Volume 32, Number 3, May/June 2000
Odds of Spousal Infidelity Are Influenced by Social And Demographic Factors
Some 11% of adults who have ever been married or cohabited have been unfaithful to their partner, according to the results of a national survey. In the first analysis of its kind to examine a constellation of factors that have not previously been studied simultaneously, the investigators assessed the influence of adults' opportunities for meeting additional sex partners, sexual values and tastes, current relationship and demographic characteristics on the odds of being unfaithful.1 When these factors were analyzed together, thinking about sex several times a day, having had a high number of prior sex partners, living in a central city, being male or black, and having been part of a couple for a long time were all associated with an increased risk of infidelity. Disapproving of sexual infidelity and sharing social networks with one's partner were associated with reduced odds of having been unfaithful.
The analysis included 2,598 men and women aged 18-59 who had participated in the 1992 National Health and Social Life Survey and who had ever been married or lived with a sexual partner. In face-to-face interviews, respondents answered questions about their social background; health; fertility; and sexual activities, attitudes and fantasies. They then filled out self-administered questionnaires indicating whether they had ever had extramarital sex. Following the survey, interviewers gathered more information about the timing of respondents' marital and cohabiting unions and their sexual histories, including whether they had had sex with someone other than their primary partner in the past 12 months.
Overall, 11% of respondents reported infidelity. Among the 1,717 adults who had been married only once, 16% reported having engaged in extramarital sex; of the 2,010 respondents who had been in a cohabiting or marital relationship in the 12 months prior to the survey, 5% reported unfaithful behavior during that period.
Most respondents expected their partners to be sexually exclusive and believed that their partners expected the same of them. Cohabiting respondents were only slightly less likely to expect their partners to be exclusive or to believe their partners expected exclusivity (94-95%) than were married respondents (98-99%, depending on whether they had lived with their spouse before marrying). Cohabiting individuals and those who married without cohabiting were about equally likely to actually be exclusive (88% vs. 92%).
Aware that concerns about social acceptability may influence respondents' reporting of sensitive issues such as extramarital sex, the researchers conducted a preliminary set of analyses on a subgroup of participants in which they compared results from the self-administered questionnaire with results from the face-to-face interviews. The data proved to be generally comparable, so the investigators used information from the interviews to perform a series of logistic regression analyses testing the effects of several types of factors on the likelihood of unfaithful behavior among respondents who had ever been married or in a cohabiting relationship.
When demographic and control variables alone were examined, being male or black were highly associated with the risk of infidelity (odds ratios, 2.2 and 1.8, respectively). In addition, for each year couples lived together, respondents became 1% more likely to be unfaithful (odds ratio, 1.1), and respondents whom the interviewers judged to be frank in their replies had an elevated risk of reporting infidelity (1.4). When the variables related to sexual tastes and attitudes were added to the analysis, respondents who thought about sex often had increased odds of being unfaithful (1.3), while those who did not approve of extramarital sex had reduced odds of experiencing infidelity (0.5). Men's likelihood of being unfaithful declined but remained statistically significant (1.6); results for the other background variables remained essentially unchanged.
In the next step of the analysis, the researchers added factors measuring characteristics of respondents' relationships. In these calculations, men and women who had cohabited were 40% more likely than others to have been unfaithful (1.4). Partners' religious, educational and age differences had no significant impact on the odds of infidelity. The background and attitudinal factors that were significant previously maintained their effects.
The last set of estimates included factors measuring respondents' opportunities for meeting additional sex partners. The more sex partners a respondent had had between age 18 and the time of first marriage or cohabitation, the more likely he or she was to be unfaithful (1.01). Men and women who lived in a central city had elevated odds of infidelity (1.5), while those who shared social networks with their partner had decreased odds of being unfaithful (0.7). The likelihood of infidelity was not significantly influenced by sexual opportunities in the workplace or by attendance at religious services. In this analysis, which encompassed all variables, black respondents, males and individuals with a high level of interest in sex remained at increased risk of being unfaithful, while those with strict attitudes about extramarital sex continued to have reduced odds of infidelity.
In a separate set of analyses, the researchers examined the behavior of married and cohabiting adults over the 12 months prior to the study. When they tested the same array of variables used in the calculations for all adults, the results were generally similar. In an additional analysis including two new variables, respondents in a cohabiting relationship and those who demonstrated a high dissatisfaction with their current relationship were more likely than others to be unfaithful (odds ratios, 2.1 and 1.3, respectively).
Commenting on the study's findings, one of the researchers observes that sexual behavior is social behavior, and suggests that interventions aimed at reducing risky sexual behavior should take into account the social contexts in which individuals make decisions related to sexual partnerships, as well as demographic risk factors.2--I. Olenick