Children and Divorce

Women are slightly more likely to be the initiators of divorce in couples with children vs. those without children

I recently wrote about how hetero divorces between couples where the wife has high relative status vs. her husband are very likely to have been initiated by the wife¹². Most hetero divorces are initiated by the wife anyways (65%), but this becomes even more likely when she has high relative status, particularly when she makes more money (79%). While there are 3,804 observations of hetero marriages between the two How Couples Meet and Stay Together (HCMST) surveys, there are only 147 divorces, so it’s important to keep in mind that the sample sizes are small. However, I took these results to suggest that breakups of couples where the woman has high relative status are more likely to be driven by the wife feeling she has better options than to be driven by the husband feeling resentful of her success. Of course, who “initiated” doesn’t tell us everything, she might’ve felt staying in the marriage was a bad option as a result of him being resentful. But still, I think the data suggests that the story is more complicated than low status men feeling emasculated by their wives’ success.

I got two comments suggesting other potential narratives for these results that I wanted to take a look at. The first was from

Ruxandra Teslo

who suggested that women who make more than their husbands might make more than most women, and perhaps women who make more in absolute terms would also be more likely to initiate divorce (

Chris Jesu Lee

suggested some other ways that class differences might influence the narrative). Basically, Ruxandra pointed out that my analysis did not separate the effects of having more wealth (and options) in an absolute sense from the effects of having more status in a relative sense. Perhaps even more women want to initiate divorce and they simply don’t because it doesn’t feel financially feasible. The second was from

Peter Gerdes

who suggested that, for couples with children, men may be loathe to initiate divorce, even if they desperately want out of the marriage, for fear of losing access to their children.

As I mentioned already, the number of divorces we’re looking at is small, 147, so properly breaking things out by class is challenging. The dataset includes a variable tracking household income as well as “who makes more” but I wasn’t able to find a variable for how much the respondent makes alone (I could be missing it, the variable coding takes some work to properly parse). However, I was able to compare couples with and without children under 13 and I found that, as Peter expected, women were even more likely to initiate divorce if they had kids. I used “# of children under 13 > 0” to define the kids vs. no kids variable since I expect couples with younger kids are the ones where custody arrangements are most relevant. But the difference with or without kids was very small, 67% with kids vs. 64% without, and women are still the major initiators of divorce in couples without kids.

The excel file used to make this chart is here on tab “kids chart”. The python scripts used to summarize the data are here (2017-2022 data) and here (2009-2015 data).

I also found that women were less likely to initiate divorce in couples that make the same (only 14 observations) vs. couples where the wife makes less (90 observations, see full results in table below³). Again, I wasn’t able to look at the wife’s personal income, but I was able to look at the average household incomes when the wife makes more, the same or less than her husband. I found that household income was lowest on average for couples where the wife makes more and highest on average for couples where the wife makes less, so it’s not clear that women who make more than their husbands make more than the average woman, I just can’t tell from the data I’ve looked at. But the relative status narrative would’ve predicted that women who make the same are more likely to leave than women who make less which I didn’t find. Unfortunately, this dataset is not sufficient to test how the absolute vs. relative status of the wife affects the chance that she initiates the divorce.

The excel file used to make this chart is here on tab “who makes more chart”.

The final observation I wanted to share was how likely the divorce is to be mutual with or without kids. In everything I’ve presented so far I followed the choice made in the paper written by Michael J. Rosenfeld, the main HCMST researcher, to count mutual breakups as 50% driven by the woman and 50% driven by the man. So, the headline number that 65% of hetero divorces are initiated by the woman breaks down as 53.7% initiated by the woman only, 23.1% initiated by the man only and 23.1% mutual. Interestingly, I found that across all the groups I looked at divorces were almost twice as likely to be mutual when the couple had no kids (28%) vs. when they had kids (17%).

The excel file used to make this chart is here on tab “mutual breakup chart”.

I don’t have a strong hypothesis for why that might be… but one idea is that when someone begins to feel unhappy in a couple without kids the other person is more likely to agree that a breakup is the right decision - most people don’t want to stay with someone who doesn’t want to be with them. But, when you add kids in the mix, I could imagine the less unhappy partner being more likely to want the more unhappy partner to make the sacrifice of staying in the relationship for the sake of the kids. Or, maybe it’s just a fluke. But if you have any other suggestions for why we might see this pattern please comment them below!

1. Rosenfeld, Michael J., Reuben J. Thomas, and Maja Falcon. 2018. How Couples Meet and Stay Together, Waves 1, 2, and 3: Public version 3.04, plus wave 4 supplement version 1.02 and wave 5 supplement version 1.0 and wave 6 supplement ver 1.0 [Computer files]. Stanford, CA: Stanford University Libraries.

2. Rosenfeld, Michael J., Reuben J. Thomas, and Sonia Hausen. 2023. How Couples Meet and Stay Together 2017-2020-2022 combined dataset. [Computer files]. Stanford, CA: Stanford University Libraries.

3. Full results in table below: