Divorce and Subjective Well-Being: A Counterfactual Approach

Lance Erickson, Brigham Young University
Kevin Shafer, Brigham Young University

Using the National Survey of Families and Households, we examine the effect of divorce on subjective well-being by applying the counterfactual model of causal inference using propensity score matching (PSM). In accordance with the counterfactual model, we argue that the effect of divorce from traditional methods is biased because the effect of divorce for those who divorce is different than the effect of divorce would be for those who do not divorce. It is likely impossible to know what the effect of divorce would be for those who don’t divorce but using PSM a reliable estimate is achievable for those who do. After matching on an estimate of the propensity to divorce, we find that the effect of divorce on divorcès is negligible. In other words, among those who divorce, divorce neither has a positive or negative effect on subjective well-being.