 49.81 45.91 33.01

    I had a go at testing this by comparing the differences at different age groups using an age group control (no education control on top of that though), comparing the differences at each year of age (for the ever divorced variable, the children variable at at least one and regtrans).

    Across the whole age group set of all given year, transplants averaged 6% higher more likely to have been divorced. At a finer scale, at 20-29, they were 3.5% less likely to have been divorced, at 30-40 the same, then differences began emerging at 41-50, where transplants were 3.5% more likely to have divorced and then 51-65 where they were 10% more likely.

    That might be pushing the data further than it goes, but that would fit with a certain view of transplants where once their kids are adults, or at least teens, they're more likely to consider the marriage as having outlived its usefulness and function, despite the fact that with more willpower and drive they would in the immediate term be likely to stay together for the kids.

    While the stay at home Natives are more likely to stay together long term for continued community support as a family, sentimental attachment, family holidays, etc. Perhaps for the same psychological reasons more or less they stay rooted to a region.


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."