The increasing ideological divide between conservative and progressive movements in the USA has become a growing concern for domestic researchers and policy groups investigating the longitudinal effects of political polarization on social behavior (Motyl, 2018; van Prooijen, 2021). Individuals on both sides of the political spectrum are becoming increasingly entrenched in their worldviews, contributing to extreme partisan behavior (Doherty, 2014), often accompanied by moral certainty regarding one’s stance on important social issues (Trivers, 1985). This polarization fosters an “us vs. them” mentality, which is deeply embedded in human social psychology.

Levine et al. (2005) explored the role of social group membership in helping behavior, demonstrating that both shared identity and the inclusiveness of group boundaries significantly influence the likelihood of helping behavior. Their findings revealed that helping behavior is enhanced when social categories are broad, encompassing both ingroup and outgroup members, whereas narrow categorizations, which emphasize group distinctions, limit helping primarily to members of one's ingroup. This finding aligns with the common ingroup identity model, which posits that emphasizing shared group membership can lead to more favorable evaluations and behaviors toward outgroup members (Nier et al., 2001). Political polarization, in this sense, can be understood as a manifestation of ingroup bias (Billig & Tajfel, 1973; Tajfel & Turner, 1979). Humans tend to categorize people into “us” and “them,” often underestimating the extent to which others share political stances (Bauman & Geher, 2003). This psychological tendency reinforces divisions and reduces cooperation between politically opposed groups (see Bauman & Geher, 2003).

However, ingroup bias itself can manifest in multiple ways. Hamley et al. (2020) emphasize that ingroup favoritism, whereby individuals treat their own group more favorably, does not always entail negative attitudes or behaviors toward outgroup members. Brewer (2017) identifies three forms of ingroup bias: Type I involves ingroup favoritism without outgroup derogation, Type II entails outgroup derogation without ingroup favoritism, and Type III consists of both ingroup favoritism and outgroup derogation (i.e., when the outgroup is treated unfairly). These distinctions are crucial, particularly in understanding the extent to which political polarization may impact helping behavior, as individuals may reserve their efforts primarily for ingroup members without necessarily wishing harm upon outgroup members (Hamley et al., 2020).

The current study sought to explore the degree to which biased political thinking might affect how people act toward members of their own political group compared with how they act toward members of the other political group. Specifically, we focused on intended helping behavior, which we predicted to be affected by whether the target of one’s helping is a member of one’s own group versus a member of the other group. Further, we did so using stimuli based on the modern political landscape by defining the ingroup and the outgroup based on affiliating with either the black lives or blue lives matter movements.

Ingroup Bias and Helping Behavior

Social Identity Theory (SIT) introduced the concept of ingroups and outgroups, suggesting that individuals have a tendency to view their ingroup more favorably than outgroups (i.e., ingroup bias or favoritism; Tajfel & Turner, 1979; Turner et al., 1987). This bias can trigger intergroup competition, conflict, or discriminatory behavior, even when individuals are simply aware of the presence of an outgroup. SIT operates on a continuum, ranging from purely interpersonal to purely intergroup behavior, with the intensity of group identification increasing with conflict. Individuals are motivated by self-esteem, which can be derived either through personal achievement or group membership. SIT outlines three core processes: social categorization, social identification, and social comparison. Social categorization allows individuals to organize their social world by grouping people into categories. Through social identification, individuals align themselves with particular groups, adopting behaviors and norms. Social comparison involves individuals evaluating their own group against others, often leading to competition or conflict as they strive to maintain a positive social identity. Social hierarchy and group comparison further contribute to a ranking of group status, with groups often striving for superiority over others. Threats to group status, particularly when one group perceives its superiority to be legitimate, may provoke intense discrimination and resistance to any challenge from another group, as this threatens the stability and legitimacy of the existing social system.

Political polarization can be viewed as a specific form of ingroup bias (Billig & Tajfel, 1973; Tajfel & Turner, 1979). This phenomenon is believed to be an evolved aspect of human social psychology (Geher & Wedberg, 2022; Wilson, 2020), suggesting that in ancestral environments, cooperation with ingroup members would have conferred evolutionary advantages. Helping members of one’s ingroup would have increased the likelihood of future reciprocation, whereas outgroup members, being less likely to offer such reciprocity, were often met with less helping behavior.

Ingroup bias and political polarization are famously connected (Ross & Nisbett, 1991). Research indicates that political polarization is linked to a reduced willingness to engage in helping behavior toward those in the opposing group (Ross & Nisbett, 1991). This aligns with Wang et al. (2024) meta-analytic findings that individuals who receive more social support are more likely to engage in helping behavior. In short, people who see others as in their outgroup tend to not only hold relatively negative opinions of said others, but they also tend to see said others as, essentially, all the same. This phenomenon, which clearly speaks to polarization, is often referred to as outgroup homogeneity (Haslam et al., 1996). In tandem, these social psychological biases make it so that people not only rate others in relatively negative terms, but they tend to see everyone in said other categories in the same way—a classic recipe for political polarization.

Helping behavior (also known as prosocial behavior) is complex from an evolutionary perspective (see Geher & Wedberg, 2022). When we think of our evolved psychology, the idea of helping others at a cost to oneself makes little sense on the surface. However, across the past several decades, various researchers have shown that helping behavior is evolvable in species that meet certain preconditions. Based on his work regarding reciprocal altruism (i.e., each individual helping the other while they help themselves), Trivers (1985) made the case that humans would have been selected to engage in helping behaviors toward others who were likely to reciprocate such helping acts into the future. Importantly, such others are likely to be members of one’s own ingroup, as members of one’s own ingroup are likely to be encountered in one’s future and, thus, might be well-positioned to help one back at a future point. Members of the outgroup, broadly defined, on the other hand, are less likely to return acts of helping. The current work explores this basic prediction in the context of modern political issues in the USA.

The Role of Personality in Helping Behavior in a Politically Polarized World

As something of a secondary set of questions, we also explored the effects of certain personality traits on the intended helping behavior measured in this research. Specifically, based on extensive recent research on understanding personality traits from an evolutionary perspective, the current work included measures of both the Dark Triad (Paulhus & Williams, 2002) and the Light Triad (Kaufman et al., 2019). The Dark Triad conceptualizes personality traits as they relate to Machiavellianism (engaging in manipulative behavior such as lying), Narcissism (inflated sense of worth and degree of selfishness) and Psychopathy (anti-social behavioral characteristics and a lack of empathy; see Paulhus & Williams, 2002). Evidence seems to suggest that the facets of the Dark Triad may influence political attitudes and behaviors, such as, increased preference for authoritarian leadership (Hart et al., 2018), outparty discrimination (Fatke, 2017) and willingness to engage in violent activism (Gøtzsche-Astrup et al., 2015).

The Light triad (Kaufman et al., 2019) is a relatively new personality measure that aims to identify and quantify other-oriented selfless behavior. The Light Triad attempts to measure personality traits as they relate to beliefs in treating people as ends unto themselves and not a mere means to an end (Kantianism), belief in the fundamental goodness of everyone (Faith in Humanity), and a belief in valuing dignity and the unique worth of each individual (Humanism; see Kaufman et al., 2019). The Light triad was also found to correlate with other personality measures including measures of compassion and empathy, interpersonal guilt, and individual life satisfaction (Kaufman et al., 2019). Kaufman and colleagues, in their research, identified a moderate negative relationship between dark triad and light triad scores, to which they suggested that these two constructs are complementary and dynamic over the course of individual development. Generally, we expected Dark Triad traits to correspond to relatively low levels of helping, while we expected Light Triad traits to correspond to relatively high levels of helping. We expected these effects across levels of the independent variables.

The Current Study

This study aimed to examine if political affiliation affects how likely one is to help an individual with opposing versus similar political views. We examined helping, an important index of perceptions of others, related to an ingroup/outgroup context, keeping in mind that helping behavior is one-dimensional and only captures a slice of attitudes toward someone from another group. We employed a randomized, between-subject design via an online survey created on Qualtrics. We hypothesized that:

(1)

Participants identifying as politically extreme would present as less helpful to outgroup members and more helpful to ingroup members and

(2)

Individuals scoring high on dark triad traits would present as less helpful to both ingroup and outgroup members. Similarly,

(3)

We predicted that higher scores on the light triad would correspond to increased helping behavior across levels of the independent variables.

Importantly, we did not have a priori predictions regarding whether relatively progressive participants versus relatively conservative participants would be more or less likely to help either ingroup or outgroup measures. Rather, we predicted members of both groups to be more helpful toward ingroup members and less helpful toward outgroup members.

Method

This study was conducted following approval from the Human Research Ethics Board (HREB) from the affiliated institution.

Participants

A total of 279 individuals participated in the online survey. Only those identifying as extremely progressive or extremely conservative were able to complete the survey and therefore were included in the analyses. The mean age was 24.8 (SD = 10.9). Participants self-reported race such that 53.5% identified as White, 13.8% Latinx, 6.9% Asian, 4.5% Black, 0.3% Native American, 3% not listed, and 1.5% chose to not disclose. 23.4% identified as male, 51.7% female, 5.4% non-binary, 1.8% not listed, and 1.5% preferred not to say. We also asked participants if they were currently in college and 70.6% indicated yes. When analyzing the two political groups separately, extremely progressive (N = 212) had a mean age of 22.9 (SD = 8.4), 69% identified as female, 19% male, and 9% non-binary. In the extremely conservative group (N = 47), the mean age was 32.8 (SD = 15.4), 30% identified as female, 55% male, and 0% non-binary.

Participants were recruited through social media platforms such as Instagram, Twitter, and Facebook. Recruitment scripts were also posted on well-known political groups on social media, various college campuses across the country, and a campus-wide email listserv. The school's subject pool system granted one credit to psychology students for taking the online survey. Psychology students were able to leave the survey at any point and still receive credit. No other incentives were given. To be eligible to participate, individuals must have been 18 years or older and fluent English-speakers.

Procedures

The survey was completed at a time of the individuals choosing, in one sitting. A total of 21 questions were presented and the approximate duration was 10 min. When participants clicked on the link to the survey included in the recruitment script, it directed them to the online survey. The first page of the survey was a consent form which included the purpose, procedures, qualifying inclusion and exclusion criteria, potential risks and benefits, a confidentiality statement, point of contact’s email address, and contact information for the ethics board. Individuals were told that the survey was completely voluntary, and they could stop participating at any point. Participants then had the option to consent and begin the survey or decline and exit.

If the individual consented, they were first asked demographic questions: age, gender, race, and college enrollment status. Then, participants were asked to disclose how they identified politically with three answer choices: extremely politically progressive, extremely politically conservative, or neither. Using this measurement scheme, we were able to have a categorical measure of political affiliation for our study’s purposes. Any participant choosing neither was immediately directed to the end of the survey where they were debriefed and thanked for their time. Credit was still given to psychology students. Those choosing extremely progressive, or extremely conservative were randomized to one of two vignettes. The vignettes depicted a scene of either a black lives matter supporter or a blue lives matter supporter (simply based on logos worn by the targets while pumping gas). Then, participants were asked to imagine that supporter moving into their neighborhood and to what extent they would help that person with varying tasks. The light triad and dirty dozen (dark triad) were then presented. After the assessments were voluntarily completed, a debriefing message appeared on the final page of the survey. The debriefing included appreciation for participation, the author's contact information, the number to the counseling center at the institution, and a message to psychology students instructing them on how to obtain subject pool credit.

Materials and Measures

All procedures, except the recruitment script, were posted on Qualtrics’ web-based platform that was used to create the online survey. As described above, participants were randomized to one of two vignettes. Subjective behavioral decisions, as used in this measure, have been used previously to measure helping behavior (e.g., Ruel et al., 2022). While this measuring system is short of tapping actual behaviors, it generally requires more thought on the part of participants relative to simply using straight-out Likert scales (Silva et al., 2019). The vignettes read as follows:

Pat, who is new in town, is wearing a Blue Lives Matter t-shirt and jeans while pumping gas next to you before you head to work. After pumping gas, Pat heads into the store and comes out with a water.

The only interchangeable phrase was Blue Lives Matter or Black Lives Matter, wherein randomization was evenly distributed among participants, no matter what political affiliation they identified as. The vignette instructions were identical, asking participants to read the statement and keep in mind that subsequent questions may pertain to it.

Helping Questionnaire

The helping questionnaire is a self-report measure consisting of 10 items (Ruel et al., 2022). It was designed simply as a measure of how much someone would help another across an array of everyday kinds of events, capturing participants’ attitudes one-dimensionally. Instructions were modified to ask participants to imagine that Pat (the supporter in the vignettes) had just moved into their neighborhood to establish a more personal connection. Each item was also modified to include the name Pat in it. Examples of the items included: “I would drive Pat to the airport,” “I would buy Pat coffee,” and “I would drive Pat to the hospital in an emergency situation.” Participants were instructed to indicate how much they agree or disagree using a 7-point Likert scale from 1 (strongly disagree) to 7 (strongly agree). The sum of all 10 items were calculated upon analysis, with 70 being the highest possible score, equating to the greatest extent of helping. In the current sample, internal consistency calculated using Cronbach’s alpha for all helping questionnaire items was 0.94.

Light Triad

The Light Triad scale is a self-report measure composed of 12 items (Kaufman et al., 2019). Participants were asked to indicate the extent to which they agreed or disagreed with each of the given statements using a 7-point Likert scale where 1 indicated strongly disagree and 7 indicated strongly agree. The scale consists of three subscales: Faith in Humanity (i.e., “I think people are mostly good”), Humanism (i.e., “I tend to admire others”), and Kantianism (i.e., “I prefer honesty over charm”). A total score was calculated by summing each individual subscale. In the current sample, internal consistency calculated using Cronbach’s alpha for all.

Light Triad items was 0.81.

Dirty Dozen (Dark Triad)

The dirty dozen, measuring dark triad traits, is a self-report measure consisting of 12 items (Jonason & Webster, 2010). Participants were asked to indicate the extent to which they agreed or disagreed with each of the given statements using a 7-point Likert scale where 1 indicated strongly disagree and 7 indicated strongly agree. The scale is comprised of three subscales: Narcissism (i.e., I tend to want others to admire me), Psychopathy (i.e., I tend to lack remorse), and Machiavellianism (i.e., I tend to manipulate others to get my way). Scores may range from 12 to 84, where 84 would indicate the highest level of dark triad traits. A total score was calculated by summing each individual subscale. In the current sample, internal consistency calculated using Cronbach’s alpha for all dirty dozen items was 0.81.