In Theory: On religion and the risk of divorce

February 15, 2014

Sociologists at the University of Texas and the University of Iowa write in the American Journal of Sociology that "Conservative religious beliefs and the social institutions they create .... increase divorce risk in the contemporary United States."

Others suggest that conservative Southern states have higher rates of divorce than the Northeast because the bluer region has lower rates of marriage. They further suggest that class is a big factor and that poverty contributes to family dysfunction.

Q: Is religion bad for marriage?

While the study controlled for factors such as income and education level, it did not make any distinction about a person's commitment to their faith. Rather, participants self identified their religious affiliation and were identified as "conservative Protestants" if they answered in the affirmative to questions such as: "Do you believe the Bible contains no errors?"


Using this criterion, there was no distinction made between "nominal Protestants" (those who identify themselves with a particular Protestant denomination) and "committed Protestants" (those who attend church regularly and engage in other religious practices such as Bible reading and prayer).

This would seem to be a rather glaring error in methodology given that earlier research indicates significant differences between divorce rates of nominal and committed Protestants. The article also failed to discuss the fact that the study also found that religiously unaffiliated persons had the highest rates of divorce, and counties with the most religiously unaffiliated persons had the highest rates of divorce.

Probably worst of all, the study authors attempt to explain the positive correlation with a series of conjectures about how teaching sexual abstinence before marriage increases early marriage, early parenting and subsequent divorce. These causative assumptions were not addressed by this study. As any psychology student completing their first course in statistics can tell you, correlations only indicate that two factors vary in the same direction, never that one factor causes another.

Pastor Ché Ahn
HRock Church


Burbank Leader Articles Burbank Leader Articles