Political polarization has become a defining feature of contemporary politics, extending across a broad spectrum of contentious issues including abortion, gun control, and immigration (Iyengar & Westwood, 2015; Mason, 2018; Park, 2014). Within this landscape, partisan identities have coalesced with increasing clarity along these divides, amplifying the polarization phenomenon. It should be noted that the strength of one’s partisan identification has emerged as a robust predictor of preferences on social policy issues, often superseding other demographic factors such as education or age (Baldassarri & Gelman, 2008). This alignment along partisan lines not only exacerbates societal divisions but also poses significant challenges to the effective functioning of democratic processes.
The entrenched nature of political polarization presents formidable obstacles to compromise and consensus-building in policy making (Jang & Sun, 2019; Strickler, 2017). As partisan identities become more deeply entrenched, the space for constructive dialogue and negotiation diminishes (Arbatli & Rosenberg, 2021), hindering the ability of policymakers to address pressing societal concerns in a collaborative manner. Furthermore, the polarization of political discourse fosters an environment where ideological rigidity takes precedence over pragmatic problem-solving, which may perpetuate gridlock and stall progress on critical issues (Hetherington & Rudolph, 2015).
Amidst this backdrop, the role of social media in shaping and reinforcing political attitudes and behaviors has come under increasing scrutiny. This study examines the impact of exposure to cross-cutting political information on social media users, focusing the potential implications for perceived political polarization. In an era where social media serve as primary sources of news and information for a significant portion of the population (Pew Research Center, 2023; Sul, 2018), understanding how exposure to diverse viewpoints influences perceptions of political polarization is essential for comprehending the dynamics of contemporary political discourse.
I collected data through an online survey of South Korean adults (N = 2,641) in March 2023. By investigating the relationship between exposure to cross-cutting political information and perceived political polarization, this study aims to contribute to a deeper understanding of the mechanisms underlying the polarization phenomenon. The current study also intends to provide constructive measures to mitigate the deleterious effects of polarization on democratic governance and societal cohesion.
Literature Review
Motivated Reasoning Theory
Motivated reasoning theory, rooted in cognitive and social psychology, posits that individuals process information in ways that align with their preexisting beliefs, desires, or emotions, leading them to favor information that supports their views while dismissing contradictory evidence (Nir, 2011). This cognitive bias occurs because people are motivated to defend their existing attitudes and maintain coherence in their worldview, which is often linked to personal identity, group affiliation, or emotional investment (Kahan, 2013; van Bavel & Pereira, 2018).
According to Kunda (1990), motivated reasoning can be divided into two types: accuracy motivation and directional motivation. Accuracy motivation occurs when individuals aim to arrive at the most accurate conclusion possible, leading them to process information more systematically and open-mindedly. On the other hand, directional motivation occurs when individuals are driven to reach a specific conclusion that aligns with their preexisting attitudes, leading to biased information processing.
Motivated reasoning influences the way individuals seek out and interpret political information (Camaj, 2019). Taber and Lodge (2006) highlight the phenomenon of confirmation bias, where individuals selectively expose themselves to information that aligns with their beliefs, and disconfirmation bias, where they scrutinize or dismiss opposing information. These cognitive processes reinforce existing beliefs and contribute to political polarization, as individuals become more entrenched in their views and resistant to counterevidence.
Another key concept within motivated reasoning theory is biased assimilation—the tendency to interpret ambiguous or conflicting evidence in a way that favors one’s existing beliefs (Lord et al., 1979). For example, when presented with mixed evidence on a contentious issue, individuals tend to interpret the evidence in ways that affirm their stance rather than reassessing their position objectively (Lord et al., 1979; Nickerson, 1998).
Nyhan and Reifler (2010) demonstrate how motivated reasoning can lead to the backfire effect, where exposure to corrective information, particularly in the political realm, can paradoxically strengthen individuals’ preexisting misconceptions rather than reducing them. This effect illustrates the power of motivated reasoning in reinforcing deeply held beliefs, even in the face of clear, factual contradictions.
Overall, motivated reasoning theory explains the psychological mechanisms that allow individuals to defend and preserve their beliefs, even when confronted with disconfirming evidence. This theory is particularly relevant in the context of political communication, where it explains how people maintain ideological divides despite exposure to diverse or contradictory information.
Exposure to Cross-Cutting Information on Social Media and Perceived Polarization
The rapid expansion of social media has given rise to critical inquiries about how exposure to cross-cutting political information influences individuals’ perceptions about political polarization. However, the direction of this impact remains contested in scholarly discourse.
On the one hand, several researchers argue that exposure to dissimilar perspectives on social media can increase perceptions of polarization. According to motivated reasoning theory, individuals are predisposed to interpret cross-cutting information in a manner that aligns with their preexisting views, thus reinforcing their ideological stances (Garrett & Stroud, 2014; Stroud, 2011). Instead of fostering greater openness to differing perspectives, exposure to cross-cutting political information may trigger defensive mechanisms that solidify ideological rigidity, leading to greater polarization (Kunda, 1990).
This is especially true in social media environments, where algorithms often amplify emotionally charged content, further intensifying users’ negative reactions to opposing views (Tsfati, 2020). Algorithms that prioritize content engagement often lead to selective exposure, which limits the diversity of viewpoints users encounter (Sunstein, 2018). Individuals are thus more likely to be exposed to ideologically consistent content, which reinforces their existing beliefs and reduces opportunities for meaningful engagement with opposing viewpoints (Flaxman et al., 2016). As a result, social media may contribute to the entrenchment of partisan divisions by amplifying motivated reasoning processes (Barberá, 2015; Iyengar & Hahn, 2009). The use of social media can, therefore, create echo chambers or filter bubbles, where individuals primarily encounter viewpoints that align with their own (Sunstein, 2018), making any exposure to opposing perspectives feel threatening.
This phenomenon can be exacerbated by the anonymity and disinhibition prevalent on many social media platforms, which can lead to hostility and incivility in political discussions (Lapidot-Lefler & Barak, 2012). The disinhibition effect, combined with the prevalence of polarized, emotive discourse, often undermines the potential for constructive engagement with opposing viewpoints. Such hostile interactions further intensify polarization by provoking strong emotional responses and creating a social environment that discourages thoughtful consideration of cross-cutting information (Kahan, 2013; Santana, 2014). In these cases, the design of social media, which often prioritizes engagement metrics over content quality, may reinforce the divides it ostensibly seeks to bridge.
However, recent several studies challenge the deterministic view that exposure to cross-cutting information invariably exacerbates political polarization. Some research suggests that, under specific conditions, exposure to opposing perspectives can encourage reflection and reduce polarization (Tynes et al., 2019). Research shows that, while selective exposure remains a prominent issue on social media, deliberate engagement with diverse viewpoints can foster understanding and reduce partisan hostility (e.g., Fishkin et al., 2021; Hartman et al., 2022). Such engagement promotes reflective thinking, which may lead to more moderate political stances and openness to alternative viewpoints.
Moreover, some studies have indicated that exposure to cross-cutting political content can foster greater empathy and reduce ideological rigidity when individuals encounter differing viewpoints in less hostile contexts. Those who actively seek out political diversity demonstrate increased tolerance for opposing viewpoints and show a willingness to revise their political beliefs after engaging with cross-cutting information (Mutz, 2002). This suggests that, under certain conditions, exposure to cross-cutting political perspectives can counterbalance the effects of selective exposure and polarization by broadening users’ understanding of complex political issues (Bakshy et al., 2015; Barberá, 2015).
In light of these contradictory findings, the debate around the impact of cross-cutting information exposure remains unresolved. Theories of motivated reasoning and selective exposure present compelling evidence for the intensification of polarization in social media, but recent research provides an equally strong case for the potential of social media to serve as a platform for depolarization through thoughtful and reflective engagement with opposing viewpoints. Thus, this study poses a research question.
RQ1: Does exposure to cross-cutting information in social media increase or decrease perceived political polarization?
Moderating Role of Political Interest
Motivated reasoning theory argues that individuals are not neutral processors of information but rather engage in selective information processing to defend and protect their pre-existing beliefs, identities, and values (Kunda, 1990; Taber & Lodge, 2006). This process is particularly salient in the political realm, where partisan identities and ideologies are central to individuals’ sense of self and worldview (Westen et al., 2006). Thus, the current study examines what roles political attitude variables play regarding polarization. One of them is political interest, which is expected to play a critical role in shaping how individuals perceive political polarization.
Research shows that individuals with higher levels of political interest exhibit greater ideological rigidity and resistance to information that challenges their beliefs (Taber & Lodge, 2006). This resistance to cognitive dissonance, coupled with the inclination towards motivated reasoning, may contribute to politically motivated individuals perceiving polarization more seriously than those with lower levels of political interest. Political interest may serve as a filter through which individuals interpret and make sense of political information, influencing their perceptions of polarization.
Moreover, the intertwining of political beliefs with individuals’ identities underscores the significance of political interest in moderating perceptions of political polarization. According to Westen et al. (2006), political beliefs often serve as markers of individuals’ identities, prompting them to defend their beliefs against opposing viewpoints as a form of identity protection. For politically motivated individuals, whose identities are closely intertwined with their political beliefs, exposure to counter attitudinal information may evoke defensive reactions aimed at preserving their sense of self and ideological coherence.
At the heart of motivated reasoning is the desire to reduce cognitive dissonance—the psychological discomfort that arises when individuals are confronted with information that contradicts their pre-existing beliefs or attitudes (Harmon-Jones & Mills, 2019). This dissonance is particularly acute in highly politicized environments, where ideological beliefs are closely tied to group identity. According to identity-protective cognition, a specific form of motivated reasoning, individuals process information in ways that affirm their group’s values, seeking to avoid identity-threatening conclusions (Kahan, 2013). For example, when politically interested individuals encounter cross-cutting political information on social media that challenges their party’s stance on a particular issue, they may engage in cognitive strategies that either dismiss the information as biased or distort it to fit their worldview (Garrett & Stroud, 2014). As a result, the perception of political polarization becomes amplified, particularly among individuals who are more politically engaged and invested in their partisan identity.
The amplification of polarization through motivated reasoning is not just a cognitive bias but is also socially reinforced in contemporary media environments. With the rise of social media platforms, individuals are more frequently exposed to ideologically homogeneous content through selective exposure, where they curate their media consumption to align with their political beliefs (Bakshy et al., 2015; Stroud, 2011). This echo chamber effect, combined with motivated reasoning, creates an environment where exposure to dissonant information does not lead to belief updating or opinion moderation but rather entrenches existing attitudes (Frimer et al., 2017). Thus, for those with high political interest, exposure to cross-cutting political information on social media, instead of fostering deliberative democracy, often reinforces perceived polarization through the mechanism of motivated reasoning. Based on the literature, I expect that politically motivated individuals, driven by their ideological commitments and identity considerations, are more susceptible to interpreting political information in a manner that reinforces their preexisting beliefs, thereby amplifying perceptions of polarization.
H1: Political interest moderates the relationship between exposure to cross-cutting information on social media and perceived political polarization, with politically motivated individuals perceiving polarization more seriously than politically less motivated individuals.
Moderating Role of Ideology
Motivated reasoning not only influences how individuals process political information but also interacts with their ideological predispositions. Decades of research in political psychology have demonstrated that conservatives and liberals exhibit distinct cognitive and motivational biases, which shape how they engage with political information (e.g., Hibbing et al., 2014; Jost et al., 2003). These ideological differences are often attributed to underlying psychological traits that influence individuals’ preferences for stability, certainty, and threat perception.
Conservatives, for instance, tend to have a higher need for cognitive closure, which reflects a desire for clear, unambiguous answers and resistance to change (Webster & Kruglanski, 1994). Conservatives, characterized by their preference for certainty and stability, exhibit greater resistance to change, making them inherently predisposed to reject new information that threatens their existing beliefs (Kruglanski & Webster, 1996). In contrast, liberals are generally more open to new experiences and tolerate ambiguity more readily (Carney et al., 2008). These psychological predispositions shape how each group engages in motivated reasoning.
Conservatives’ and liberals’ different cognitive styles contribute to differential perceptions of polarization, especially when exposed to dissonant information on social media. Biased assimilation—the tendency to accept congruent evidence while dismissing incongruent evidence—is a key component of motivated reasoning that operates across the ideological spectrum but is moderated by individuals’ psychological needs and motivations (Taber & Lodge, 2006). As a result, ideological differences shape not only the interpretation of political information but also the extent to which individuals perceive the political landscape as polarized.
In general, the threat posed by information that challenges their beliefs triggers defensive mechanisms aimed at preserving ideological coherence and identity (Denefrio et al., 2018; Greenberg & Jonas, 2003; Westen et al., 2006). This defensive stance is particularly pronounced among conservatives, who may exhibit stronger emotional and cognitive reactions when confronted with counterattitudinal information in social media (Jost et al., 2003; Kunda, 1990). Conservative individuals often demonstrate heightened threat sensitivity, which amplifies their defensive responses to perceived challenges to their worldview (Altemeyer, 1998; Hetherington & Weiler, 2009). Thus, exposure to counterattitudinal information may evoke stronger cognitive dissonance and emotional discomfort among conservatives, leading to a more pronounced polarization perception (Jost et al., 2008; Taber & Lodge, 2006). This defensive reasoning is not limited to conservatives, as liberals also engage in motivated reasoning, albeit in different forms. While liberals might be more open to cross-cutting information, they too selectively integrate new information in ways that affirm their core values, particularly on issues like climate change and social justice (Cohen, 2003; Kahan et al., 2011).
For conservatives, the heightened need for cognitive closure makes them more likely to engage in motivated reasoning when confronted with dissonant political information. Conservative individuals are more inclined to reject or reinterpret cross-cutting information that threatens their ideological worldview because it disrupts their preference for order and stability (Jost et al., 2003). This tendency is also linked to their greater sensitivity to perceived threats, which triggers defensive mechanisms aimed at protecting their social and political identity (Lavine et al., 2002). The result is a selective interpretation of political information that further intensifies the perception of political polarization (Frimer et al., 2017).
This resistance to ambiguity and preference for certainty can lead conservatives to adopt more rigid and inflexible cognitive strategies when processing information, particularly when it conflicts with their beliefs. Moreover, the need for cognitive closure is closely linked to authoritarianism, a trait commonly associated with conservative ideologies (Altemeyer, 1998). Authoritarians tend to adhere to conventional norms and values, exhibiting a preference for order, conformity, and clear-cut distinctions between right and wrong. This adherence to authority and tradition further reinforces the need for cognitive closure among conservatives, as it aligns with their ideological predispositions.
I thus hypothesize that conservatives, because of their stronger cognitive need for closure and greater aversion to dissonance, are more likely to perceive heightened polarization when exposed to cross-cutting political information compared with liberals, who may exhibit more cognitive flexibility but still engage in motivated reasoning to protect their ideological identity.
H2: Ideology moderates the relationship between exposure to cross-cutting information in social media and perceived political polarization, with conservatives perceiving polarization more seriously than liberals.
Role of Education
Education has long been associated with enhanced cognitive skills, critical thinking, and a greater capacity for evidence-based reasoning (Stanovich & West, 2007). Higher education levels are often associated with greater critical thinking abilities, analytical skills, and access to a broader range of information sources (Halpern, 1998). These skills enable individuals to engage in more sophisticated information processing and reasoning, reducing the likelihood of succumbing to motivated reasoning when confronted with conflicting political information. Individuals with higher education levels are more likely to possess the cognitive resources necessary to critically evaluate and integrate complex information, including counter attitudinal political content encountered on social media.
As such, individuals with higher education levels are theoretically better equipped to engage in unbiased information processing, as they possess the cognitive tools to evaluate competing claims, recognize bias, and consider alternative viewpoints. Cognitive sophistication, a trait often associated with higher education, may help individuals mitigate the effects of motivated reasoning by enabling them to resist the pull of identity-protective cognition and evaluate cross-cutting political information more objectively (Kahan, 2013).
However, research also suggests that higher education may not necessarily inoculate individuals against motivated reasoning. Individuals with higher levels of education tend to have stronger partisan identities and ideological affiliations (Ehret et al., 2018). They are more likely to align themselves with a particular political ingroup, whether it be liberal, conservative, or another ideological orientation. This alignment fosters a sense of belonging and solidarity within the ingroup and reinforces individuals’ values, beliefs, and attitudes consistent with their political affiliation (Ehret et al., 2018). Thus, higher levels of education may play a significant role in exacerbating political polarization by reinforcing individuals’ identification with their ideological ingroup and strengthening their commitment to its values and beliefs (Drummond & Fischhoff, 2017). As individuals become more educated, they are more exposed to elite cues and ideological messaging that align with their ingroup’s worldview. This exposure reinforces their existing attitudes and contributes to the entrenchment of partisan divisions (Ehret et al., 2018). In short, education may strengthen partisan identities and make individuals more resistant to cross-cutting information (Baek & Wojcieszak, 2009).
In this sense, education may reinforce motivated reasoning, particularly among those with strong ideological commitments. Indeed, some studies argue that individuals with higher education levels are more adept at generating arguments to support their pre-existing beliefs, making them even more susceptible to motivated reasoning (e.g., Drummond & Fischhoff, 2017; Kahan, 2013). This paradox of cognitive sophistication suggests that education might not reduce motivational reasoning but rather increase individuals’ ability to selectively interpret information in ways that align with their ideological identities.
This study explores the dual role of education in either mitigating or exacerbating motivated reasoning and its impact on perceived political polarization. The expectation is that education may moderate the relationship between exposure to cross-cutting information and perceived polarization, though its effects may be contingent on individuals’ political sophistication and ideological alignment.
RQ2: How does educational attainment moderate the relationship between exposure to cross-cutting information in social media and perceived political polarization?
Method
Data Collection
The study collaborated with Science Research, a survey research firm, to leverage its large database of over 100,000 South Korean adults. This database included essential demographic details such as age, gender, and email addresses, which allowed for precise targeting of the sample. The expertise of the firm in data collection ensured the methodological rigor required to conduct a large-scale survey like this.
The study employed a stratified quota sampling approach to build a quota sample, focusing on two key demographic variables: age and gender. The researchers selected a sample of 4,000 potential respondents from the data pool. The sample was quota-controlled to ensure that the proportions of people in each age and gender group reflected their proportions in the overall South Korean population.
By doing this, the research team ensured that the sample was demographically representative, thereby increasing the external validity of the survey findings. This method guarantees that underrepresented groups are included in the study, particularly if certain demographics (e.g., older individuals or specific gender groups) tend to have lower survey response rates (Kalton, 2020). Because quota sampling relies on pre-determined proportions of demographic subgroups, it allows researchers to overcome some limitations of purely random sampling (Vehovar et al., 2016). For instance, in random sampling, you might end up with over- or under-representation of certain groups. Quota sampling controls for this by ensuring that the sample includes a proportional mix of different demographic groups (Bryman, 2016).
Once the sample of 4,000 individuals was identified, each person received an email invitation to participate in the survey. The invitation included a unique URL for each recipient to access the online survey platform and details about the time required to complete the survey. Participants were entered into a drawing for one of twenty $10 gift cards as a thank you for their time and feedback. To further encourage participation, the research team sent a reminder email to those who had not responded after ten days. This reminder provided an additional opportunity for non-respondents to complete the survey, which is a common technique to minimize non-response bias (Lynn, 2013).
Out of the 4,000 invitations, 2,641 respondents fully completed the survey. This resulted in a 66.0% response rate, a high figure for an online survey (Fincham, 2008). This high response rate suggests that the sampling and recruitment methods were effective in engaging the target audience.
The core purpose of quota sampling was to ensure that the survey results could be generalized to the larger South Korean population, specifically in terms of age and gender distribution. As such, the final sample was nearly balanced, with 49.7% males and 50.3% females, reflecting the gender composition of the South Korean electorate.
Measures
Perceived Political Polarization
Perceived political polarization was assessed using a validated scale adapted from Hetherington and Rudolph (2015). Participants were presented with four statements related to their perceptions of political polarization in society and asked to indicate their level of agreement or disagreement on a 7-point scale, ranging from 1 (strongly disagree) to 7 (strongly agree). The items included in the scale were: (1) “I believe that political polarization in society is increasing,” (2) “There is a growing divide between liberals and conservatives in the country,” (3) “Political debates seem to be more heated and divisive than they used to be,” and (4) “I perceive a lack of willingness among political parties to compromise and work together.” (α = .85, M = 4.8, SD = 2.6).
Exposure to Cross-Cutting Information
Exposure to cross-cutting political information on social media was measured with three items on a 5-point scale from 1 (never) to 5 (very frequently): (1) “How often do you come across political posts or articles in social media that present viewpoints different from your own?” (2) “To what extent do you encounter political content in social media that makes you question or reconsider your own beliefs?” (3) “In the past month, how frequently have you seen political information in social media that contradicts your opinions?” (α = .86, M = 2.9, SD = 1.9).
Political Interest
Political interest was assessed using a single-item measure asking participants to rate their level of interest in politics on a scale from 1 (not at all interested) to 7 (extremely interested) (M = 4.8, SD = 3.8).
Political Ideology
Political ideology was assessed on the following scale: (1) extremely liberal, (2) liberal, (3) slightly liberal, (4) moderate, (5) slightly conservative, (6) conservative, (7) extremely conservative (M = 4.2, SD = 3.0).
Control Variables
To assess general social media use, participants were asked: “How much time do you typically spend on social media a day?” Respondents provided their answers on a 7-point scale, ranging from “less than 30 minutes” to “more than three hours” (M = 4.1, SD = 2.8).
Participants were asked to report how many days in the past week they had engaged with five types of traditional media: (1) network television news, (2) local television news, (3) printed newspapers, (4) radio news, and (5) web-only news sites. Their responses spanned from 0 to 7 days per week, and these values were averaged to establish a traditional media use index (α = .83, M = 3.3, SD = 2.2).
Participants reported their gender (50.3% females), educational attainment (median, some college) and their age as of their last birthday (M = 40.9, SD = 22.8). The annual family household income for 2022 was chosen from a 9-point scale ranging from 1 (less than 20 million KRW [$14,000]) to 9 (more than 100 million KRW [$71,000]) (Median = 50 million KRW [$35,000] - 60 million KRW [$43,000]) (see Table 1).
Analytical Procedures
To examine the research question and hypotheses, this study conducted a series of hierarchical ordinary least squares (OLS) regressions, where the independent variables were introduced into the model in sequential blocks. These measures were introduced in a specific order: demographics, political attributes, media use, and the interaction terms. To mitigate potential issues linked to multicollinearity in the moderation analysis, the means were centered before being utilized to create the interaction terms (Jaccard et al., 1990).
Results
In Table 2 hierarchical regression analysis reveals participants who often encounter ideologically incongruent content exhibited lower levels of polarization (β = –0.15, p < .01, RQ1). Analysis also showed an interaction effect between exposure to cross-cutting content and perceived political polarization, with political interest negatively moderating the relationship (β = 0.12, p < .01). The negative direct effect was offset among those with higher levels of political interest. In other words, people who were more interested in politics were also more likely to perceive political polarization to be getting worse compared to people who were less interested in politics. Thus, H1 is supported.
The analysis also finds ideology moderates the link between exposure to cross-cutting content and perceived political polarization (β = 0.11, p < .01), supporting H2. Conservatives tend to perceive stronger polarization than liberals when they encounter cross-cutting political information on social media.
The analysis shows that education also moderates the association between exposure to cross-cutting content and perceived political polarization (β = 0.13, p < .01, RQ2). Well-educated people tend to perceive stronger polarization than less-educated people when they encounter cross-cutting political information on social media (see Table 2).
Discussion
The findings of this study offer a nuanced understanding of motivated reasoning theory and its implications for political polarization in the context of social media. More than anything, this study found a negative relationship between exposure to cross-cutting political information and perceived polarization (RQ1). According to motivated reasoning (Kunda, 1990; Taber & Lodge, 2006), exposure to conflicting information should activate defensive mechanisms, leading individuals to reject or distort opposing viewpoints, thus intensifying their perception of polarization (Garrett & Stroud, 2014). However, our finding demonstrates the opposite effect.
Regarding this result, several explanations can be addressed. One plausible reason may be connected to the unique nature of social media environments. Traditional models of motivated reasoning assume that individuals process information in a way that aligns with their ideological commitments to reduce cognitive dissonance (Kunda, 1990). This process is thought to be exacerbated by selective exposure—where individuals actively seek information that supports their beliefs (Stroud, 2011). However, on social media, users are often exposed to diverse viewpoints unintentionally, through algorithms, shared content, or discussions within broader networks (Bakshy et al., 2015; Barberá, 2015). This incidental exposure to cross-cutting information may not elicit the same defensive reactions predicted by motivated reasoning theory, because the exposure is passive rather than deliberate (Bakshy et al., 2015).
Another plausible explanation could be rooted in deliberative democracy theory, which emphasizes the value of rational discourse and engagement with diverse perspectives in reducing political polarization (Gutmann & Thompson, 2004). In contrast to motivated reasoning, deliberative democracy theory suggests that exposure to opposing viewpoints in a structured and civil manner can foster greater understanding and tolerance, leading to a reduction in perceived polarization (Mutz, 2006). On social media, certain platforms or discussions may provide spaces for more thoughtful, reflective engagement with cross-cutting content, which can encourage users to reevaluate their political positions rather than defend them. This can result in reduced perceptions of polarization, particularly when users engage in meaningful exchanges that highlight shared values or common ground.
The contact hypothesis from social psychology (Zhou et al., 2019) offers another potential explanation. This theory posits that increased exposure to individuals or groups with different perspectives can reduce intergroup hostility and foster more positive attitudes (Allport, 1954). Applying this framework to the context of political communication, frequent exposure to cross-cutting political content may serve as a form of “contact” with opposing viewpoints, gradually reducing political hostility and fostering greater understanding. Over time, such exposure may challenge stereotypes or misconceptions about opposing political groups, leading individuals to perceive the political landscape as less polarized (Mutz, 2002). This is especially plausible in social media contexts where repeated exposure to cross-cutting information can occur organically, through shared content or comments, facilitating a reduction in perceived polarization over time.
Taken together, while motivated reasoning theory predicts that exposure to cross-cutting information should reinforce perceptions of political polarization, the negative relationship observed in this study suggests that other theoretical mechanisms may be at play. Further research is needed to explore the conditions under which motivated reasoning may be mitigated, as well as to examine the role of social context and platform design in shaping how individuals process political information online.
However, the impact of cross-cutting information exposure on social media does not fully materialize when individual characteristics are considered. Our results indicate that exposure to diverse perspectives on social media does not uniformly reduce or increase perceived polarization but is contingent on individual differences.
First, this study’s support for H1, which found that political interest moderates the relationship between cross-cutting information and perceived political polarization, aligns with existing literature on motivated reasoning (Garrett & Stroud, 2014; Taber & Lodge, 2006). Individuals with higher political interest exhibited stronger perceptions of polarization when exposed to cross-cutting content, reflecting the heightened motivation to defend ideological stances. This suggests that political interest amplifies the cognitive biases inherent in motivated reasoning, leading politically engaged individuals to experience greater polarization even when exposed to dissonant information. This finding contributes to the broader discourse on how political interest shapes the way individuals process information and highlights the importance of considering individual differences in polarization research.
Second, the result that ideology moderates the effect of cross-cutting information on perceived polarization (H2) further illustrates the role of motivated reasoning in reinforcing partisan identities. Conservatives, characterized by their preference for certainty and stability, exhibit greater resistance to change and are more prone to defensive reactions when confronted with challenging information (Jost et al., 2003; Kruglanski & Webster, 1996). This defensive mechanism is further accentuated by the need for cognitive closure (Hibbing et al., 2014), which aligns with conservative predispositions towards maintaining existing beliefs and social structures (Jost et al., 2003; Webster & Kruglanski, 1994). Thus, conservatives are more likely to perceive more polarization than liberals when encountering cross-cutting political information. This finding indicates that individuals will engage in biased assimilation to protect their ideological beliefs (Kahan, 2013; Lord et al., 1979). By highlighting the differential effects of cross-cutting information on conservatives and liberals, this study contributes to a more refined understanding of how ideological predispositions interact with motivated reasoning to shape perceptions of political polarization.
Third, the results regarding the role of education (RQ2) present an intriguing paradox. While education is often associated with greater cognitive sophistication and critical thinking skills (Stanovich & West, 2007), our findings suggest that more educated individuals perceive higher levels of polarization when exposed to cross-cutting content. This aligns with research suggesting that individuals with higher education may be more adept at rationalizing their pre-existing beliefs (Drummond & Fischhoff, 2017; Kahan, 2013). Thus, rather than mitigating the effects of motivated reasoning, education may actually enhance individuals’ ability to engage in selective information processing. This paradox challenges conventional assumptions about the role of education in promoting open-mindedness and underscores the need for further investigation into how cognitive sophistication interacts with motivated reasoning in polarized political environments.
Higher levels of education can intensify political polarization as individuals strengthen their allegiance to their ideological groups and their associated values and beliefs (Drummond & Fischhoff, 2017). Research indicates that education not only correlates with increased partisanship but also deepens ideological divides (Boxell et al., 2017). Drummond and Fischhoff (2017) suggest that this phenomenon occurs because education exposes individuals to elite cues and ideological messaging that resonate with their group’s worldview. Furthermore, education may lead individuals to more rigorously defend their ideological positions, eventually exacerbating polarization (Lelkes & Westwood, 2017). Moreover, highly educated individuals often inhabit social circles comprised primarily of like-minded peers and result in forming echo chambers where dissenting viewpoints are marginalized (Bakshy et al., 2015). Thus, while education is traditionally viewed as a tool for fostering informed citizenship, its role in exacerbating political polarization should also be recognized.
The implications of these findings are significant for understanding the role of social media in political polarization. Social media platforms offer opportunities for exposure to diverse political perspectives (Steinfeld & Lev-on, 2024), and this study suggests that such exposure does lead to reduced polarization, negating motivated reasoning theory. However, the effects of cross-cutting information are contingent upon individual-level factors such as political interest, ideology, and education, which can exacerbate perceptions of polarization. This underscores the importance of considering the psychological mechanisms underlying information processing when evaluating the impact of social media on political polarization.
In terms of practical implications, the findings point to the challenges of fostering constructive political discourse in online environments. Given that politically interested and ideologically committed individuals are more likely to perceive heightened polarization when exposed to cross-cutting content, efforts to promote engagement with diverse perspectives on social media may be met with resistance. To mitigate these effects, social media platforms and policymakers might consider designing interventions that encourage more reflective and less reactive engagement with political content, particularly among highly politicized users. Such interventions could include promoting civility in online discourse, providing contextual information to reduce the emotional intensity of political debates, and encouraging users to critically engage with opposing viewpoints.
Additionally, the role of education in moderating motivated reasoning highlights the importance of fostering critical media literacy skills. While education alone may not be sufficient to reduce motivated reasoning, targeted media literacy programs that emphasize the importance of evaluating evidence and considering alternative viewpoints could help individuals overcome the cognitive biases that contribute to polarization. These programs could be particularly beneficial for politically interested individuals, who may be more motivated to engage with political content but also more prone to biased information processing.
In conclusion, this study provides insights into the complex interplay between exposure to cross-cutting information on social media, individual characteristics, and perceived political polarization. While social media platforms offer opportunities for cross-cutting exposure, their effectiveness in reducing prejudice and fostering empathy is contingent upon various factors, including political interest, ideology, and education levels. While motivated reasoning theory offers a theoretical framework for understanding the potential benefits of exposure to diverse viewpoints, the reality is more nuanced and complicated.
Limitations and Future Research
This study primarily focuses on individuals with access to and engagement with social media, potentially excluding those with limited Internet access or who do not use social media platforms. Furthermore, the survey’s reliance on self-reported data introduces the possibility of response bias, where participants may provide socially desirable answers or inaccurately report their behaviors and attitudes. Future research should continue to explore other individual-level factors that influence how cross-cutting information is processed (for example, political knowledge, political sophistication, and partisan affiliation) and investigate potential strategies for reducing the negative effects of motivated reasoning on political polarization. Lastly, caution is needed in applying our findings to other social contexts, because this study was conducted in South Korea. Despite some limitations, this study reveals valuable insights about the role of social media and individual characteristics in shaping perceptions of political polarization.