For almost three decades, population experts have pointed to the US Census data and other federal sources of information about divorce and concluded that the divorce rate peaked in the late 1970s and has been declining since then. This seems to be wrong.
A recent report by Sheela Kennedy and Steven Ruggles from the Minnesota Population Center at the University of Minnesota challenges these common findings. In a 2014 report in Demography, Kennedy and Ruggles write that rather than going down, the divorce rate in the US has been steadily increasing for the past 30 years. This is a startling and important reexamination of recent trends in divorce.
Here is their case. First, the demographers take a careful look at the quality of the data that has been used to calculate divorce rates over the past 30 years. They compare multiple sources of data. In general, population experts have known there have been various problems in the reporting of data, but have generally thought that several more recent methods could be relied on to overcome these problems. Not so says Kennedy and Ruggles. They find that even the better data sources have flaws in the ways that data are collected and may have badly distorted the true divorce rate.
The good news is that in 2008, the US Census Bureau added divorce-related questions to the American Community Survey. This survey annually collects data from a representative group of Americans and is generally the best scientific assessment of the US population. These data gives researchers better information about the divorce rate. The other contribution made by Kennedy and Ruggles is to consider patterns of divorce among different age groups. For many years, the pattern of divorce across age groups indicated that divorce increased in young people until around 25 years of age and then steadily declined among older people. Looking more closely at these data, it becomes apparent that in the more recent decades, the divorce rate has not been declining as rapidly for those over age 35. The result is that people well into their 60s are divorcing at a higher rate than in previous decades. Based on these new data and this careful attention to the changing patterns of divorce among older Americans, the authors conclude, "The age-standardized refined divorce rate increased substantially after 1990 and is now at an all-time high" (Kennedy & Ruggles, 2014).
So what are we to make of this? First, this is an important reassessment of trends in divorce. These findings suggest an important shift in our thinking about marital stability. Divorce rates have been going up and its happening among couples well into their later years. Some of these divorces are among second marriages, but more long-term marriages are also breaking up. Our efforts to help couples strengthen marriages need more attention, not less.
Also, these authors provide an excellent analysis and correct for previous ways of reporting divorce data. Going forward, it is important for other scholars to use their approach and to rely on the data in the American Community Survey.
Another lesson from this work is that "divorce rates" cannot continue to be our primary measure of "marital stability." At least part of the decline in marital breakup among younger people is that more couples, especially young couples, are cohabitating instead of, or prior to, marriage. There is much evidence that these relationships are more unstable than marriages (See Cherlin, 2010). However, these data are not included when we talk about divorce rates. If we want a good estimate of "relationship stability," we need data that combines cohabitation breakups due to divorce. These combined rates would give us a much better picture of the real patterns of stability or instability in couple relationships.
Robert Hughes, Jr., PhD, is a professor of Human Development & Family Studies at the University of Illinois at Urbana-Champaign and writes about the science of divorce at divorcescience.org.