From the Mouth of Two or Three Surveys

My post a few days ago looked at whether members of the Church in the US reported a lower likelihood of identifying as “divorced” than non-members in Pew data. 

However, afterwards some friends raised valid concerns about the fact that remarried divorcees would have identified as “remarried.” Therefore, if Latter-day Saints were remarried at a higher rate or were quicker to remarry after a divorce, both very plausible given our emphasis on marriage, that could explain the difference. 

I since discovered that the General Social Survey, a large survey taken almost every year, has a question that asks married or widowed people whether they were ever divorced or separated. Combined with the marital status question, we can use this to create a measure of “ever been divorced.” 

Now, the General Social Survey only has a handful of self-identified members per survey, so you have to combine a lot of years to get a large enough sample of members to say anything interesting, so here I combined all survey years from 2004-2018 (the latest year available); this gives us 220 randomly surveyed members. 

Of those who have ever been married in the survey, 28% of members have been divorced at some point, while 42% of non-members have been divorced. This difference is highly significant, with a less than one in a thousand chance that it happened by chance. Taking into account age and/or year of the survey does not change things. 

Now, this is not saying what the probability that a marriage will end in divorce is. As I noted in my last post, this is technically very difficult. This is saying, when you take a random sample of ever-marrieds in the US, how many report having ever been divorced. 

So, we now have two completely different surveys with different methodologies that support the notion that in the United States members of the Church get divorced less than their non-Latter-day Saint counterparts. Investigating why this is would take me down rabbit holes that I don’t have time to explore right now, but in my opinion it’s now clear that self-identified Latter-day Saints do get divorced less. 

Code is available on my Github page for any wonks who want to play around with different variables and combinations. 


3 comments for “From the Mouth of Two or Three Surveys

  1. I have been married to the same woman for over 50 years, and life is very good.

    I am wondering where you are going with this? Are you going to infer that lower divorce rates indicate more healthy societies for example? My understanding is that 70% of divorces are initiated by women. And that lower rates often indicate women who have less freedom. looking at the countries that have rates under 30 I wouldn’t want to be a woman trying to get out of a bad marriage in most of them.

  2. All these things shall give thee experience, and shall be for thy good. The title of the previous article seems to apply to this one.

Comments are closed.