Executive Summary: “Growing Evidence for a ‘Divorce Divide’? Education, Race, and Marital Dissolution Rates in the U.S. since the 1970s” by Steve P. Martin

In this study, I use the Survey of Income and Program Participation (SIPP) and supplementary data to measure trends in marital dissolution rates across educational and racial groups in the United States.  The outcome of interest is a marital dissolution within ten years of a first marriage, where a marital dissolution can be either a divorce or a permanent separation.  I compare marital dissolution rates for women of different education levels, and I also compare trends based on women’s years of marriage.

The results show an increasing gap in marital dissolution rates, as marital dissolution rates have declined substantially for college graduates, but held steady or even risen for women of lower education.  A few percentages summarize these findings. Firstly, among women with a bachelor’s degree or higher who married in the 1970s, about 25 percent had a marital dissolution within ten years.  For women with a bachelor’s degree or higher who married in the early 1990s, about 17 percent had a marital dissolution within ten years.  Similarly, among women with no bachelor’s degree (including some college, high school graduate, or no high school diploma) or higher who married in the 1970s, about 34 percent had a marital dissolution within ten years.  Of women without a bachelor’s degree or higher who married in the early 1970s, about 35 percent had a marital dissolution within ten years.

Supplementary analyses yield several additional findings.  When examining other data sources, I find a pattern of a widening gap in marital dissolution rates in all of them. I’ve compared results across a number of educational categories, and found consistent evidence for a widening gap.  In particular, marital dissolution rates appear to be increasing for women with no high school diploma.  My study also compares results for men to results for women.  In the SIPP data, the gap is growing most rapidly for women, but in census data, it appears that the gap may be growing more rapidly for men.  Men’s responses may not be as reliable as women’s, so gender differences might simply reflect response error.  I’ve briefly analyzed trends in marital dissolution extending back to 1960.  Highly educated women have always had lower marital dissolution rates than other women, but the diverging trend appears to have started in the mid-1970s.  In addition, the growing gap in marital dissolution rates is not explained by recent increases in women’s overall educational attainment, nor by statistical controls for recent increases in age at marriage timing and premarital childbearing.  Finally, I examine the gap for different groups of married women with children.  If anything, the divergence in marital dissolution rates is especially pronounced among married couples with young children.

Together, these results suggest a growing association between socioeconomic disadvantage and marital dissolution.  However, it is premature to speculate that socioeconomic disadvantage is the cause of the growing gap in marital dissolution rates, because many other factors have affected marriage in recent decades.