Predicting divorce
Predicting divorce involves using statistical analysis and mathematical models to assess the likelihood of marital dissolution. Contrary to the common belief that about 50% of marriages end in divorce, research suggests that the actual figure is closer to 40% for first marriages. This discrepancy arises because the statistics often compare newlyweds to those divorcing in the same year, without accounting for the average eight-year marriage duration prior to divorce. Various risk factors contribute to divorce probability, including pre-marital influences such as parental divorce, educational attainment, and the age at marriage, as well as in-marriage dynamics like communication styles, financial issues, and commitment levels.
Two primary methods are utilized to predict divorce: empirical approaches that analyze data from couples, and mathematical models that identify current relational features that could indicate future risks. Notably, research by Professor John Gottman highlights how communication styles and certain negative interaction patterns can significantly predict divorce, achieving a high accuracy rate in assessments. Understanding these factors can provide valuable insights for couples and professionals seeking to strengthen marriages or anticipate challenges.
On this Page
Subject Terms
Predicting divorce
SUMMARY: Statistical data analysis and mathematical models can be used to predict the likelihood of divorce.
There is a common misconception that one out of every two marriages ends in divorce. The 50% number comes from dividing the number of divorces in a given year (about 1.3 million) by the number of marriages in that same year (about 2.6 million). The mistake is failing to realize that, in any given year, the people getting divorced are probably not the same as those getting married, because the average length of a marriage before a divorce is about eight years (the overall length of marriage, on average, is about 24 years). Hence, those getting married in any given year have an eight-year lag in their projections for divorce. This lag means that the numerator and denominator of the above ratio are not comparable. Instead, experts suggest that about two out of every five marriages end in divorce (or about 40%).
Because of the propensity for some to remain married, for some to divorce more than once, and for some to never marry, only about one out of every five people are predicted to experience a divorce in their lifetime. However, these figures mask the distribution of divorce rates by category—40% of all first marriages end in divorce, 60% of second marriages end in divorce, and 73% of all third marriages end in divorce. There are also some differences by age group, with divorce rates highest for those in their early 20s and declines steadily in subsequent age groups.
There are two main ways to predict divorce: empirical (or statistical) methods that take advantage of data gathered on married and divorced couples; and mathematical models that try to make a priori predictions of future divorce using features of existing marriages or theoretical assumptions based on extensive work in the area.
Empirical Methodology
Empirical work suggests that indicators predicting divorce can be separated into two groups: factors present before marriage and factors that occur within the marriage. Some of the more common risk factors brought into a marriage include parental history of divorce (children of divorced parents are more likely to divorce), educational attainment (those with lower levels of education are more likely to divorce), and age (those who marry younger are more likely to get divorced). The risk factors that arise within the marriage include communication styles (couples with poor or destructive communication have a greater chance of divorce), finances (couples with financial problems, including a large disparity in spending habits, disposable income, and wealth goals, are at a greater risk for divorce), infidelity, commitment to the marriage (a lack of commitment or a dissimilarity in the amount of commitment often leads to divorce), and dramatic change in life events.
Mathematical Models
Mathematical models seek to discover features of current relationships that will put a couple at risk for future divorce. Professor John Gottman argues that the way couples communicate can often predict divorce. His research, which is based on analyzing hundreds of videotaped conversations between married couples, claims a 94% accuracy rate. The work also monitors pulse rates and other physiological data that, when combined with the observations, leads to what he calls the “bitterness rating.” The rating is based on six signs. The first sign posits that when a conversation starts with accusations, criticisms, or negativity, the discussion is likely to end badly. However, he argues that the opposite is also true. The second sign encompasses four patterns of negative interaction that can be deleterious to a marriage: criticism, contempt, defensiveness, and stonewalling. The third sign is “flooding,” in which negativity of one partner overwhelms the positive feelings of the spouse until there is virtually nothing left but discontent. The fourth sign recognizes that physiological changes, such as increases in adrenaline and blood pressure, often lead to feelings of entrapment and serve to poison an otherwise benign conversation. The fifth sign identifies the fact that some marital discord is unchanged by the repeated attempt by one partner to repair the damage done to the relationship. Finally, the sixth sign involves one or both people rewriting the history of their relationship to be largely negative. Once people reach the sixth sign, Gottman argues, divorce is likely.
Bibliography
Booth, Alan, and John N. Edwards. “Age at Marriage and Marital Instability.” Journal of Marriage and the Family 47 (1985).
Gottman, John, and Nan Silver. The Seven Principles for Making Marriage Work. London: Orion, 2004.
Martin, Teresa Castro, and Larry L. Bumpass. “Recent Trends in Marital Disruption.” Demography 26 (1989).
South, Scott, and Glenna Spitze. “Determinants of Divorce Over the Marital Life Course.” American Sociological Review 51 (1986).
Wolfinger, Nicholas H. “Trends in the Intergenerational Transmission of Divorce.” Demography 36 (1999).