Premarital sex and the risk of divorce.
Data from the 1988 US National Survey of Family Growth (NSFG) were utilized to assess the impact of premarital sexual activity on subsequent marital stability. Among white NSFG subjects first married in 1965-85, virgin brides were significantly less to have become separated or divorced (25%) than women who had not been virgins at marriage (35%). A bivariate profit model was applied to examine 3 possible explanations for this phenomenon: 1) a direct causal effect; 2) an indirect effect through other intervening variables that are positively related to both premarital sex and divorce (e.g. early marriage or a premarital birth); and 3) a selectivity effect, in which virgins differ from nonvirgins in terms of their family backgrounds or attitudes toward the institution of marriage. The lower risk of divorce on the part of white women with no premarital sexual experience persisted even after numerous intervening and background variables were controlled. However, when unobserved factors affecting both the likelihood of being a virgin and the likelihood of marital disruption were taken into account, the impact of virginity status on the risk of divorce lost its significance. The effects of other predictors of divorce (age at marriage, family background, religion) remained unchanged from single-equation estimates. It is speculated that the unobserved variables that selectively protect virgins from the risk of divorce reflect a set of traditional attitudes about marriage. For example, women who adhere to traditional values and expectations early in life may be programmed to postpone sexual activity until after marriage and to reject divorce as an option if they are unhappy or dissatisfied. Studies that explore long-standing attitudes toward the institution of marriage are urged to confirm or refute this hypothesis.