Back in July of 2012 I pointed out that the conventional wisdom on the trend in US divorce rates appeared to be changing. The reason is the official data set gathered by the National Center for Health Statistics has been missing data for California, Georgia, Hawaii, Indiana, Louisiana, and Minnesota. Here is the divorce data the National Marriage Project published in their 2011 State of Our Unions report:
However, the National Center for Family & Marriage Research (NCFMR) at Bowling Green State University used a new data set (American Community Survey) to calculate divorce rates by age in the US for 2010. When they compared this result with their statistic from 1990 they determined that divorce rates had remained flat for 20 years:
The overall U.S. divorce rate has remained essentially unchanged over the past 20 years. In 1990, 19 people divorced for every 1,000 marrieds versus 18 per 1,000 in 2010.
Back in July of 2012 when I emailed the Director of the National Marriage Project about this he indicated that his team was considering using the new ACS data as well. Since then their State of Our Unions report for 2012 has come out. Here is the new divorce trend using ACS data for 2011:
Note that Indiana and Louisiana divorce data is missing from 1990 and 2000 in the chart above, and California data is also missing for 2000. Since the 2011 data point uses the ACS no states are missing. I’m not sure why the two groups show different estimates for the 1990 divorce rate (19 vs 20.9), or why the NCFMR calculated the rate for 2010 at 18 per 1,000 and the National Marriage Project calculated 2011 almost four points higher at 21.7. The ACS data appears to jump around a bit from year to year because when I calculated the overall US divorce rate for 2009 using ACS data I came up with 19 per 1,000 married couples.