We collected data for two separate projects in this survey and based the target sample size on the project that needed a larger number of participants. A power analysis using G*Power (Faul et al., 2009), assuming an effect size of f = .15, with 2 groups and four measurements and an estimated correlation of 0.50 between repeated measures suggested that we would need a sample size of 294 to achieve 90% power. However, because the other project required a larger sample size, we recruited 425 participants, as preregistered. We recruited equal numbers of gender-balanced samples of Democrats and Republicans (based on their pre-screening information) on Prolific. We excluded 16 participants who did not identify as Democrat or Republican. Two participants revoked their consent after being debriefed and were therefore also excluded, resulting in a final sample size of 407 (198 women, 201 men, 8 nonbinary people; 74.20% White; 78.62% straight; M_age = 42.13, SD = 14.27; see Table S1 in the online supplement for more information).

The survey was advertised as a study about views on decision-making among children and adults and contained a section on children’s bodily autonomy for a separate project. The two parts (children and adults) were presented in randomized order, and the data about children’s autonomy were not analyzed in relation to the data presented here. In the part about adults, participants first responded to questions about decision-making regarding women’s bodies, followed by questions about decision-making regarding men’s bodies.

Each item was presented in the form of a statement describing a woman’s (or man’s) desire to do something or not do something (e.g., “A woman wants minor cosmetic procedures (e.g., laser hair removal);” “A woman does not want to have sex with her husband”). Participants were told to imagine that the woman (or man) in question was married to a man (or woman) and indicated the percentage to which this decision should be made by the woman (or man) herself (himself), her husband (his wife), the government, and religious authorities, with the total percentage summing up to 100 (see Figure S1 in the online supplement for example). Thus, for each individual item, we had 4 different measures, with the percentage allocated to the woman/man herself/himself measuring support for bodily autonomy.

The section about women contained a total of 23 different items, 9 about reproductive health (e.g., “A woman wants to start prescription birth control”), 9 about how women present their bodies (e.g., “A woman wants to get a facial piercing”), and 5 items about health-related decisions (e.g., “A woman wants to get drunk”). In line with our pre-registration, we ran an exploratory factor analysis on the 23 measures capturing to what extent participants thought women should make decisions about their bodies (see pp. 8-10 in the online supplement). This analysis initially suggested a 3-factor solution. However, these factors were highly correlated (rs > .68) with high cross-loadings across factors, as were the two factors in a 2-factor solution (r = .76). We therefore included all items in a single support for women’s bodily autonomy scale that was highly reliable (α = .97).

The section about men included 21 items. However, some of them referred not to their own bodily autonomy but their wife’s (e.g., “A man wants his wife to terminate her pregnancy”). After excluding these items, 18 items remained, all of which mirrored items included for women (e.g., “A man wants to get drunk”). These items also formed a highly reliable scale (α = .96), which we refer to as support for men’s bodily autonomy.

After responding to the measures about women and men, participants indicated to what extent freedom, liberty, justice, equality, unity, and security were important values to them. Liberty and freedom were highly correlated (r = .67) and formed our measure of importance of liberty. Lastly, participants provided demographic information before being debriefed and paid.

Our study makes several contributions. First and foremost, we are (to our knowledge) the first to show that Republicans, compared to Democrats, are less supportive of women’s bodily autonomy across a wide range of behaviors that go beyond reproductive rights and public toplessness.

Second, we show that this is potentially a reflection of Republicans’ general opposition to bodily autonomy for anybody, regardless of gender/sex. Research has already shown that Republicans and Democrats have different definitions of freedom and liberty and views on which freedoms they perceive to be under threat (Armaly & Enders, 2024). We add to this literature by showing that despite valuing liberty and freedom to a high extent, bodily autonomy does not seem to fall under this umbrella for Republicans—at least not to the same extent as it does for Democrats. Indeed, it is possible that Republicans interpret “liberty” primarily as freedom from government, whereas Democrats view liberty as freedom from any interference from others. Future research should examine these nuances further.

The pattern that Republicans are less supportive of bodily autonomy for women and men is in line with a large body of literature showing the link between political affiliation and right-wing authoritarianism (i.e., support for, submission to, and aggressive defense of authorities and traditional norms) (e.g., de Oliveira Santos & Jost, 2024) and Republicans’ reliance on the moral foundation of authority and tradition (Graham et al., 2009). Our work adds to this literature by showing that the body is another site where these beliefs and values take shape and where submission to authorities such as spouses, the government, and religion, are valued over individual self-determination. Given that partisan differences were especially pronounced for women’s bodily autonomy, this pattern has implications for understanding the continued perpetuation of gender roles and gender inequality.

Our work is not without limitations. First, our sample was not representative. Indeed, we specifically recruited only self-identified Democrats and Republicans and aimed for a gender-balanced sample within each category. We did so to eliminate participant gender/sex as a confounding variable, but this also means that our sample does not reflect actual differences in the population, where women are more likely to identify as Democrats and men are more likely to identify as Republicans.

Relatedly, Democrats and Republicans in our sample differed notably in religiosity (with 47% of Democrats, but only 16% of Republicans indicating they were agnostic or atheist). While these numbers are higher than in the general population, they reflect real-world differences in religiosity between Democrats and Republicans (Pew Research Center, 2024). Future research should examine the role that religiosity plays in views about bodily autonomy in general, and women’s bodily autonomy specifically.

Moreover, our investigation was limited to the bodily autonomy of (presumed) cisgender women and men who were married to men and women respectively. Future research should examine support for bodily autonomy of other groups, for example, transgender people, whose bodily autonomy (i.e., their ability to access gender-affirming care) is a polarizing issue in the United States. Likewise, given the high percentage of decision-making power Republicans allocated to spouses, it would be interesting to examine where this decision-making power goes for women and men who are not married or who are married to a same-sex partner.

Republican politicians such as Donald Trump often emphasize the importance of freedom and liberty. Yet they show a lack of support for one fundamental freedom: the right to bodily autonomy. We hope that this research can serve as a first step in exploring this interesting paradox and as a springboard for future research into attitudes towards bodily autonomy more broadly—a topic that has received surprisingly little attention in social psychology.