This study examined whether married Americans who view pornography, either at all or in greater frequencies, have a higher likelihood of experiencing a marital separation over time compared to those who do so seldom or not at all. Findings affirmed that married pornography users were more than twice as likely to experience a marital separation in the 6 years following their reported pornography viewing compared to those who did not view pornography. Looking at the association between pornography viewing frequency and marital separation demonstrated that the relationship was technically curvilinear, with marital separation by 2012 increasing with 2006 pornography use to a point and then decreasing at frequencies of at least weekly. Ancillary analyses, however, showed that this group of Americans with high initial pornography viewing frequency and low likelihood of marital separation was a small minority in the sample, too small in fact to be statistically distinguishable from either abstainers or moderate pornography users. Tests for interactions with gender also showed that the associations between pornography use and marital separation held regardless of whether the participant was male or female.

Before discussing the implications of this research, several data limitations should be mentioned in order to chart a path for future studies. First, the data were unable to clarify what exactly led to the marital separation between 2006 and 2012, or, if pornography use was involved, how it contributed to the split. While this study has gone beyond previous work to establish with longitudinal data that pornography use at one time does predict later marital separation, even after controlling for relevant confounders, more detail would be needed to clarify exactly how pornography use might be involved. A second limitation is that PALS does not indicate what type of pornography was being used or whether participants were viewing pornography with their partners. Previous research suggests that certain types of pornography might be related to committed romantic relationships in different ways (Willoughby & Busby, 2016), and there is a growing literature suggesting that a central issue is whether partners or spouses are viewing pornography as a couple or in isolation (Newstrom & Harris, 2016; Willoughby et al., 2016). Future research would ideally make use of longitudinal dyadic data where both spouses’ pornography viewing can be taken into account, and, preferably, with more precise measures of pornography use so as to clarify what sort of content is being consumed.

A third limitation is that, while these data were longitudinal and thus allow for the establishment of temporal precedence between pornography use (Time 1) and marital separation (Time 2), they cannot rule out the possibility that there are self-selection issues that may account for both the pornography use at Time 1 and marital separation later on. While the analysis controlled for Time 1 marital happiness, sexual satisfaction, and previous divorce, other confounding factors could include whether participants had divorced parents or other family members who divorced (Amato, 2010) and personality characteristics which are related to both pornography use (Egan & Parmar, 2013) and divorce (Roberts, Kuncel, Shiner, Caspi, & Goldberg, 2007). Another intervening factor could include the subjective evaluations and experiences of participants’ spouses (Bridges et al., 2003; Stewart. & Szymanski, 2012; Zitzman & Butler, 2009). Again, dyadic data would better isolate other factors that might have contributed to the experience of marital separation, beyond more specific measures of pornography use. And finally, the analysis necessarily focused on heterosexual couples since there would have been so few same-sex married couples in 2006 when Wave 1 of PALS was collected. Because pornography is likely associated with the relationship outcomes of same-sex couples somewhat differently than heterosexual couples, future studies would ideally use large enough samples of same-sex couples along with heterosexual couples to discern any differences.

Despite these limitations, this study has contributed to the literature on pornography use and relationship outcomes in several key ways. First, while numerous studies have found a negative association between pornography use and marital quality (Doran & Price, 2014; Muusses et al., 2015; Perry, 2016, 2017a; Stack et al., 2004; Yucel & Gassanov, 2010), the current study has affirmed recent research (Daines & Shumway, 2012; Perry & Davis, 2017; Perry & Schleifer, 2017), suggesting that the potential influence of pornography use may extend to the stability of marital relationships. Married pornography users, in general, were more likely to experience a marital separation over time compared to those who abstain from pornography. But why? Because of the nature of the dependent variable, it is impossible to tell who initiated the separation (the primary consumer or their spouse) or how pornography use played a part. Following previous research, it could be that the sexual scripts internalized from consistent pornography use contributed to marital dissatisfaction for the consumer and/or that some sort of discrepancy between the pornography consumption of the consumer and her/his spouse contributed to feelings of insecurity or betrayal. The pornography use measure was also a limitation in this regard since the association between Time 1 pornography use and Time 2 marital separation may be due to certain types of pornography being viewed and/or whether it was being viewed in private rather than as a couple (Willoughby & Busby, 2016; Willoughby et al., 2016).

Part of the explanation may be found in the apparent curvilinear relationship between pornography viewing frequency and the likelihood of marital separation. It is possible that married persons who were viewing pornography at the highest amounts in 2006 were those who had either discussed their pornography use with their partner or were, in fact, viewing it regularly with their partner. This would help explain why their pornography use did not seem to be associated with higher rates of marital separation, namely because it was not done in isolation. Alternatively, because the data only show pornography use in 2006, it could also be that married participants who were viewing pornography at the highest frequencies in 2006 at some time stopped using pornography and their long-term marital stability improved as a result. It should also be kept in mind that the group of participants who both viewed pornography at the highest frequencies and 2006 and experienced relatively high marital stability up to 2012 were a small minority of the sample (only 5%). The majority of the sample, rather, fit the predicted pattern—those who abstained from pornography experienced greater marital stability over time, while those who viewed pornography in increasing amounts (up to two or three times a month) experienced an increasing likelihood of marital separation in the following 6 years. Ultimately, longitudinal and dyadic qualitative interviews would be ideal to unpack the specific role that pornography use may have played.

Another unanswered puzzle is why the association between pornography use in 2006 and later marital separation was not moderated by gender, especially when previous empirical research and leading theories suggest that men’s relationships are more strongly associated with pornography use than women’s (Doran & Price, 2014; Perry, 2017a; Perry & Davis, 2017; Wright et al., 2017; Yucel & Gassanov, 2010). Perry and Schleifer’s (2017) study of pornography use and divorce using GSS panel data also found no moderating effect of gender. It may be that, while men’s marital quality may be more sensitive to pornography use compared to women given men’s different use patterns and potential susceptibility to dominant sexual scripts in pornography (Sun et al., 2016; Wright et al., 2014), viewing pornography may relate to their marital stability in a qualitatively different way. Greater insights could be gleaned from data that take into consideration, not only specific measures of pornography consumption, but how men and women interpret their partner’s pornography use (Willoughby et al., 2016), and how that relates to marital stability over time.

As technology and cultural norms continue to evolve, it is important to consider how the observed connection between pornography use and marital separation or divorce may change over time. Perry and Schleifer (2017) find that the link between beginning pornography use and divorce declines with age, which they theorize may be due both to stronger bonds among older couples and greater frequency of pornography use among younger Americans. Price et al. (2016) show that Americans from the 1980s onward (the first to experience the Internet as adolescents) show the most dramatic increase from previous cohorts in their pornography use. Moreover, more recent cohorts are more likely than previous generations to continue watching pornography as they get older. To the extent that pornography use contributes to marital separation, as more Americans regularly view pornography at younger ages and sustain that viewership into their later years, it is possible that greater instances of marital disruption in the U.S. could be a consequence. Indeed, because the average participant age for PALS was 47 years, the associations between pornography viewing and marital separation in this study may actually be suppressed by the older sample (Perry & Schleifer, 2017). However, it could also be possible that pornography’s overall negative association with marital stability will attenuate with time. As pornography use continues to become more mainstream and commonplace, couples might have more open dialogue about its use and it would not be a source of conflict. More longitudinal data are needed to track how the prevalence of pornography use connects to relationship dynamics over longer periods of time.