PDF of full study – Reconsidering the Pornography Use and Abortion Support Relationship – A Reply to Tokunaga, Wright, and McKinley (2015)
Tokunaga, Wright, and McKinley (2015) argue that pornography usage significantly influences one’s later support of abortion. To support their position, hey relied on panel data from the General Social Survey (GSS) from 2006 through 2010 and regressed abortion support onto self-reported pornography use from 2 years’ prior to measurement. In a later analysis, Wright and Tokunaga (2018) claimed this relationship was better explained by a broader attitude structure called sexual liberalism which they argued is influenced by the acquisition, activation and application of sexual scripts embedded in pornographic content. The current study reexamines this claim by analyzing sexual liberalism factors, such as attitudes towards gay marriage, extramarital sex and political and religious beliefs from the most recent GSS panel data and their 2016 survey. Results indicate these factors are collectively stronger predictors of support of abortion than pornography usage alone. An examination of a three wave panel dataset (2010, 2012, and 2014) demonstrated a lack of time-order relationship between pornography usage and support for abortion. It is argued that sexual liberalism provides a better explanation for the previously found relationship between abortion support and pornography usage. Specifically, both abortion support and pornography use are two of many indicators of sexual liberalism, a higher-order attitude construct. Suggestions are presented to further test the relationship between sexual liberalism and pornography usage.
URI – http://hdl.handle.net/10477/80033
EXCERPTS:
The current paper revisits the potential effects of pornography use proposed in a recent pair of analyses (Tokunaga, Wright & McKinley, 2015; Wright & Tokunaga, 2018) in relation to abortion rights attitudes. The authors argue that exposure to pornography’s sexual scripts leads to liberal perceptions of sex and reproduction and liberal sexual attitudes are positively associated with an increased support for abortion. The relationship between pornography use and support for abortion has received no attention outside the work of these two papers.
It is the position of this thesis that the arguments from these earlier papers are based on invalid conclusions drawn from panel data that analyzed the relationship between support for abortion (measured at later data point) and self-reported pornography use (measured at an earlier data point). To this end, the internal validity concerns of past work casts doubt on the legitimacy of the claim that pornographic media use predicts abortion support. To be clear, it stands to reason the two factors are logically related to one another; however it is my position that the weight of evidence indicating pornographic media use causes one to be pro-choice is light.
It is argued the relationship between pornography exposure and abortion support is better explained by a liberalism model, where pornography use, abortion attitudes, political identity, and other factors indicate an individual’s overall sexual liberalism.
In a 2012 Pew Research survey, 80% of liberal Democrats and 31% of conservative Republicans surveyed believed abortion should be legal. Survey participants from this same analysis who reported they seldom or never attend religious events were twice as likely to support abortion compared to those who attend religious services once a week or more. Participants that graduated college were 30% more likely to support abortion than participants that had a high school diploma or less (Pew Research Center, 2012).
Among the 816 participants who reported their position on abortion at all three waves, 415 of those individuals were also surveyed on pornography consumption in the same three Surveys. Amongst those whose abortion position differed at T1 and T3, only 24 participants (5.8%) reported pornography usage at T1, 19 at T2 (6.0%) and 26 at T3 (6.3%). It seems very difficult to draw pornography usage as a cause for change in attitudes towards abortion due to the paucity of participants who reported pornography usage and who also changed their stance to be in favor of abortion.
It should be noted that the GSS measure used by Tokunaga & Wright simply asked if participants had seen an X-rated movie in the last year. This measure puts the frequent porn viewer and the occasional porn consumer in the same response category. What is more, the nature of the question is confusing as it specifically asks about X-rated movies and does not address other forms of pornography such as brief clips or other forms of sexual media one typically streams. The use of interval or continuous measures of pornography would better measure any possible effect of watching pornography.
The usage of dichotomous measure over continuous interval measures can result in a 20% to 66% loss in of the variance that may be accounted for on the original variables (Cohen, 1983).
Religious Views: A scale from 1 (very religious) to 4 (not religious) surveyed participants on how much they considered themselves to be a religious person.
Only 23.5% of those that define themselves as conservative or extremely conservative were in favor of abortion compared to 74.2% of liberal or extremely liberal participants. Religion also was a major predictor of attitudes towards abortion, as 24.3% of ‘very religious’ participants and 70.4% of ‘nonreligious’ participants reported support for abortion. Thus, it appears religiosity and political ideology are strongly related to the two main variables of interest.
Consistent with the correlational analysis, religious views (B = -.063, p < .01) and political views (B = -.052, p < .01) each have a significant role in predicting support for abortion.
The more a participant identified as liberal or nonreligious, the less likely they were to oppose Abortion.
Pornography usage at T1 (p = .46) was also not a significant predictor of abortion support at T2.
Pornography was most strongly correlated with religious views in the correlational analysis. Of those surveyed on pornography and religious beliefs, only 27.8% viewed pornography within the past year. Forty-three percent of all nonreligious participants reported pornography use, compared to only 13.7% of those that identified as very religious.
While pornography usage appeared to have a moderate relation to support for abortion in the correlational and regression analysis, this relationship was not as strong to other items related to sexual liberalism such as support for gay marriage and extramarital sex.
Other demographic variables such as education, religion and political views were more significant predictors of support for abortion than usage of pornography.
Based on these findings from both the 2016 GSS survey and GSS panel, pornography usage and abortion both appear to be related to individuals’ political and religious beliefs as well as other attitudes related to sex. However, the relationship between these two variables was weaker than the relationships each shares with other items related to sexual liberalism. The addition of items assessing sexually liberal attitudes in our logistic regression accounted for variance explained by relationship between pornography usage and support for abortion compared to the relationship previously seen in Tokunaga, Wright and McKinley (2015).
Conclusion
Wright and Tokunaga have gained recognition on the 3am model and sexual liberalism with numerous publications within the last decade. In order for their work to reach the next level of notoriety, perhaps they should administer their own surveys with scales that strongly measure their constructs instead of relying on secondary datasets that contain single-item general variables.
It is preliminary to claim any sort of relationship between pornography usage and support for abortion based on analyses from the GSS dataset. How can it be certain that the later support of abortion is directly a result of earlier pornography usage? A continuous measure of pornography would better assess the effect exposure to sexual media has on attitudes related to sex. It would be insightful to survey those whom changed their attitudes on abortion over time to identify similarities (reached certain age, experienced life event, changed political attitudes) that may have influenced their shift in attitudes.
Future research should further describe the definition of sexual liberalism and investigate the time-order relationship between these attitudes and use of pornography. In its current state, the relationship between pornography usage and support for abortion faces a ‘chicken or the egg’ problem as it has not been empirically demonstrated that pornography usage leads to greater support for abortion. The argument presented here is that these attitudes and behaviors are symptoms of a larger ideological construct known as sexual liberalism and are not directly related.
;