Pets and politics: Associations between pet ownership, political views, and voting intentions

The survey began with an introduction that explained the study’s purpose and asked for consent to participate (see appendix). The next set of questions asked for demographic information including age, gender, LGBTQ+ identity, race, ethnicity, education, number of children, marital/relationship status, income, US state of residence, and pets owned.

A total of 954 surveys, completed by participants from every US state (Table 1), were used for analysis. The number of participants from Democrat leaning states was 419 (43.7%) with 535 (56.3%) from Republican leaning states. The sample consisted of 515 (54.0%) women, 401 (42.0%) men, 35 (3.7%) nonbinary people, and 3 (0.3%) who chose to self-describe or not answer. Because of the small number of nonbinary people, further analysis was limited to women and men. The participants had a mean age of 37 (SD = 12.6; range = 18–80). A total of 245 (25.7%) indicated they identify as LGBTQ+. Most participants (711, 74.5%) identified as non-Hispanic White (823, 86.3%) and having some college (228, 23.9%), a 4-year college degree (309, 32.4%), or a graduate school degree (152, 15.9%). A total of 628 participants (65.8%) reported having no children, and nearly equal numbers of participants reported being single/never married (443, 46.4%) vs. married/partnered (408, 42.8%). The largest number of participants reported a gross income between $30,000–49,000 (174, 18.2%) and $50,000–69,000 (179, 18.8%). A total of 215 (22.5%) participants had no pets, 308 (32.3%) reported having at least one cat, 175 (18.3%) reported having at least one dog, and 256 (26.8%) reported having at least one dog and at least one cat (Table 2).

From the list of 20 current topics (n = 951), those rated as a significant concern by the largest percentage of participants included healthcare (583, 61.3%), corruption and ethics in government (548, 57.6%), abortion rights (502, 52.8%), economy/jobs (499, 52.5%), and climate change/environment (456, 47.9%) (Fig. 1). The mean level of concern for each item was calculated and differences in concern level were assessed across pet group (cat, dog, both, none). Significant differences were seen for several current topics: climate change and environment (F = 12.23, p < 0.001; cat owners significantly more concerned than dog owners, owners of both, none); LGBTQ+ rights (F = 5.01, p = 0.002; cat owners significantly more concerned than owners of both); immigration reform (F = 3.07, p = 0.027; cat owners significantly less concerned than owners of both); gun control (F = 4.59, p = 0.003; cat owners significantly more concerned than dog owners); religious freedom (F = 3.64, p = 0.013, cat owners significantly less concerned than owners of both); national security and terrorism (F = 4.58, p = 0.003; cat owners significantly less concerned than owners of both); taxes and government spending (F = 4.28, p = 0.005; cat owners significantly less concerned than dog owners); crime and public safety (F = 2.91, p = 0.034, cat owners significantly less concerned than owners of both), voting rights and election integrity (F = 2.92, p = 0.033, cat owners significantly more concerned than dog owners), and animal welfare (F = 22.71, p < 0.001; those with no pets significantly less concerned than dog owners, cat owners or owners of both).

When asked their views about the legality of abortion (n = 949), the majority of participants reported feeling it should be “Legal in all circumstances” (408, 43.0%) or “Legal in most circumstances except, for example, certain gestational limits” (350, 36.9%). A total of 136 participants (14.3%) reported feeling it should be “Illegal in most circumstances except, for example, rape, incest, or when there is a danger to the mother’s life,” and 27 (2.8%) reported feeling it should be “Illegal in all circumstances.” Twenty-eight (3.0%) participants reported being “Unsure/don’t know.” An association was found between pet ownership and views on the legality of abortion. Cat owners were more likely to indicate that they feel that abortion should be legal in all circumstances (162, 53.1%) compared to dog owners (72, 41.1%), owners of both dogs and cats (92, 35.9%), and participants with no pets (82, 38.5%) (X² = 36.48 (df = 12), p <0.001). An association was also seen between LGBTQ+ status and support for the legality of abortion. Specifically, participants who identified as LGBTQ+ were more likely to indicate feeling that abortion should be legal in all circumstances (159, 65.2%) compared to those who did not identify as LGBTQ+ (233, 34.4%) (X² = 80.69 (df = 4), p < 0.001).

When participants were asked to indicate the likelihood they would help a friend or family member who wanted to get an abortion obtain one (n = 949), 486 (51.2%) reported very likely and 178 (18.8%) reported somewhat likely. A total of 135 (14.2%) reported neither likely nor unlikely, 51 (5.4%) reported somewhat unlikely, and 99 (10.4%) reported very unlikely. Cat owners were more likely to indicate they are very likely to help a friend or family member obtain an abortion (182, 59.7%) than dog owners (82, 46.9%), owners of both dogs and cats (117, 45.7%), and participants with no pets (105, 49.3%) (X² = 28.45 (df = 12), p = 0.005). Participants who identified as LGBTQ+ were more likely to indicate they are very likely to help a friend or family member obtain an abortion (191, 78.3%) compared to those who did not identify as LGBTQ+ (280, 41.3%) (X² = 102.11 (df = 4), p < 0.001).

Participants were asked to indicate their support for or opposition to two types of abortion: medication and procedural (n = 949). Most participants strongly (587, 61.9%) or somewhat (130, 13.7%) supported medication abortion. (Table 3). Cat owners were more likely to strongly support medication abortion (221, 72.5%) than dog owners (97, 55.4%), owners of both dogs and cats (141, 55.1%), or participants with no pets (128, 60.1%) (X² = 32.28 (df = 12), p < 0.001). Participants who identified as LGBTQ+ were more likely to indicate that they strongly support medication abortion (213, 87.3%) than those who did not identify as LGBTQ+ (353, 52.1%) (X² = 95.99 (df = 4), p < 0.001).

A total of 528 (55.6%) participants strongly supported procedural abortion (Table 3). Cat owners were more likely to strongly support procedural abortion (199, 65.2%) than dog owners (88, 50.3%), owners of both dogs and cats (122, 47.7%), or participants with no pets (119, 55.9%) (X² = 31.84 (df = 12), p < 0.001). Participants who identified as LGBTQ+ were more likely to indicate that they strongly support procedural abortion (194, 79.5%) than those who did not identify as LGBTQ+ (317, 46.8%) (X² = 85.28 (df = 4), p < 0.001).

Participants responded to a series of additional questions about the US healthcare system’s role in ensuring the right to reproductive autonomy, indicating their agreement level with each statement (n = 949). The majority strongly agreed with each statement, indicating support such that “Every woman should have the right to parent or not parent a child if she chooses” (593, 62.5%), “It is the responsibility of the healthcare system to ensure that every woman has access to a full range of contraceptive options” (609, 64.2%), “It is the responsibility of the healthcare system to ensure that every woman has access to abortion care” (506, 53.3%), and “The healthcare system should play a role in helping everyone achieve full autonomy over their reproductive decisions” (551, 58.1%) (Table 4). The responses to these four statements were summed to obtain a score regarding views on the US healthcare system’s role in ensuring the right to reproductive autonomy.

When asked about political party affiliation (n = 949), 160 participants (16.9%) reported considering themselves a Republican and 415 (43.7%) reported considering themselves a Democrat. Participants who identified as Independent or “something else” (n = 362) were asked if they lean more toward the Republican party or the Democratic party, to which 68 (18.8%) reported Republican and 144 (39.8%) reported Democrat (Table 5). There was a significant association between pet ownership and self-identified political party (Republican, Democrat, or Independent) (X²= 29.69, df = 6, p < 0.001). Cat owners were more likely to identify as a Democrat (157, 55.1%) than dog owners (74, 46.3%), owners of both dogs and cats (92, 36.9%), or those with no pets (92, 45.5%). There was also a significant association between LGBTQ+ status and self-identified political party (X²= 42.43, df = 2, p < 0.001) with those who identified as LGBTQ+ more likely to report being a Democrat (135, 58.5%) than those who did not identify as LGBTQ+ (264, 41.0%).

When asked how participants would rate themselves on a continuum from extremely liberal to extremely conservative (n = 948), 208 (21.9%) rated themselves as extremely liberal, 221 (23.3%) rated themselves as somewhat liberal, and 122 (12.9%) rated themselves as leaning liberal. A total of 179 (18.9%) reported feeling in the middle, while 79 (8.3%) reported leaning conservative, 90 (9.5%) reported feeling somewhat conservative, 44 (4.6%) reported feeling extremely conservative, and 5 (0.5%) indicated they preferred to not answer. To compare responses based on pet ownership, the seven categories from extremely liberal to extremely conservative were combined into three categories: liberal, in the middle, and conservative. There was a significant association between pet ownership and self-description of political leanings (X² = 30.07, df = 6, p < 0.001), with cat owners more likely to self-identify as liberal (208, 68.4%) compared to dog owners (100, 57.5%), owners of both dogs and cats (125, 49.6%), and participants with no pets (118, 55.4%). There was also a significant association between LGBTQ+ status and political leanings (X² = 107.60, df = 2, p < 0.001) with those who identified as LGBTQ+ more likely to report being a liberal (209, 86.0%) than those who did not identify as LGBTQ+ (321, 47.7%).

When participants were asked if they plan to vote in the 2024 Presidential election (n = 948), a total of 772 (81.4%) indicated yes, 69 (7.3%) indicated no, 97 (10.2%) indicated unsure, and 10 (1.0%) were either unable to vote or preferred to not answer. Those who indicated that they plan to vote were asked who they plan to vote for, with 501 (65.0%) indicating that they plan to vote for Harris/Walz, 166 (21.5%) for Trump/Vance, 38 (4.9%) for a third-party candidate, 47 (6.1%) unsure, and 20 (2.5%) preferred to not answer. When evaluating the association between plans to vote for Harris/Walz or Trump/Vance and pet ownership, there was a significant difference (X² = 34.39, df = 3, p < 0.001) between the four groups. Cat owners were more likely to report they plan to vote for Harris/Walz (189, 86.7%) compared to dog owners (83, 74.8%), owners of both dogs and cats (119, 61.7%), or participants with no pets (110, 75.9%). Participants who identified as LGBTQ+ were more likely to report planning to vote for Harris/Walz (172, 94.0%) than those who did not identify as LGBTQ+ (312, 67.1%) (X² = 50.24, df = 1, p < 0.001).

The topic of pets within the context of US politics is not new. Most American presidents have had pets, and many have capitalized on pet-owner relationships to help connect with the general population (Maltzman et al., 2012). Many presidents’ pets played a significant role in shaping public perception. President Harding’s Airedale, Laddie Boy, for example, became a national celebrity and was given a chair at cabinet meetings while Franklin Roosevelt’s black Scottie, Fala, traveled abroad and joined Roosevelt at international meetings (Mutz, 2010). In fact, Barack Obama was the first elected president to not have a pet, something he changed quickly by promising his daughters a dog after the election (Mutz, 2010). Pets, and their care, have also been used to depict politicians in a negative light. The Obama reelection campaign, for example, drew media attention to the fact that Mitt Romney, a rival political candidate, drove to Canada with his dog crate strapped to the top of his car posting images of Romney’s car and crate with the message “how loving owners transport their dogs” (Axelrod, 2012).

In other examples, the love and connection people have with their pets have been used to create division. Recently, there have been several media stories suggesting that Haitian immigrants in a small US city (Springfield, Ohio) are eating people’s pets (Ulmer and Ulmer, 2024). This is despite numerous authority figures and county representatives indicating that there is no support or evidence for these unsubstantiated claims (Epstein and Cabral, 2024; Helmore, 2024), which appear to have been entirely fabricated (Murray, 2024; Reuters Fact Check, 2024). This type of xenophobia and weaponizing of animals and animal welfare has a long, sordid history (Jenkins and Rudd, 2022). The dog-eating stereotype, for example, has historically been utilized to belittle Asians and Asian immigrants (Ho, 2018), and in 2016, Faye Stewart, who was running for the Senate in Oregan, accused Vietnamese refugees of “harvesting” pets for food (Bai, 2016).

Similarly, the negative connotations and perceptions of women cat owners are not new. Yet the suggestion that cat owners, and women cat owners in particular, are more neurotic or unhappy than dog owners or the general population helps perpetuate a negative opinion of this population. Neuroticism is linked to poorer physical and mental health and a lower quality of life (Costa and McCrae, 1992; John and Srivastava, 1999; Roberts et al., 2009; Yoon et al., 2013; Mader et al., 2023). Furthermore, research has documented an association between neuroticism and childhood trauma (Roy, 2002; Moskvina et al., 2007; Ogle et al., 2014).

Our study, however, found no differences in life satisfaction scores based on gender or pet ownership. In other words, women with cats were no more likely to report feeling unsatisfied with their lives than men, dog owners, or people with no pets. This runs contrary to the notion expressed by Republican vice presidential nominee JD Vance who was recorded referring to “a bunch of childless cat ladies who are miserable at their own lives and the choices that they’ve made” (Treisman, 2024).

While pet ownership may not play a role in life satisfaction, it does appear to predict political views; a significantly higher percentage of cat owners reported being liberal compared to dog owners, owners of both dogs and cats, or participants with no pets. Cat owners were also more likely to report they plan to vote for Harris/Walz (87%) compared to dog owners, owners of both dogs and cats, or participants with no pets. Similar trends were seen regarding views pertaining to the legality of abortion; while the majority of cat owners indicated they feel that abortion should be legal in all circumstances, less than half of dog owners, owners of both dogs and cats, or participants with no pets agreed.

These results support previous research on the personality characteristics and voting trends among dog and cat owners (Ivanski et al., 2021). Research related to the 2016 US Presidential election, for example, revealed that the US states with the highest percentage of cat owners were more liberal-leaning, while US states with the highest percentage of dog owners tended to be conservative-leaning (Bratskeir, 2016). Furthermore, 7 of the 10 states in which former President Trump had the most support, even after controlling for population density, were also among the 10 states in the United States with the highest percentage of dog owners (Ivanski et al., 2021).

The relationship between pet ownership and support for former President Trump echoes previous research suggesting that being a dog person is related to conservatism (Ivanski et al., 2021). Greater conservatism predicts a preference for dogs over cats, even after controlling for age, gender, income, and ethnicity (Ivanski et al., 2021). In addition to conservatism, dog people have also been seen to be higher in competitiveness and SDO, defined as the degree to which people support group-based hierarchy and inequality (Alba and Haslam, 2015; Dallago et al., 2008). SDO is the psychosocial basis for the development of hierarchy-enhancing attitudes, values, and beliefs that provide moral and intellectual justification for practices that result in an unequal allocation of resources (Dallago et al., 2008). SDO strongly correlates with a number of social attitudes, ideological beliefs, and behavior patterns that promote inequality in the relationships between groups including prejudice, racism, sexism, and nationalism (Sidanius et al., 1992; Pratto et al., 1994).

Highly correlated to SDO is Right Wing Authoritarianism (RWA), a set of related beliefs and attitudes characterized by conventionalism (adherence to conventional values), authoritarian submission (valuing obedience and respect for authority), and authoritarian aggression (punitive attitudes toward those who diverge from conventional values) (Duckitt, 1989; Dallago et al., 2008; Manson, 2020; Nacke and Riemann, 2023). RWA is related to conservativism (Tarr and Lorr, 1991) and was a significant predictor in determining whether people voted for Trump in the 2016 election (MacWilliams, 2016). The findings that SDO and RWA are correlated with a preference for dogs (McGreal, 2014; Alba and Haslam, 2015) are perhaps explained by the favorable views of authority held by those with RWA beliefs, and the fact that dogs are seen as more loyal and obedient than cats (McGreal, 2014; Alba and Haslam, 2015).

Given the relationship between higher levels of SDO for dog owners when compared to cat owners (Alba and Haslam, 2015), the results of this study that find cat owners are more liberal than dog owners is perhaps unsurprising. Yet, it is interesting to note that owners of both dogs and cats are more conservative than cat owners without dogs. One possible reason may lie within the research pertaining to the relationship between number of children and political views. Although we did not ask cat owners how many cats they had, nor dog owners how many dogs they had, by definition, those who indicated they had one or more dogs and one or more cats had at least two pets in the household. It is possible that this group had more pets per household than dog-only owners or cat-only owners. Since pets are often viewed as children (Turner, 2001; Volsche, 2018; Owens and Grauerholz, 2019; Volsche, 2019), perhaps increased number of pets correlates with increased conservatism.

Parenthood and number of children have been linked to social conservatism, which can be characterized as a tendency to prefer safer, more conventional behaviors, belief in harsher moral judgments, and resistance to cultural changes (Kerry and Murray, 2018; Kerry et al., 2022). Furthermore, while there is a common perception that people tend to become more conservative with age as a consequence of greater insights and experiences, some studies suggest that parenthood, not age, is driving the age-conservatism relationship (Kerry and Murray, 2018; Kerry and Murray, 2020; Kerry et al., 2022). Evidence of this relationship between conservatism and parenthood can be seen in previous US voting patterns. In the 2020 Presidential election, Trump received more votes in US counties with higher birth rates, while the most pro-Biden counties had total birth rates approximately 25% lower than the most pro-Trump counties. Yet, the relationship between political leanings and birth rate is fairly recent; in the 1970s, there was little or no difference in birth rates between liberal and conservative women. This relationship, however, has continued to grow, whereby conservatives are now more likely to be parents than liberals (Kuziemko et al., 2018; Stone, 2020). Changes to the relationship between parenthood and political leanings are also likely related to the many transformations of gender roles and women’s rights that have occurred in the USA since the 1970s including the greater likelihood for women to be highly educated and employed outside the home, better able to delay or prevent pregnancy through contraception, and finding purpose and satisfaction from a variety of sources, including but not limited to, motherhood (Goldin and Katz, 2002; Bernard, 2017). Emerging research and political polls indicate that the American public has become more liberal and more supportive of abortion rights following the Dobbs decision which overturned Roe v. Wade (Thomson-DeVeaux, 2023; McCarthy, 2024; Pew Research Center, 2024). Further research is needed in order to understand how these factors impact people’s reproductive decision making and to ensure the right to parent – or not parent.

In addition to conservatism, having children has been linked with views of abortion. For example, people with more children tend to have more conservative views pertaining to abortion and same-sex marriage (Vogl and Freese, 2020). Any causal relationship between parenting and conservatism remains unclear, however. It is unknown whether conservative people marry and have more children or if getting married and having children makes one more conservative; it has been suggested that it is likely both of these options (Stone, 2020). Holding conservative beliefs and attitudes might make people more likely to marry and have children earlier; alternatively, being responsible for children may influence people’s views, leading them to favor socially conservative ideals (Weeden and Kurzban, 2014). Because many people view their pets as children (Volsche, 2018; Volsche, 2019), it is possible that more pets, similar to increased likelihood of parenting, might be correlated with more conservative views. This cannot be answered in the current study but warrants further exploration.

In addition to exploring the relationship between pet ownership and political views, this study offered some insights into the political views of individuals who identify as LGBTQ+. While there has been increasing amounts of research pertaining to how the public views LGBTQ+ issues, far less research has explored how LGBTQ+ people themselves feel about politics (Peer et al., 2017). The limited research in peer-reviewed studies assessing LGBTQ+ people’s political views has determined that the cohort is more likely to vote Democratic and hold more liberal views than those who do not identify as LGBTQ+ (Hertzog, 1996; Schaffner and Senic, 2006; Lewis et al., 2011; Egan, 2012).

While LGBTQ+ Americans hold distinctive political views on several issues, there is substantial attitudinal diversity within this population group, with some studies suggesting that bisexual respondents are less likely than lesbian/gay respondents to identify as liberal or to vote for Democratic candidates (Lewis et al., 2011; Swank, 2019; Worthen, 2020; Jones, 2021). Yet other studies have discerned small or no differences in ideology or political party based on sexual orientation (Herek et al., 2010). Findings from our study support previous research indicating that individuals who identify as LGBTQ+ are more likely to indicate they are liberal (86%), less likely to be Republican (5%), and plan to vote for Harris/Walz (94%) than those who do not identify as LGBTQ+ (liberal: 48%; Republican: 23%; vote for Harris/Walz: 50%) (Migdon, 2024; Ring, 2024).

The views of LGBTQ+ people pertaining to abortion have received even less attention, even though LGBTQ+ people are as or more likely than the general population of women to experience unplanned pregnancies and abortions, likely due to barriers when accessing sexual and reproductive health services and contraception as well as increased rates of sexual violence among LGBTQ+ populations (Herek et al., 2010; Everett et al., 2017; Assifi et al., 2020; Bowler et al., 2023). LGBTQ+ participants in our study were more likely to indicate they support abortion in all circumstances (65%) and would help a friend or family member obtain an abortion (78%) compared to those who did not identify as LGBTQ+ (legal in all circumstances: 34%; help a friend/family member: 41%). Given the fact that LGBTQ+ people are as likely or more likely than the general population to have an abortion (Herek et al., 2010; Everett et al., 2017; Assifi et al., 2020; Bowler et al., 2023), their views on the subject are especially important to understand.

This study has several limitations inherent in cross-sectional research, including the potential biases of self-reports. It is possible that participants interested in completing this study are more knowledgeable and involved in politics than the general population. The sample was relatively liberal leaning and consisted of participants within the United States with limited racial and ethnic diversity. It is important to note that this lack of racial and ethnic diversity limits our ability to generalize these findings to other populations. We had a limited number of participants who identified as an ethnicity other than white, and given that comparing white with “non-white” participants is not the best practice (Jenkins and Rudd, 2022), we were unable to make meaningful comparisons based on race or ethnicity. We also had a limited number of participants who identified as non-binary and, consequently, did not have adequate statistical power to make meaningful comparisons between women, men, and gender-expansive identities. In addition, the sample was derived from Prolific, an online crowdsourcing platform and viable alternative to MTurk (Peer et al., 2017). Yet, given that participants in online crowdsourcing platforms tend to be younger, more highly educated, and more white than the general US population, caution is advised when generalizing our results (Tang et al., 2022). We also did not ask participants how many cats and/or dogs they owned so we were unable to determine if the number of pets is a predictor of political views, nor did we assess for urban versus rural geographical location.

In conclusion, our study helps illuminate the association between pet ownership and political views. We discovered that cat owners are more likely to report being liberal, support abortion, and plan to vote for Harris/Walz in the upcoming US presidential election than dog owners, owners of both a dog and cat, and non-pet owners. We also found that participants who identify as LGBTQ+ are more likely to report being liberal, support abortion, and plan to vote for Harris/Walz in the upcoming US presidential election than those who do not identify as LGBTQ+. Lastly, life satisfaction did not significantly differ by pet ownership or gender, contrary to popular stereotypes about cat owners. These results offer a springboard for further research to better understand the relationship between pet ownership and political views. For example, these results warrant more comprehensive research on the psychological traits associated with dog versus cat ownership, and the association between these traits and political attitudes across different demographic groups. Analyzing the role of emotional connections with pets in political identity formation, as well as the potential causal mechanisms behind the observed correlation between pet ownership and political leanings, could also prove of value. Further research regarding gender and sexuality, in addition to actual voting behaviors versus voting intentions, could also help further our understanding of the intersection between political views, pet ownership, gender, and sexuality. Lastly, it is unknown if the relationship we detected between pet ownership and political views can be generalized to those outside the United States. Research on this topic in other countries would help us better understand if these findings would be similar in other countries.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The authors confirm that the research meets any required ethical guidelines, including adherence to the legal requirements of the study country.

The authors would like to thank L.W. for her help in the study design.

All authors contributed to the conceptualization, investigation, methodology, project administration, resources, writing – original draft, writing – review and editing.

The authors report no funding for this study.

Data available upon request.

This research does not reflect the authors’ universities’ support or endorsement of any particular candidate or political party.