Feminists have been telling us for decades that consumption of pornography trains men to treat women as sex objects and teaches them that it’s acceptable for women to be subjugated to their male partners. Some critics have claimed that it encourages rape and violence toward women, although there is no convincing evidence that nonviolent pornography causes physical aggression toward women. Studies of pornography’s effects on attitudes toward women’s equality have reported conflicting results.

A new study by Taylor Kohut and colleagues at Western University in London, Ontario seems to be more definitive than previous studies. The data come from the General Social Survey (GSS), a large-scale interview survey of American public opinion conducted by the National Science Foundation. Participants were 10,946 men and 14,101 women interviewed between 1975 and 2010. Pornography consumption was measured by a question asking participants whether they had seen an X-rated film in the past year. (Twenty-three percent said they had.) Attitudes toward women were measured using five questions or sets of questions.

• Feminist identification. “Do you think of yourself as a feminist or not?”
• Women holding positions of power. Three questions measuring attitude toward women holding political power, i.e., “If your party nominated a woman for president, would you vote for her if she were qualified for the job?”
• Working outside the home. Five questions, i.e., “All in all, family life suffers when the woman has a full-time job.”
• Abortion. Three questions measuring whether a legal abortion should be available in cases of rape, pregnancy of an unmarried woman, or an unwanted child for any reason.
• Traditional family. Ten items measuring support for a family in which “women take the main responsibility for care of the home and children, while men take the main responsibility for supporting the family financially.”

There are good reasons to believe that this is a strong study. It has a large sample that is representative of the American adult population. In addition, these questions are embedded in a much longer survey, making it unlikely that participants were sensitized to the purpose of the analysis.

The data were analyzed by separately comparing men and women who had or had not watched pornography. Pornography consumption had no effect on self-identification as a feminist, or on attitudes toward the traditional family. However, contrary to predictions from feminist theory, both men and women who had viewed pornography in the past year had more positive attitudes toward women holding positions of power, were more supportive of women working outside the home, and had less negative attitudes toward abortion. While these differences were not large, they were statistically significant due to the large sample size.

The data fail to support the hypothesis that pornography encourages subordination of women. However, the authors carefully avoid claiming that the results show that pornography encourages liberal or feminist attitudes toward women’s equality. This is a correlational study, and correlation does not imply causation. A positive correlation between pornography consumption and liberal sexual attitudes could mean that watching pornography causes feminist attitudes (implausible), or that people with liberal sexual attitudes are more likely to watch pornography (much more likely), or a that some third variable, such as a non-religious upbringing, causes both.

A recent experiment done in Denmark by Gert Martin Hald and colleagues appears to contradict the Kohut study. The Danish study had two parts—a survey and an experiment. The participants, 200 young adult Danes, were asked about their prior pornography consumption and given scales measuring attitudes toward women and hostile sexism. Among the men (but not the women), the greater the pornography consumption, the less egalitarian their attitudes toward women were and higher they were in hostile sexism.

The second part was an experiment. Participants were randomly assigned to watch either 30 minutes of nonviolent pornography or a control film. Attitudes were measured after the film, and it was found that, for both men and women, pornography watching led to greater hostile sexism. However, this only occurred for those participants who were low in the personality trait of agreeableness, described as a tendency to be suspicious and antagonistic toward others. Highly agreeable (friendly and cooperative) people did not show any effect.

Are these results inconsistent with the Kohut study? Maybe, but maybe they can be reconciled. The American and Danish surveys found opposite results.  But as Kohut points out, the participants in the Hald study differed from theirs in important ways. They were Danes rather than Americans. They were young adults with an average age of 25, while the American sample had an average age of 45. Most importantly, they had agreed to watch a pornographic film. Only about 70% of those contacted agreed to participate. Thus, the Danish sample is younger and presumably more liberal in their sexual attitudes. To put it differently, the more inclusive American sample includes older people who are more likely to have conservative sexual attitudes and probably don’t watch pornography.

What about the results of the Hald experiment? Suppose it’s true, as Kohut found, that in a representative sample of people of all ages, those with more liberal sexual attitudes are more likely to watch pornography. Suppose it’s also true, as Hald found, that pornography’s content undermines those egalitarian attitudes, at least among some audience members, making them more sexist than they would have otherwise been. Nevertheless, they might still be more liberal than those who don’t watch pornography at all.

This is admittedly highly speculative. Disentangling these possibilities will require a longitudinal study in which both pornography consumption and sexual attitudes are measured among a representative sample of adults over a period of several years.