The Value Clarification Hypothesis: Explaining a Striking Asymmetry in the Link Between Education and Moral Attitudes

The Value Clarification Hypothesis: Explaining a Striking Asymmetry in the Link Between Education and Moral Attitudes | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article The Value Clarification Hypothesis: Explaining a Striking Asymmetry in the Link Between Education and Moral Attitudes Simon Karlsson, Kimmo Eriksson, Chris G. Sibley, Irina Vartanova, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7658969/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Social scientists have long documented a robust association between higher education and liberal moral attitudes. However, this relationship may be more complex than previously understood. Using large-scale survey data from the United States (General Social Survey), we demonstrate that the education-liberalism link is profoundly asymmetric across political ideology. Education strongly predicts liberal attitudes among self-identified liberals but has virtually no association with attitudes among strong conservatives. We explain this striking pattern by introducing and testing the Value Clarification Hypothesis, which proposes that education acts as a value-clarifying rather than uniformly liberalizing force. Higher education enhances individuals' ability to align their attitudes with their more fundamental moral values, but this process produces different outcomes depending on one's ideological starting point. For liberals, whose values center on preventing harm and promoting fairness, exposure to moral arguments consistently reinforces positions justified by such arguments. For conservatives, whose broader moral framework creates cross-cutting considerations, the net effect is negligible. A preregistered replication using similar survey data from New Zealand (the New Zealand Attitudes and Values Survey) confirms both the asymmetric pattern and our theoretical explanation. These findings reframe debates about political bias in higher education and offer new insights into how educational experiences shape political belief systems. higher education political attitudes moral psychology political socialization ideology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Social scientists have long observed a robust link between educational attainment and moral attitudes: individuals with higher levels of education tend to hold more liberal views on a wide range of moral issues (Hyman and Wright 1979; Lipset 1959; Weil 1985). This finding, documented across numerous countries and decades, has fueled a long-standing debate about its origins. Does higher education actively foster liberal attitudes through socialization and the development of critical thinking skills (Surridge 2016)? Or does the correlation reflect a selection effect, where individuals with pre-existing liberal attitudes are simply more likely to seek out advanced education (Campbell and Horowitz 2016)? While this debate has been fruitful, it may overlook a more fundamental feature of the phenomenon. As the articles above illustrate, the link between education and liberalism is typically treated as an ideologically uniform effect. Yet it is not! Our analysis reveals that this relationship is profoundly asymmetric: education strongly predicts liberal attitudes among self-identified liberals but has virtually no effect among conservatives. This finding presents a puzzle that existing theories, focused on the general effects of education, are not well-equipped to explain. To resolve this puzzle, we introduce and test the Value Clarification Hypothesis, an explanation derived from Moral Argument Theory (Strimling et al. 2019). We propose that higher education functions less as a uniformly liberalizing force and more as a catalyst that helps individuals develop attitudes whose justifications are consistent with the individuals’ underlying values. For liberals, this process reinforces liberal positions. For conservatives, however, the net effect is negligible because their values are consistent with justifications for both sides of most moral issues (Graham et al. 2009; Strimling et al. 2019). This framework not only accounts for the ideological asymmetry but also generates a more nuanced prediction: the interaction should depend on the specific moral arguments available for each issue. We test these propositions across three studies, using large-scale survey data from two countries. Study 1 establishes the core asymmetric pattern in U.S. data. Study 2 tests our theoretical explanation by examining how the pattern varies across different types of moral issues. Study 3 provides a preregistered replication in New Zealand, confirming both the phenomenon and our explanation. Our findings offer a more precise account of whose attitudes are most influenced by education and on which issues. By moving beyond the simple selection-versus-socialization dichotomy, we provide new insights into how educational experiences shape the contours of contemporary political divisions. Background: Explanations for the Education-Attitudes Link and the Role of Ideology The well-established link between higher education and liberal moral attitudes has been a central topic of inquiry in the social sciences for decades. Explanations for this phenomenon have historically fallen into two broad categories: those emphasizing selection effects and those proposing direct causal effects (Surridge 2016). The selection hypothesis posits that the correlation is largely spurious; individuals with liberal predispositions are simply more likely to seek out and complete higher education (Gross 2013). In contrast, theories of causal effects contend that the educational experience itself actively fosters liberal attitudes, either through cognitive development (Stubager 2008) or social influence (Surridge 2016; Werfhorst and Graaf 2004). While most research suggests both selection and causal mechanisms are at play (Simon 2022), the precise nature of the causal link remains a subject of intense debate. The Missing Piece: Ideological Moderation To further refine this inquiry, it is crucial to distinguish between broad political ideology and specific moral attitudes. Some researchers have sought to isolate the "pure" effect of education by including ideology as a statistical control (e.g., Weil 1985; Ma-Kellams 2014). However, treating ideology as a mere control variable implicitly assumes that the effect of education is uniform across the ideological spectrum. This approach fails to ask a potentially more important question: does the strength of the relationship between education and attitudes depend on one's ideology? This oversight represents a significant gap in the literature. In Study 1, we directly explore this moderation and reveal a clear pattern: the link is strong among liberals but gradually weakens across the ideological spectrum until it becomes negligible for the most conservative groups. This finding presents a theoretical puzzle. Why would the influence of education be so potent for liberals yet so inert for conservatives? Existing theories do not offer a ready explanation. To explain this ideologically asymmetric link between education and moral attitudes, we draw on Moral Argument Theory (Eriksson and Strimling 2015), which builds upon Moral Foundations Theory (MFT; Graham, Haidt, and Nosek 2009). Ideology as a Proxy for Moral Values: The Moral Foundations Framework MFT posits that moral reasoning is guided by a set of intuitive foundations, most notably the "individualizing" foundations of Harm (preventing suffering) and Fairness (ensuring justice and reciprocity), and the "binding" foundations of Loyalty (fidelity to one's group), Authority (respect for tradition and hierarchy), and Purity (avoiding contamination and degradation). A key, well-replicated insight is that liberals prioritize the individualizing foundations almost exclusively, while conservatives draw on all foundations more equally (Graham et al. 2011; Haidt and Graham 2007). Our model rests only on this robust empirical distinction, which we confirm in Study 3, while remaining agnostic about other, more contested aspects of MFT. From Values to Arguments for Specific Issue Positions: Moral Argument Theory Moral Argument Theory (MAT) extends this framework by focusing on the intrinsic properties of moral attitudes themselves. MAT posits that any stance on an issue is connected to a specific profile of moral arguments, and there is broad societal agreement on which arguments support which positions (Strimling et al. 2019; Vartanova et al. 2021). This allows us to measure an attitude's individualizing argument advantage : the degree to which it is supported by arguments about harm and fairness relative to the opposing attitude. To illustrate the concept of argument advantage, consider same-sex marriage as an example. Arguments in favor are naturally and easily framed in terms of Fairness ("all couples deserve equal rights") and Harm prevention ("denying marriage causes psychological and material harm to same-sex couples"). Arguments against same-sex marriage, by contrast, are more likely to draw on appeals to tradition (Authority: "marriage has always been between a man and woman") or sanctity (Purity: "marriage is sacred and should not be redefined"). Thus, a positive attitude to same-sex marriage has a strong advantage with respect to individualizing arguments. If we know an attitude’s individualizing argument advantage, we can say how well it fits an individual’s values depending on how liberal those values are. For strong liberals, who are assumed to value only individualizing arguments, the attitudes that are consistent with their values are those that have an individualizing argument advantage. The stronger an attitude’s individualizing argument advantage is, the more unambiguous is its fit with liberal values. For strong conservatives, who are assumed to value individualizing and binding arguments equally high, both attitudes on a given issue are consistent with those values. Thus, regardless of which position is supported by individualizing arguments and which position is supported by binding arguments, none of them fit conservative values better than the other one. Fig. 1 illustrates this interaction between the attitude's argumentative properties and the liberalness of the person’s values in determining the attitude-values fit. The Value Clarification Hypothesis This brings us to the core of our explanation, which we formalize as the Value Clarification Hypothesis. Prior research has shown that the values-attitudes fit is better among people with higher cognitive ability (Eriksson et al. 2023), consistent with higher cognitive ability providing better argument comprehension (Britt et al. 2016). Prior research has further demonstrated better values-attitudes fit among people who discuss moral issues more frequently (Eriksson et al. 2022a), consistent with greater exposure to arguments for both sides supporting the alignment process. The Value Clarification Hypothesis says that education catalyzes increased values-attitudes fit in three ways: First, by selecting students with higher cognitive ability. Second, by providing students training that further increases their ability to comprehend arguments. Third, by providing an environment where students are more likely to be exposed to discussion of issues, thereby giving them greater exposure to arguments for both sides of issues. The effect of this alignment process is fundamentally asymmetric, due to the interaction between liberalness and individualizing arguments discussed above. For a strong liberal, the effect of value clarification is a stronger adoption of attitudes with individualizing argument advantage. For a strong conservative, however, no moral attitude has a better fit than the opposite attitude, so the net effect is expected to be negligible. This framework generates two clear hypotheses: Hypothesis 1 (The Ideological Moderation Effect): The positive association between education and holding liberal moral attitudes will be moderated by ideology, such that the association is strongest among liberals and weakest (or non-existent) among conservatives. Hypothesis 2 (The Argument Advantage Effect): The ideological moderation effect described in H1 will itself be conditional on the nature of the moral issue. The effect of education on liberals' attitudes will be strongest for issues with the highest individualizing argument advantage. Fig. 2 illustrates these theoretical predictions and the underlying assumption that liberals have a strong preference for individualizing moral arguments whereas conservatives don’t. To illustrate that the ideology is a continuous spectrum, the figure also shows that moderates are expected to have results in-between liberals and conservatives. In the studies, a seven-step ideology scale is used, capturing the spectrum in an even more fine-grained way. Study Design and Empirical Strategy Our three studies are designed to test the presence of the complex patterns predicted by the Value Clarification Hypothesis. Study 1 documents the core asymmetric pattern. Using U.S. General Social Survey data, we test whether the association between education and moral attitudes is moderated by political ideology (H1). This establishes the individual-level constraint that any adequate theory must explain. Study 2 identifies the issue-level pattern. Using the same dataset, we introduce the concept of "individualizing argument advantage" and test whether this property of moral issues moderates the education-ideology interaction (H2). This establishes a second, cross-level constraint on viable explanations. Study 3 provides preregistered replication in New Zealand. Using the New Zealand Attitudes and Values Survey, we test both patterns to demonstrate their robustness across national contexts and confirm key theoretical assumptions. We argue that the Value Clarification Hypothesis provides a better explanation for the complex and surprising empirical patterns we uncover in these studies than any previous theory of the education attitudes link. This can be described as an "inference to the best explanation" approach (Lipton 2000). The patterns the study is designed to establish are those illustrated in Fig. 2, panels B and C. The Value Clarification Hypothesis readily explains these patterns as follows: The education-attitudes link is driven by education helping people align their attitudes with their preferences for individualizing arguments. Therefore, the link will be strongest among strong liberals who have the strongest preference for individualizing arguments, while it will vanish among strong conservatives as they don’t have that preference. Moreover, this pattern will be moderated by the size of the individualizing argument advantage of the specific attitude. Now consider how previously discussed alternative theories could explain the same patterns. One theory is direct socialization, that colleges teach their students liberal attitudes. Another theory is that adoption of liberal attitudes is facilitated by the cognitive development that higher education stimulates. Both these theories can explain why liberal attitudes are more common among the highly educated than in the population at large. However, they do not explain why attitudes do not differ between educated and uneducated conservatives. A third theory is that people self-select to higher education based on their politics. While this theory can explain why there are more liberals among the highly educated than in the population at large, it does not explain why attitudes differ between educated and uneducated liberals or why the education-attitudes link varies systematically across attitudes. Our model incorporates a more nuanced role for selection, distinguishing between selection on attitudes and selection on cognitive ability. While simple selection on prior attitudes is insufficient, selection on cognitive ability (Deary and Johnson 2010) plays a contributing role. Because individuals with higher cognitive ability are more likely to attend university, the student population is pre-selected for stronger argument comprehension skills. This means the value clarification process has likely already begun before university. Higher education then acts as an accelerator for this process. Our model sees cognitive selection and educational socialization not as competing explanations, but as intertwined forces. Study 1: The Asymmetric Link Between Education and Attitudes in the U.S. Study 1 tests our core proposition: that the link between education and liberal moral attitudes is moderated by political ideology. We hypothesized that the association would be strongest among liberals and weakest among conservatives (H1). Method Data and Participants. We analyzed data from the General Social Survey (GSS), a nationally representative, biennial survey of the U.S. population covering the period from 1972 to 2021 (Davern et al. 2024). Measures. We measured moral attitudes, education, and political ideology as follows. Moral Attitudes. We used a set of 98 moral attitude items from the GSS, following prior research (Eriksson et al. 2022b). To facilitate analysis, all items were first dichotomized by removing neutral options and grouping graded responses. We then empirically determined the liberal stance for each issue by examining its correlation with participants' political ideology so that the more liberal position was consistently coded as 1 and the more conservative position as 0. Education. Participants' educational attainment was measured by the total number of years of schooling completed, ranging from 0 to 20. Political Ideology. Participants' political ideology was assessed with the item "polviews," a 7-point scale. To aid interpretation, we recoded this variable so that it ranged from 1 (Extremely conservative) to 7 (Extremely liberal). Analysis Plan. To test for moderation, we ran a separate logistic regression for each of the 98 moral attitude items. In each model, the dichotomous attitude was the outcome variable, and the predictors were the participant's education level, their political ideology, and the interaction term between education and ideology. To test Hypothesis 1, we examined the sign and significance of the interaction term for each moral issue. With ideology coded such that higher values indicate greater liberalism, we expected a positive coefficient, indicating that the positive effect of education on holding a liberal attitude is amplified as ideology becomes more liberal. Results The results provided strong support for Hypothesis 1. Across the 98 moral issues, the interaction term between education and ideology was in the expected positive direction for almost all items: 95 of 98, or 96.9% (95% CI from 91.3% to 99.4%, p < .001). To visualize this consistent pattern, we estimated the strength of the education-attitude link separately for each of the seven ideological groups. As shown in Fig. 3, the association between education and holding liberal attitudes was close to zero among the most conservative respondents, and increased in strength in a stepwise fashion across the ideological spectrum. The strongest link between education and liberal attitudes was observed among the most liberal respondents. Discussion The findings from Study 1 establish a clear and robust empirical pattern: the well-documented link between education and liberal moral attitudes is profoundly asymmetric. The association is robust in explaining attitudes among liberals, but has virtually no predictive power among conservatives. This asymmetry constitutes the theoretical puzzle that the remainder of this paper seeks to explain. Having established this phenomenon, we now turn to Study 2 to test our proposed theoretical explanation. Study 2: Explaining the Asymmetry with Moral Arguments Study 1 established that the link between education and liberal attitudes is strongest for liberals. Study 2 tests our theoretical explanation for this phenomenon: that this interaction is itself driven by the moral content of the issues. Specifically, we test Hypothesis 2, which predicts that the moderating effect of ideology will be most pronounced for issues where the liberal position has a strong individualizing argument advantage. Method Data and Participants. The analysis for Study 2 used the same GSS dataset as Study 1. Measures. The variables Education and Political Ideology were measured and coded identically to Study 1. Moral Attitudes . As in Study 1, all responses were dichotomized by excluding neutral options and collapsing graded responses. However, in this study, we retained the original coding with 1 indicating “agree” and 0 indicating “disagree,” because the argument advantage variable (described below) captured whether an opinion was liberal or conservative. Individualizing Argument Advantage. This new variable was measured at the issue level. We used data from a prior study in which participants evaluated the arguments for and against each of the 98 moral attitudes (Vartanova et al., 2021). For each attitude, we calculated the proportion of respondents who endorsed Harm- and Fairness-based arguments for that position, and subtracted the proportion who endorsed those same arguments for the opposing position. This yielded a continuous score for each of the 98 issues, where a high positive score indicates a strong individualizing argument advantage for the liberal attitude. (See the Online Resource for the full list of items and their advantage scores.) Analysis Plan. To test Hypothesis 2, we used a multilevel logistic regression model that included all participants and all moral issues simultaneously. This approach accounts for the nested structure of the data (responses nested within individuals, and individuals nested within issues). The outcome variable was the dichotomous moral attitude. The model included fixed effects for education, ideology, the individualizing argument advantage of the issue, all two-way interactions, and the crucial three-way interaction term (Education × Ideology × Argument Advantage). Our primary test of Hypothesis 2 was the sign and significance of this three-way interaction term. We predicted a positive coefficient, indicating that the two-way interaction between education and ideology (found in Study 1) becomes stronger as the argument advantage of the liberal attitude increases. Results The multilevel model revealed a significant and positive three-way interaction between education, ideology, and individualizing argument advantage, providing strong support for Hypothesis 2 (for further details, see Table S1 in the Online Resource). To illustrate this complex interaction, we plotted the strength of the education-attitude link as a function of the argument advantage score, separately for each of the seven ideological groups (see Fig. 4). The results show a striking pattern. For "Extremely liberal" individuals, there is a steep, positive relationship: the more an attitude is supported by individualizing arguments, the stronger its association with education. This relationship becomes progressively flatter as we move across the ideological spectrum. For "Extremely conservative" individuals, the line is almost perfectly flat, indicating that the link between their education and their attitudes is entirely insensitive to the argument advantage of the issue. Discussion The findings from Study 2 provide strong support for our theoretical model. The data confirm that the asymmetric effect of education is not random but is systematically related to the moral content of the issues themselves. Education appears to amplify the process of value-argument alignment, but this only occurs for individuals whose values are aligned with the arguments that define the liberal side of the debate (liberals). For conservatives, whose values create more complex considerations, this alignment process does not produce a clear directional effect. Having established our proposed mechanism in this exploratory analysis, we turn to Study 3 for a preregistered replication. Study 3: Preregistered Replication in New Zealand Study 3 provides a preregistered replication of the key findings from Studies 1 and 2 in a different national context. This provides a robust test of the generalizability of both the ideological asymmetry (H1) and our proposed explanation for it (H2). Additionally, this study provides a direct test of a core assumption of our theory: that liberal ideology serves as a proxy for a strong preference for individualizing moral arguments. Method Data and Participants. We analyzed data from the New Zealand Attitudes and Values Survey (NZAVS), a national longitudinal panel study that has collected data annually since 2009 (Sibley 2023). Our analysis includes all 14 waves of data available at the time of the study. Preregistration. The hypotheses, measures, and analysis plan for this study were preregistered prior to data analysis. Due to space constraints on the preregistration platform, the study was preregistered in multiple documents: https://aspredicted.org/1JV_PK9 (specifies the selection of moral issues in the NZAVS, as well as the procedures for gathering and calculating argument advantages), https://aspredicted.org/FW3_2R5 (specifies certain preliminary analyses), and https://aspredicted.org/Y3X_KL5 (specifies the analyses for all hypotheses about the education-attitudes link, referred to as H2a, H2b, and H2c in the preregistration). Analytic deviations from preregistration. For consistency with Hypothesis 2 in the GSS data, we retained the original coding of dichotomized attitude variables in the NZAVS rather than reverse-coding when the argument advantage was negative. This decision does not affect H1, and H2 remains supported when analyzed according to the preregistered plan (see the Online Resource). In addition, inspection of model output revealed that the preregistered multilevel logistic regression models produced non-normal random intercepts for several issues, with distributions that were highly skewed or bimodal. As a robustness check, we re-estimated each hypothesis using models that do not assume normally distributed random intercepts. Results were consistent across specifications (see the Online Resource). Measures. In addition to the variables in Study 2, this study also includes respondents' preference for individualizing arguments. Moral Attitudes. We identified 36 moral attitude items from the NZAVS that met the inclusion criteria. Of these, 29 exhibited a significant argument advantage and were retained for analysis, as specified in the preregistration. All items were dichotomized by removing neutral responses and collapsing graded options into binary categories (1 = agree, 0 = disagree). In line with the preregistered plan, the liberal position on each issue was determined by its individualizing argument advantage: if the advantage was negative, the liberal stance was coded as “disagree”; if positive, as “agree.” Notably, this approach produced results highly consistent with the ideology–attitudes association used in Study 1, with 26 of the 29 issues yielding the same directional sign. Education. Educational attainment was measured on an 11-point scale, ranging from 0 (no qualification) to 10 (doctorate degree). Political Ideology. In the NZAVS, political ideology was measured with a 7-point item ranging from Extremely liberal (1) to Extremely conservative (7), with a neutral midpoint at 4. Because this item has been included in multiple survey waves, we calculated each participant’s mean response across all available waves. We classified respondents as liberal if their mean score was ≤ 3 and as conservative if their mean score was ≥ 5. The binary scale was used in preregistered analyses, whereas figures displayed the full 7-point scale for consistency with Studies 1 and 2. Individualizing Argument Advantage. Argument advantage scores for the 36 NZAVS items were collected in a new survey of 537 New Zealand participants, following the same procedure used for the GSS items. (See the Online Resource for a detailed description of this data collection and the full list of items and their advantage scores). Preference for Individualizing Arguments. In one wave of the survey, participants completed the 30-item Moral Foundations Questionnaire (MFQ; Graham et al. 2011). From this, we calculated each participant's preference for individualizing concerns by subtracting their average score on the binding foundations (Loyalty, Authority, Purity) from their average score on the individualizing foundations (Harm, Fairness). Analysis Plan. First, to test our theoretical assumption, we examined the correlation between participants' political ideology and their preference for individualizing arguments. Second, to replicate Study 1, we tested for the Education × Ideology interaction (H1). Third, to replicate Study 2, we tested for the Education × Ideology × Argument Advantage interaction (H2). As a robustness check of H2, we also ran a linear regression of the interaction effect obtained in H1 on the individualizing argument advantage to assess whether the advantage positively predicted the size of the interaction, as we expected in the preregistration. Due to the panel nature of the NZAVS data, all models were multilevel, accounting for repeated observations nested within individuals. Results The data provided strong support for our core theoretical assumption of a positive relationship between political liberalism and preference for individualizing arguments ( r = .54, 95% CI [.51, .56], p < .001). As shown in Fig. 5A, the preference for individualizing arguments was captured by ideological liberalism: while no significant preference was found among the most conservative respondents, the preference increased stepwise with greater liberalism, becoming very strong among the most liberal respondents. Second, the results replicated the findings from Study 1, providing further support for H1. Across the 29 moral issues, the interaction term between education and ideology was in the expected direction for 26 of 29 issues (89.6%, 95% CI from 72.6% to 97.8%, p < .001). As shown in Fig. 5B, the positive association between education and liberal attitudes was strongest among the most liberal New Zealanders and disappeared among the most conservative. Finally, we successfully replicated the findings from Study 2. The three-way interaction between education, ideology, and argument advantage was positive and significant, supporting H2 (see Table S1 in the Online Resource). In addition, the robustness check confirmed that argument advantages were positively associated with the Education × Ideology interaction obtained in H1 (β = .376, 95% CI [.240, .511], p < .001). As shown in Fig. 5C, the relationship between education and attitudes was highly sensitive to the argument advantage of the issue for liberals, but this sensitivity diminished and vanished for conservatives. Discussion The results of Study 3 provide strong, preregistered support for our theoretical account. We successfully replicated the core findings from the U.S. data in a different Western democracy, demonstrating the robustness of both the asymmetric link between education and attitudes and its moderation by the moral content of the issues. Furthermore, we confirmed a key assumption of our model: that the liberal-conservative ideological dimension maps directly onto the preference for individualizing moral foundations. This confirmatory evidence provides a solid foundation for the generalizability of our theory. General Discussion For decades, social scientists have sought to understand why higher education is associated with more liberal moral attitudes. Across three studies, we identified a novel empirical phenomenon—that the education-liberal attitudes link is profoundly asymmetric—and provided a theoretical model to account for it. Our findings in the United States, replicated in a preregistered study in New Zealand, demonstrate that the effect of education is conditional not only on who is being educated but also on the moral arguments surrounding an issue. Revisiting the Education-Attitudes Link: The Value Clarification Hypothesis Our central theoretical contribution is the Value Clarification Hypothesis, which posits that education enhances individuals' ability to connect their underlying moral values to specific attitudes through a process of value-argument alignment. The striking asymmetry arises because the moral value systems of liberals and conservatives are structured differently. This explanation was supported by our finding that the effects were strongest for issues with a clear individualizing argument advantage, and that ideological liberalism itself maps strongly onto a preference for these same arguments. Rather than operating as a uniform liberalizing force, higher education appears to function as a value-clarifying mechanism. The empirical patterns we document present significant challenges to existing theoretical accounts. The strength of the Value Clarification Hypothesis lies not only in fitting these data, but in generating specific, cross-level predictions that were confirmed across multiple national contexts through preregistered replication. While we cannot claim definitive causal proof, the convergent evidence provides substantial support for our proposed mechanism over existing alternatives. Implications for Understanding Political Bias in Higher Education These findings reframe the public debate about ideological bias in academia. The fact that the education-attitudes link was essentially zero for strong conservatives suggests that education is not a monolithic liberalizing force. This suggests the link is driven less by indoctrination and more by a process of intellectual exploration that allows students to build more coherent belief systems based on their pre-existing values. Broader Implications for Political Socialization Our research contributes to broader understanding of how education shapes political belief systems. The findings help explain previously observed inconsistencies in the education-attitudes link across different contexts. For example, the finding that the link is weaker in some countries (Weakliem 2002; Weil 1985) may be partly explained by those countries having different ideological compositions in their populations. Our model suggests that future research should account for the ideological distribution of samples when estimating the effects of education. The pattern also has implications for understanding political polarization. If higher education primarily serves to clarify and strengthen existing ideological commitments rather than fostering convergence toward liberal positions, this could contribute to the increasing polarization observed in many democracies. Educated liberals and conservatives may emerge from university with more crystallized and internally consistent—but divergent—worldviews. Limitations and Future Directions Our research has several important limitations. The fact that we obtained similar findings in the United States and New Zealand suggest some degree of cross-cultural generalizability. However, as both societies are Western democracies, replication in culturally more distant contexts is required to answer how far the findings can be generalized. Another limitation is that we did not measure whether education in fact enhances argument exposure and comprehension, as we assume it does. While beyond the scope of this study, this could be examined in future research. Conclusion The link between education and moral attitudes is more complex and systematic than previously understood. It is not a uniform liberalizing tide that lifts all boats. Instead, it is a highly specific current, powerful for those with a liberal compass and nearly still for those with a conservative one. By identifying this asymmetry and providing a robust theoretical explanation for it, this research advances our understanding of political socialization and offers a new lens through which to view the enduring relationship between what we know and what we believe. In an era of increasing concern about political polarization and the role of educational institutions in democratic societies, these findings suggest that universities may be less politically transformative than either critics or advocates claim. Rather than systematically converting students to liberal positions, higher education appears to help individuals—particularly those with liberal inclinations—develop more coherent belief systems aligned with their pre-existing values. This nuanced understanding has important implications for ongoing debates about academic freedom and intellectual diversity in higher education. Our findings suggest that concerns about uniform liberal indoctrination may be overstated, while also indicating that educational experiences do have differential effects across the political spectrum. The key insight is that education functions less as a liberalizing force and more as a value-clarifying mechanism that helps students connect their values to specific positions through enhanced exposure to and comprehension of moral arguments. By identifying these previously unrecognized patterns and providing a theoretical framework to explain them, this research offers new directions for understanding how educational experiences shape political belief systems. While we do not claim definitive proof of our proposed mechanism, the empirical patterns we document present clear challenges to existing theories and point toward more sophisticated models of educational influence on political attitudes. Declarations Data Availability Data from the General Social Survey (GSS; http://doi.org/10.5255/UKDA-Series-200006) are publicly available. Data from the New Zealand Attitudes and Values Study (NZAVS) are not publicly available but can be requested from the study administrators. Argument advantage scores for all items included in the analyses are provided in the Online Resource. Ethical Statement This study relied on secondary analysis of publicly available, anonymized survey data from the General Social Survey (GSS) and data collected under ethical oversight from the New Zealand Attitudes and Values Study (NZAVS). No additional ethics approval was required. Funding This work was supported by the Knut and Alice Wallenberg Foundation [grant number 2022.0191] and the Swedish Research Council [grant number 2021–01654]. Acknowledgement The authors have no acknowledgements to declare. References Britt, M. Anne, Kristopher J. Kopp, Amanda M. Durik, Dylan Blaum, and Peter Hastings. 2016. “Identifying General Cognitive Abilities Involved in Argument Comprehension and Evaluation.” Zeitschrift Für Pädagogische Psychologie / German Journal of Educational Psychology 30(2–3):79–95. doi: 10.1024/1010-0652/a000173. Campbell, Colin, and Jonathan Horowitz. 2016. “Does College Influence Sociopolitical Attitudes?” Sociology of Education 89(1):40–58. doi: 10.1177/0038040715617224. Davern, Michael, Rene Bautista, Jeremy Freese, Pamela Herd, and Stephen L. Morgan. 2024. “General Social Survey 1972-2024. [Machine-Readable Data File]. Principal Investigator, Michael Davern; Co-Principal Investigators, Rene Bautista, Jeremy Freese, Pamela Herd, and Stephen L. Morgan. Sponsored by National Science Foundation. NORC Ed. Chicago: NORC, 2024: NORC at the University of Chicago [Producer and Distributor].” Deary, Ian J., and Wendy Johnson. 2010. “Intelligence and Education: Causal Perceptions Drive Analytic Processes and Therefore Conclusions.” International Journal of Epidemiology 39(5):1362–69. doi: 10.1093/ije/dyq072. Eriksson, Kimmo, and Pontus Strimling. 2015. “Group Differences in Broadness of Values May Drive Dynamics of Public Opinion on Moral Issues.” Mathematical Social Sciences 77:1–8. doi: 10.1016/j.mathsocsci.2015.06.004. Eriksson, Kimmo, Irina Vartanova, Isabela Hazin, and Pontus Strimling. 2023. “Cognitive Ability and Ideology Join Forces in the Culture War: A Model of Opinion Formation” edited by J. Druckman. PNAS Nexus 2(6):pgad205. doi: 10.1093/pnasnexus/pgad205. Eriksson, Kimmo, Irina Vartanova, and Pontus Strimling. 2022a. “How Does Political Discussion Frequency Impact Political Moral Opinions? The Moral Argument Theory of Opinion Dynamics.” Frontiers in Psychology 13. Eriksson, Kimmo, Irina Vartanova, and Pontus Strimling. 2022b. “Opinion Trends on Moral Issues in the United States and the United Kingdom Explained by the Applicability of Generally Accepted Arguments.” International Journal of Public Opinion Research 34(1):edac001. doi: 10.1093/ijpor/edac001. Fitzmaurice, Garrett M., Nan M. Laird, and James H. Ware. 2011. Applied Longitudinal Analysis . 2nd edition. Hoboken, New Jersey: Wiley. Graham, Jesse, Jonathan Haidt, and Brian A. Nosek. 2009. “Liberals and Conservatives Rely on Different Sets of Moral Foundations.” Journal of Personality and Social Psychology 96(5):1029–46. doi: 10.1037/a0015141. Graham, Jesse, Brian Nosek, Jonathan Haidt, Ravi Iyer, Sena Koleva, and Peter Ditto. 2011. “Mapping the Moral Domain.” Journal of Personality and Social Psychology 101:366–85. doi: 10.1037/a0021847. Gross, Neil. 2013. Why Are Professors Liberal and Why Do Conservatives Care? Harvard University Press. Haidt, Jonathan, and Jesse Graham. 2007. “When Morality Opposes Justice: Conservatives Have Moral Intuitions That Liberals May Not Recognize.” Social Justice Research 20(1):98–116. doi: 10.1007/s11211-007-0034-z. Hyman, Herbert Hiram, and Charles Robert Wright. 1979. Education’s Lasting Influence on Values . University of Chicago Press. Lipton, Peter (2000). Inference to the Best Explanation. In W. Newton-Smith, A companion to the philosophy of science. Malden, Mass.: Blackwell. pp. 184–193. Lipset, Seymour Martin. 1959. “Some Social Requisites of Democracy: Economic Development and Political Legitimacy.” American Political Science Review 53(1):69–105. doi: 10.2307/1951731. Ma-Kellams, Christine, Aida Rocci Ruiz, Jacqueline Lee, and Andrea Madu. “Not All Education Is Equally Liberal: The Effects of Science Education on Political Attitudes.” Journal of Social and Political Psychology 2, no. 1 (2014): 1. https://doi.org/10.5964/jspp.v2i1.259. Sibley, C. G. 2023. “Sampling Procedure and Sample Details for the New Zealand Attitudes and Values Study.” doi: https://doi.org/10.31234/osf.io/wgqvy. Simon, Elizabeth. 2022. “Demystifying the Link between Higher Education and Liberal Values: A within-Sibship Analysis of British Individuals’ Attitudes from 1994–2020.” The British Journal of Sociology 73(5):967–84. doi: 10.1111/1468-4446.12972. Strimling, Pontus, Irina Vartanova, Fredrik Jansson, and Kimmo Eriksson. 2019. “The Connection between Moral Positions and Moral Arguments Drives Opinion Change.” Nature Human Behaviour 3(9):922–30. doi: 10.1038/s41562-019-0647-x. Stubager, Rune. 2008. “Education Effects on Authoritarian–Libertarian Values: A Question of Socialization1.” The British Journal of Sociology 59(2):327–50. doi: 10.1111/j.1468-4446.2008.00196.x. Surridge, Paula. 2016. “Education and Liberalism: Pursuing the Link.” Oxford Review of Education 42(2):146–64. doi: 10.1080/03054985.2016.1151408. Vartanova, Irina, Kimmo Eriksson, Isabela Hazin, and Pontus Strimling. 2021. “Different Populations Agree on Which Moral Arguments Underlie Which Opinions.” Frontiers in Psychology 12. Weakliem, D. L. 2002. “The Effects of Education on Political Opinions: An International Study.” International Journal of Public Opinion Research 14(2):141–57. doi: 10.1093/ijpor/14.2.141. Weil, Frederick D. 1985. “The Variable Effects of Education on Liberal Attitudes: A Comparative- Historical Analysis of Anti-Semitism Using Public Opinion Survey Data.” American Sociological Review 50(4):458. doi: 10.2307/2095433. Werfhorst, Herman G. Van de, and Nan Dirk de Graaf. 2004. “The Sources of Political Orientations in Post-Industrial Society: Social Class and Education Revisited.” The British Journal of Sociology 55(2):211–35. doi: 10.1111/j.1468-4446.2004.00016.x. Additional Declarations No competing interests reported. Supplementary Files OnlineResource.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7658969","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":528765736,"identity":"505ab2ff-a917-4587-b541-914f1325a242","order_by":0,"name":"Simon Karlsson","email":"","orcid":"","institution":"The Institute for Future Studies","correspondingAuthor":false,"prefix":"","firstName":"Simon","middleName":"","lastName":"Karlsson","suffix":""},{"id":528765737,"identity":"07a52a89-d174-42e9-97e4-231a16b7663f","order_by":1,"name":"Kimmo Eriksson","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6klEQVRIiWNgGAWjYBACCWYog40HTNnwsDcfIE1LGg/PsQQCWuAsiJbDDAS1SLZzp274wGCTz8dz/PGHjzvOy/CwsT9g/FGBW4s0M++2mzMY0izbeHvMJGeeuc3Dw8ZjwMxzBrcWOaCW2zwMhw3Y+HnYmHnbbvPYy/cwMDO2EdTyH6iF/fFn3rZzPGCH/fyH32FALQcM2HgbDKR52w4AtTAYMPA24PF+M8gvBskGbDxngH5pSwb75TDPMdxaJM6f3XbjQ4WdgXxPOjDE2uzsgQ57+PBHDW4tEGCAxj9ASMMoGAWjYBSMAvwAAG7UQ9BXQb1sAAAAAElFTkSuQmCC","orcid":"","institution":"The Institute for Future Studies","correspondingAuthor":true,"prefix":"","firstName":"Kimmo","middleName":"","lastName":"Eriksson","suffix":""},{"id":528765738,"identity":"7d21b31f-2019-442c-b0de-cc460043dcfb","order_by":2,"name":"Chris G. Sibley","email":"","orcid":"","institution":"University of Auckland","correspondingAuthor":false,"prefix":"","firstName":"Chris","middleName":"G.","lastName":"Sibley","suffix":""},{"id":528765739,"identity":"2e56d7c8-6d3b-46ca-800b-8cb49a75787e","order_by":3,"name":"Irina Vartanova","email":"","orcid":"","institution":"The Institute for Future Studies","correspondingAuthor":false,"prefix":"","firstName":"Irina","middleName":"","lastName":"Vartanova","suffix":""},{"id":528765740,"identity":"8116ad30-dbeb-42c9-bea1-d90c926c6a10","order_by":4,"name":"Pontus Strimling","email":"","orcid":"","institution":"The Institute for Future Studies","correspondingAuthor":false,"prefix":"","firstName":"Pontus","middleName":"","lastName":"Strimling","suffix":""}],"badges":[],"createdAt":"2025-09-19 13:23:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7658969/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7658969/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":93457348,"identity":"d75bfe9c-6ada-4c7d-b9f6-bf3d9756a64f","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":847232,"visible":true,"origin":"","legend":"","description":"","filename":"AnonymizedManuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/4876055aba61d3c6befb0be3.docx"},{"id":93459106,"identity":"fbc03879-ae79-4943-a330-19b0652005fa","added_by":"auto","created_at":"2025-10-14 05:54:03","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6969,"visible":true,"origin":"","legend":"","description":"","filename":"ced66effea9c4d059276b1604aa18482.json","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/52137523cd558393efa003c9.json"},{"id":93458363,"identity":"46be316a-a069-42a3-b3d2-f84ab6cfb98e","added_by":"auto","created_at":"2025-10-14 05:46:03","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42067,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource.docx","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/b9a2b7d714ee734d4a41dd08.docx"},{"id":93457351,"identity":"ff4d334c-4b15-4d15-9bc1-08ade25e7159","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":89381,"visible":true,"origin":"","legend":"","description":"","filename":"ced66effea9c4d059276b1604aa184821enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/57e70dfdc10fb41759cc3b76.xml"},{"id":93458361,"identity":"89dd5088-6f5d-486e-8321-7b82ef7589c8","added_by":"auto","created_at":"2025-10-14 05:46:03","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":86184,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/859c40f4eceb2da411fdb4a1.png"},{"id":93458366,"identity":"c8650a2f-a425-4758-9128-96d3801850b4","added_by":"auto","created_at":"2025-10-14 05:46:03","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":336606,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/36ce65596e31f2798aa4802c.png"},{"id":93457354,"identity":"a3920d78-23d5-4254-8cad-40f1b2c1a748","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53604,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/5c11f25a0b7c603e93ea8580.png"},{"id":93458369,"identity":"4e7e6601-7ba6-4d68-be09-1ec48c32ddb1","added_by":"auto","created_at":"2025-10-14 05:46:03","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":157325,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/0ffe788f7dd82c1e504edef7.png"},{"id":93459111,"identity":"2345ac90-dad9-447f-8432-550a2a605005","added_by":"auto","created_at":"2025-10-14 05:54:04","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":170802,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/4e9c8b9a3ea63a2ff8101898.png"},{"id":93459105,"identity":"c7c91eb0-1521-4363-9797-bac0de7d2837","added_by":"auto","created_at":"2025-10-14 05:54:03","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36974,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/7f5bf7cbc2cff75f093f8744.png"},{"id":93457357,"identity":"a45ea9ba-af91-48b4-9784-63dff17c95e5","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":76162,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/078d95318b45f320f062cdf5.png"},{"id":93458365,"identity":"49862be7-4eb6-484d-8eb7-511dfa252778","added_by":"auto","created_at":"2025-10-14 05:46:03","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":22290,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/7fbfbce9ec286fd05712b481.png"},{"id":93457363,"identity":"5e4713cc-39f8-4ad3-885b-1d231e528c2a","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":81655,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/7ea119f26e9aeb398e4fa452.png"},{"id":93459109,"identity":"b9c3d4aa-9d42-426d-bcdb-dc495cd9eb69","added_by":"auto","created_at":"2025-10-14 05:54:03","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":71775,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/dbe8e17621563498fdce485d.png"},{"id":93457364,"identity":"2fef8bd3-90a4-48ba-889f-fb0e465057a1","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"xml","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":86852,"visible":true,"origin":"","legend":"","description":"","filename":"ced66effea9c4d059276b1604aa184821structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/b3bed18279cc7eedca396ed4.xml"},{"id":93457361,"identity":"82631543-c6b3-474d-a6e2-1d08d32f09a4","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"html","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":95963,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/b3a45032974e3d54d19639cb.html"},{"id":93457342,"identity":"f30fcfc6-efa7-4c72-b3f7-f236f594f491","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":43052,"visible":true,"origin":"","legend":"\u003cp\u003eA key proposition from Moral Argument Theory.\u003cstrong\u003e \u003c/strong\u003eThe model illustrates that the fit between a moral attitude and an individual's values is a product of two factors: the attitude’s connection to individualizing arguments and the liberalness of the individual’s moral values.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/50f33da7e98de4025b2a7825.png"},{"id":93457343,"identity":"1d40f481-79ab-4301-8d8d-e7cc2bf95f27","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":85077,"visible":true,"origin":"","legend":"\u003cp\u003eTheoretical predictions for the moderation of the education-attitudes link.\u003cstrong\u003e \u003c/strong\u003e(A) Following MFT, we assume that the preference for individualizing moral arguments is strong among liberals, weaker among moderates, and non-existent among strong conservatives. (B) The association between education and liberal moral attitudes is assumed to stem from the preference for individualizing moral arguments, and is therefore predicted to be strong among liberals, weaker among moderates, and non-existent among strong conservatives (Hypothesis 1). (C) The sign and strength of the association between education and a specific moral attitude is predicted to depend on its individualizing argument advantage (Hypothesis 2).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/584b8befcd4d2fc79a308d03.png"},{"id":93457344,"identity":"37bee467-a021-4245-b594-73319253edd4","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":35861,"visible":true,"origin":"","legend":"\u003cp\u003eThe link between education and liberal attitudes is negligible among very conservative participants and increases steadily with political liberalism.\u003cstrong\u003e \u003c/strong\u003eThis graph shows the average strength of the link between education and liberal attitudes. Link strength is estimated as the standardized coefficient from regressions of liberal attitudes on education, stratified by ideology, and represents the change in log-odds associated with a one-standard-deviation increase in education. Error bars indicate 95% confidence intervals.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/1e0b760e533ece32e1fb3e82.png"},{"id":93457347,"identity":"f8614136-90eb-4899-b5ec-a2324013cc98","added_by":"auto","created_at":"2025-10-14 05:38:03","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":101825,"visible":true,"origin":"","legend":"\u003cp\u003eThe three-way interaction between education, ideology, and argument advantage. Each panel plots the strength of the education-attitude link (y-axis) against the attitude's individualizing argument advantage (x-axis) for a given ideological group. Link strength is estimated as the standardized coefficient from regressions of attitudes on education, and thus represents the change in log-odds associated with a one-standard-deviation increase in education. As predicted by the value-argument alignment model, the association between education and attitudes is strongly dependent on moral arguments for liberals, but this dependency vanishes for conservatives.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/e71f5808ec1ccd476ff12a04.png"},{"id":93459825,"identity":"6e3ec822-ea5e-4c40-b2fb-63be73262069","added_by":"auto","created_at":"2025-10-14 06:02:03","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":105497,"visible":true,"origin":"","legend":"\u003cp\u003ePreregistered replication of the theoretical model in New Zealand data. Panel A confirms a core theoretical assumption, showing that preference for individualizing moral concerns increases with ideological liberalism. Panel B replicates the finding from Study 1 (H1), showing that the positive link between education and liberal attitudes (measured by standardized coefficients averaged across attitudes) is strongest for liberals and vanishes for conservatives. Panel C replicates the finding from Study 2 (H2), showing that the education-attitude link is highly sensitive to an issue's argument advantage for liberals, while this sensitivity disappears for conservatives. In Panels B and C, link strength is estimated as standardized coefficients, representing the change in log-odds associated with a one-standard-deviation increase in education. Error bars indicate 95% confidence intervals. Note the similarity between this figure and the illustration of theoretical predictions in Fig. 2.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/7f814986bf3a3f51499d2835.png"},{"id":102964478,"identity":"b8b91d9b-d2fd-4c1d-82f9-3612eea6e0c4","added_by":"auto","created_at":"2026-02-19 04:22:26","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1221622,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/270c5f20-b64a-4e69-aba4-8275462ce902.pdf"},{"id":93458359,"identity":"a8c3bdea-66cc-411a-b42b-b1b5a3e518b0","added_by":"auto","created_at":"2025-10-14 05:46:03","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":42067,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource.docx","url":"https://assets-eu.researchsquare.com/files/rs-7658969/v1/e6008ef5ab76ebbcbe10ee71.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"The Value Clarification Hypothesis: Explaining a Striking Asymmetry in the Link Between Education and Moral Attitudes","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSocial scientists have long observed a robust link between educational attainment and moral attitudes: individuals with higher levels of education tend to hold more liberal views on a wide range of moral issues (Hyman and Wright 1979; Lipset 1959; Weil 1985). This finding, documented across numerous countries and decades, has fueled a long-standing debate about its origins. Does higher education actively foster liberal attitudes through socialization and the development of critical thinking skills (Surridge 2016)? Or does the correlation reflect a selection effect, where individuals with pre-existing liberal attitudes are simply more likely to seek out advanced education (Campbell and Horowitz 2016)?\u003c/p\u003e\n\u003cp\u003eWhile this debate has been fruitful, it may overlook a more fundamental feature of the phenomenon. As the articles above illustrate, the link between education and liberalism is typically treated as an ideologically uniform effect. Yet it is not! Our analysis reveals that this relationship is profoundly asymmetric: education strongly predicts liberal attitudes among self-identified liberals but has virtually no effect among conservatives. This finding presents a puzzle that existing theories, focused on the general effects of education, are not well-equipped to explain.\u003c/p\u003e\n\u003cp\u003eTo resolve this puzzle, we introduce and test the Value Clarification Hypothesis, an explanation derived from Moral Argument Theory (Strimling et al. 2019). We propose that higher education functions less as a uniformly liberalizing force and more as a catalyst that helps individuals develop attitudes whose justifications are consistent with the individuals’ underlying values. For liberals, this process reinforces liberal positions. For conservatives, however, the net effect is negligible because their values are consistent with justifications for both sides of most moral issues (Graham et al. 2009; Strimling et al. 2019). This framework not only accounts for the ideological asymmetry but also generates a more nuanced prediction: the interaction should depend on the specific moral arguments available for each issue.\u003c/p\u003e\n\u003cp\u003eWe test these propositions across three studies, using large-scale survey data from two countries. Study 1 establishes the core asymmetric pattern in U.S. data. Study 2 tests our theoretical explanation by examining how the pattern varies across different types of moral issues. Study 3 provides a preregistered replication in New Zealand, confirming both the phenomenon and our explanation. Our findings offer a more precise account of whose attitudes are most influenced by education and on which issues. By moving beyond the simple selection-versus-socialization dichotomy, we provide new insights into how educational experiences shape the contours of contemporary political divisions.\u003c/p\u003e"},{"header":"Background: Explanations for the Education-Attitudes Link and the Role of Ideology","content":"\u003cp\u003eThe well-established link between higher education and liberal moral attitudes has been a central topic of inquiry in the social sciences for decades. Explanations for this phenomenon have historically fallen into two broad categories: those emphasizing selection effects and those proposing direct causal effects (Surridge 2016). The selection hypothesis posits that the correlation is largely spurious; individuals with liberal predispositions are simply more likely to seek out and complete higher education (Gross 2013). In contrast, theories of causal effects contend that the educational experience itself actively fosters liberal attitudes, either through cognitive development (Stubager 2008) or social influence (Surridge 2016; Werfhorst and Graaf 2004). While most research suggests both selection and causal mechanisms are at play (Simon 2022), the precise nature of the causal link remains a subject of intense debate.\u003c/p\u003e\n\u003ch2\u003eThe Missing Piece: Ideological Moderation\u003c/h2\u003e\n\u003cp\u003eTo further refine this inquiry, it is crucial to distinguish between broad political ideology and specific moral attitudes. Some researchers have sought to isolate the \u0026quot;pure\u0026quot; effect of education by including ideology as a statistical control (e.g., Weil 1985; Ma-Kellams 2014). However, treating ideology as a mere control variable implicitly assumes that the effect of education is uniform across the ideological spectrum. This approach fails to ask a potentially more important question: does the \u003cem\u003estrength\u003c/em\u003e of the relationship between education and attitudes \u003cem\u003edepend on\u003c/em\u003e one\u0026apos;s ideology?\u003c/p\u003e\n\u003cp\u003eThis oversight represents a significant gap in the literature. In Study 1, we directly explore this moderation and reveal a clear pattern: the link is strong among liberals but gradually weakens across the ideological spectrum until it becomes negligible for the most conservative groups. This finding presents a theoretical puzzle. Why would the influence of education be so potent for liberals yet so inert for conservatives? Existing theories do not offer a ready explanation.\u003c/p\u003e\n\u003cp\u003eTo explain this ideologically asymmetric link between education and moral attitudes, we draw on Moral Argument Theory (Eriksson and Strimling 2015), which builds upon Moral Foundations Theory (MFT; Graham, Haidt, and Nosek 2009).\u003c/p\u003e\n\u003ch2\u003eIdeology as a Proxy for Moral Values: The Moral Foundations Framework\u003c/h2\u003e\n\u003cp\u003eMFT posits that moral reasoning is guided by a set of intuitive foundations, most notably the \u0026quot;individualizing\u0026quot; foundations of Harm (preventing suffering) and Fairness (ensuring justice and reciprocity), and the \u0026quot;binding\u0026quot; foundations of Loyalty (fidelity to one\u0026apos;s group), Authority (respect for tradition and hierarchy), and Purity (avoiding contamination and degradation). A key, well-replicated insight is that liberals prioritize the individualizing foundations almost exclusively, while conservatives draw on all foundations more equally (Graham et al. 2011; Haidt and Graham 2007). Our model rests only on this robust empirical distinction, which we confirm in Study 3, while remaining agnostic about other, more contested aspects of MFT.\u003c/p\u003e\n\u003ch2\u003eFrom Values to Arguments for Specific Issue Positions: Moral Argument Theory\u003c/h2\u003e\n\u003cp\u003eMoral Argument Theory (MAT) extends this framework by focusing on the intrinsic properties of moral attitudes themselves. MAT posits that any stance on an issue is connected to a specific profile of moral arguments, and there is broad societal agreement on which arguments support which positions (Strimling et al. 2019; Vartanova et al. 2021). This allows us to measure an attitude\u0026apos;s \u003cem\u003eindividualizing argument advantage\u003c/em\u003e: the degree to which it is supported by arguments about harm and fairness relative to the opposing attitude. To illustrate the concept of argument advantage, consider same-sex marriage as an example. Arguments in favor are naturally and easily framed in terms of Fairness (\u0026quot;all couples deserve equal rights\u0026quot;) and Harm prevention (\u0026quot;denying marriage causes psychological and material harm to same-sex couples\u0026quot;). Arguments against same-sex marriage, by contrast, are more likely to draw on appeals to tradition (Authority: \u0026quot;marriage has always been between a man and woman\u0026quot;) or sanctity (Purity: \u0026quot;marriage is sacred and should not be redefined\u0026quot;). Thus, a positive attitude to same-sex marriage has a strong advantage with respect to individualizing arguments.\u003c/p\u003e\n\u003cp\u003eIf we know an attitude\u0026rsquo;s individualizing argument advantage, we can say how well it fits an individual\u0026rsquo;s values depending on how liberal those values are.\u0026nbsp;\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eFor strong liberals, who are assumed to value only individualizing arguments, the attitudes that are consistent with their values are those that have an individualizing argument advantage. The stronger an attitude\u0026rsquo;s individualizing argument advantage is, the more unambiguous is its fit with liberal values.\u003c/li\u003e\n \u003cli\u003eFor strong conservatives, who are assumed to value individualizing and binding arguments equally high, both attitudes on a given issue are consistent with those values. Thus, regardless of which position is supported by individualizing arguments and which position is supported by binding arguments, none of them fit conservative values better than the other one.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eFig. 1 illustrates this interaction between the attitude\u0026apos;s argumentative properties and the liberalness of the person\u0026rsquo;s values in determining the attitude-values fit.\u003c/p\u003e\n\u003ch2\u003eThe Value Clarification Hypothesis\u003c/h2\u003e\n\u003cp\u003eThis brings us to the core of our explanation, which we formalize as the Value Clarification Hypothesis. Prior research has shown that the values-attitudes fit is better among people with higher cognitive ability (Eriksson et al. 2023), consistent with higher cognitive ability providing better argument comprehension (Britt et al. 2016). Prior research has further demonstrated better values-attitudes fit among people who discuss moral issues more frequently (Eriksson et al. 2022a), consistent with greater exposure to arguments for both sides supporting the alignment process. The Value Clarification Hypothesis says that education catalyzes increased values-attitudes fit in three ways: First, by selecting students with higher cognitive ability. Second, by providing students training that further increases their ability to comprehend arguments. Third, by providing an environment where students are more likely to be exposed to discussion of issues, thereby giving them greater exposure to arguments for both sides of issues.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe effect of this alignment process is fundamentally asymmetric, due to the interaction between liberalness and individualizing arguments discussed above. For a strong liberal, the effect of value clarification is a stronger adoption of attitudes with individualizing argument advantage. For a strong conservative, however, no moral attitude has a better fit than the opposite attitude, so the net effect is expected to be negligible. This framework generates two clear hypotheses:\u003c/p\u003e\n\u003cul class=\"decimal_type\"\u003e\n \u003cli\u003e\u003cstrong\u003eHypothesis 1 (The Ideological Moderation Effect):\u003c/strong\u003e The positive association between education and holding liberal moral attitudes will be moderated by ideology, such that the association is strongest among liberals and weakest (or non-existent) among conservatives.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eHypothesis 2 (The Argument Advantage Effect):\u003c/strong\u003e The ideological moderation effect described in H1 will itself be conditional on the nature of the moral issue. The effect of education on liberals\u0026apos; attitudes will be strongest for issues with the highest individualizing argument advantage.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eFig. 2 illustrates these theoretical predictions and the underlying assumption that liberals have a strong preference for individualizing moral arguments whereas conservatives don\u0026rsquo;t. To illustrate that the ideology is a continuous spectrum, the figure also shows that moderates are expected to have results in-between liberals and conservatives. In the studies, a seven-step ideology scale is used, capturing the spectrum in an even more fine-grained way.\u003c/p\u003e"},{"header":"Study Design and Empirical Strategy","content":"\u003cp\u003eOur three studies are designed to test the presence of the complex patterns predicted by the Value Clarification Hypothesis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStudy 1\u003c/strong\u003e documents the core asymmetric pattern. Using U.S. General Social Survey data, we test whether the association between education and moral attitudes is moderated by political ideology (H1). This establishes the individual-level constraint that any adequate theory must explain.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStudy 2\u003c/strong\u003e identifies the issue-level pattern. Using the same dataset, we introduce the concept of \"individualizing argument advantage\" and test whether this property of moral issues moderates the education-ideology interaction (H2). This establishes a second, cross-level constraint on viable explanations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStudy 3\u003c/strong\u003e provides preregistered replication in New Zealand. Using the New Zealand Attitudes and Values Survey, we test both patterns to demonstrate their robustness across national contexts and confirm key theoretical assumptions.\u003c/p\u003e\n\u003cp\u003eWe argue that the Value Clarification Hypothesis provides a better explanation for the complex and surprising empirical patterns we uncover in these studies than any previous theory of the education attitudes link. This can be described as an \"inference to the best explanation\" approach (Lipton 2000).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe patterns the study is designed to establish are those illustrated in Fig. 2, panels B and C. The Value Clarification Hypothesis readily explains these patterns as follows: The education-attitudes link is driven by education helping people align their attitudes with their preferences for individualizing arguments. Therefore, the link will be strongest among strong liberals who have the strongest preference for individualizing arguments, while it will vanish among strong conservatives as they don’t have that preference. Moreover, this pattern will be moderated by the size of the individualizing argument advantage of the specific attitude.\u003c/p\u003e\n\u003cp\u003eNow consider how previously discussed alternative theories could explain the same patterns. One theory is direct socialization, that colleges teach their students liberal attitudes. Another theory is that adoption of liberal attitudes is facilitated by the cognitive development that higher education stimulates. Both these theories can explain why liberal attitudes are more common among the highly educated than in the population at large. However, they do not explain why attitudes do not differ between educated and uneducated conservatives.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eA third theory is that people self-select to higher education based on their politics. While this theory can explain why there are more liberals among the highly educated than in the population at large, it does not explain why attitudes differ between educated and uneducated liberals or why the education-attitudes link varies systematically across attitudes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOur model incorporates a more nuanced role for selection, distinguishing between selection on attitudes and selection on cognitive ability. While simple selection on prior attitudes is insufficient, selection on cognitive ability (Deary and Johnson 2010) plays a contributing role. Because individuals with higher cognitive ability are more likely to attend university, the student population is pre-selected for stronger argument comprehension skills. This means the value clarification process has likely already begun before university. Higher education then acts as an accelerator for this process. Our model sees cognitive selection and educational socialization not as competing explanations, but as intertwined forces.\u003c/p\u003e"},{"header":"Study 1: The Asymmetric Link Between Education and Attitudes in the U.S.","content":"\u003cp\u003eStudy 1 tests our core proposition: that the link between education and liberal moral attitudes is moderated by political ideology. We hypothesized that the association would be strongest among liberals and weakest among conservatives (H1).\u003c/p\u003e\n\u003ch2\u003eMethod\u003c/h2\u003e\n\u003cp\u003e\u003cstrong\u003eData and Participants.\u003c/strong\u003e We analyzed data from the General Social Survey (GSS), a nationally representative, biennial survey of the U.S. population covering the period from 1972 to 2021 (Davern et al. 2024).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMeasures.\u003c/strong\u003e We measured moral attitudes, education, and political ideology as follows.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eMoral Attitudes.\u003c/em\u003e We used a set of 98 moral attitude items from the GSS, following prior research (Eriksson et al. 2022b). To facilitate analysis, all items were first dichotomized by removing neutral options and grouping graded responses. We then empirically determined the liberal stance for each issue by examining its correlation with participants\u0026apos; political ideology so that the more liberal position was consistently coded as 1 and the more conservative position as 0.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eEducation.\u003c/em\u003e Participants\u0026apos; educational attainment was measured by the total number of years of schooling completed, ranging from 0 to 20.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePolitical Ideology.\u003c/em\u003e Participants\u0026apos; political ideology was assessed with the item \u0026quot;polviews,\u0026quot; a 7-point scale. To aid interpretation, we recoded this variable so that it ranged from 1 (Extremely conservative) to 7 (Extremely liberal).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAnalysis Plan.\u003c/strong\u003e To test for moderation, we ran a separate logistic regression for each of the 98 moral attitude items. In each model, the dichotomous attitude was the outcome variable, and the predictors were the participant\u0026apos;s education level, their political ideology, and the interaction term between education and ideology. To test Hypothesis 1, we examined the sign and significance of the interaction term for each moral issue. With ideology coded such that higher values indicate greater liberalism, we expected a positive coefficient, indicating that the positive effect of education on holding a liberal attitude is amplified as ideology becomes more liberal.\u003c/p\u003e\n\u003ch2\u003eResults\u003c/h2\u003e\n\u003cp\u003eThe results provided strong support for Hypothesis 1. Across the 98 moral issues, the interaction term between education and ideology was in the expected positive direction for almost all items: 95 of 98, or 96.9% (95% CI from 91.3% to 99.4%, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001).\u003c/p\u003e\n\u003cp\u003eTo visualize this consistent pattern, we estimated the strength of the education-attitude link separately for each of the seven ideological groups. As shown in Fig. 3, the association between education and holding liberal attitudes was close to zero among the most conservative respondents, and increased in strength in a stepwise fashion across the ideological spectrum. The strongest link between education and liberal attitudes was observed among the most liberal respondents.\u003c/p\u003e\n\u003ch2\u003eDiscussion\u003c/h2\u003e\n\u003cp\u003eThe findings from Study 1 establish a clear and robust empirical pattern: the well-documented link between education and liberal moral attitudes is profoundly asymmetric. The association is robust in explaining attitudes among liberals, but has virtually no predictive power among conservatives. This asymmetry constitutes the theoretical puzzle that the remainder of this paper seeks to explain. Having established this phenomenon, we now turn to Study 2 to test our proposed theoretical explanation.\u003c/p\u003e"},{"header":"Study 2: Explaining the Asymmetry with Moral Arguments","content":"\u003cp\u003eStudy 1 established that the link between education and liberal attitudes is strongest for liberals. Study 2 tests our theoretical explanation for this phenomenon: that this interaction is itself driven by the moral content of the issues. Specifically, we test Hypothesis 2, which predicts that the moderating effect of ideology will be most pronounced for issues where the liberal position has a strong individualizing argument advantage.\u003c/p\u003e\n\u003ch2\u003eMethod\u003c/h2\u003e\n\u003cp\u003e\u003cstrong\u003eData and Participants.\u003c/strong\u003e The analysis for Study 2 used the same GSS dataset as Study 1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMeasures.\u003c/strong\u003e The variables \u003cem\u003eEducation\u003c/em\u003e and \u003cem\u003ePolitical Ideology\u003c/em\u003e were measured and coded identically to Study 1.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eMoral Attitudes\u003c/em\u003e. As in Study 1, all responses were dichotomized by excluding neutral options and collapsing graded responses. However, in this study, we retained the original coding with 1 indicating \u0026ldquo;agree\u0026rdquo; and 0 indicating \u0026ldquo;disagree,\u0026rdquo; because the argument advantage variable (described below) captured whether an opinion was liberal or conservative.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eIndividualizing Argument Advantage.\u003c/em\u003e This new variable was measured at the issue level. We used data from a prior study in which participants evaluated the arguments for and against each of the 98 moral attitudes (Vartanova et al., 2021). For each attitude, we calculated the proportion of respondents who endorsed Harm- and Fairness-based arguments for that position, and subtracted the proportion who endorsed those same arguments for the opposing position. This yielded a continuous score for each of the 98 issues, where a high positive score indicates a strong individualizing argument advantage for the liberal attitude. (See the Online Resource for the full list of items and their advantage scores.)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAnalysis Plan.\u003c/strong\u003e To test Hypothesis 2, we used a multilevel logistic regression model that included all participants and all moral issues simultaneously. This approach accounts for the nested structure of the data (responses nested within individuals, and individuals nested within issues). The outcome variable was the dichotomous moral attitude. The model included fixed effects for education, ideology, the individualizing argument advantage of the issue, all two-way interactions, and the crucial three-way interaction term (Education \u0026times; Ideology \u0026times; Argument Advantage). Our primary test of Hypothesis 2 was the sign and significance of this three-way interaction term. We predicted a positive coefficient, indicating that the two-way interaction between education and ideology (found in Study 1) becomes stronger as the argument advantage of the liberal attitude increases.\u003c/p\u003e\n\u003ch2\u003eResults\u003c/h2\u003e\n\u003cp\u003eThe multilevel model revealed a significant and positive three-way interaction between education, ideology, and individualizing argument advantage, providing strong support for Hypothesis 2 (for further details, see Table S1 in the Online Resource).\u003c/p\u003e\n\u003cp\u003eTo illustrate this complex interaction, we plotted the strength of the education-attitude link as a function of the argument advantage score, separately for each of the seven ideological groups (see Fig. 4). The results show a striking pattern. For \u0026quot;Extremely liberal\u0026quot; individuals, there is a steep, positive relationship: the more an attitude is supported by individualizing arguments, the stronger its association with education. This relationship becomes progressively flatter as we move across the ideological spectrum. For \u0026quot;Extremely conservative\u0026quot; individuals, the line is almost perfectly flat, indicating that the link between their education and their attitudes is entirely insensitive to the argument advantage of the issue.\u003c/p\u003e\n\u003ch2\u003eDiscussion\u003c/h2\u003e\n\u003cp\u003eThe findings from Study 2 provide strong support for our theoretical model. The data confirm that the asymmetric effect of education is not random but is systematically related to the moral content of the issues themselves. Education appears to amplify the process of value-argument alignment, but this only occurs for individuals whose values are aligned with the arguments that define the liberal side of the debate (liberals). For conservatives, whose values create more complex considerations, this alignment process does not produce a clear directional effect. Having established our proposed mechanism in this exploratory analysis, we turn to Study 3 for a preregistered replication.\u003c/p\u003e"},{"header":"Study 3: Preregistered Replication in New Zealand","content":"\u003cp\u003eStudy 3 provides a preregistered replication of the key findings from Studies 1 and 2 in a different national context. This provides a robust test of the generalizability of both the ideological asymmetry (H1) and our proposed explanation for it (H2). Additionally, this study provides a direct test of a core assumption of our theory: that liberal ideology serves as a proxy for a strong preference for individualizing moral arguments.\u003c/p\u003e\n\u003ch2\u003eMethod\u003c/h2\u003e\n\u003cp\u003e\u003cstrong\u003eData and Participants.\u003c/strong\u003e We analyzed data from the New Zealand Attitudes and Values Survey (NZAVS), a national longitudinal panel study that has collected data annually since 2009 (Sibley 2023). Our analysis includes all 14 waves of data available at the time of the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreregistration.\u003c/strong\u003e The hypotheses, measures, and analysis plan for this study were preregistered prior to data analysis. Due to space constraints on the preregistration platform, the study was preregistered in multiple documents: https://aspredicted.org/1JV_PK9 (specifies the selection of moral issues in the NZAVS, as well as the procedures for gathering and calculating argument advantages), https://aspredicted.org/FW3_2R5 (specifies certain preliminary analyses), and https://aspredicted.org/Y3X_KL5 (specifies the analyses for all hypotheses about the education-attitudes link, referred to as H2a, H2b, and H2c in the preregistration).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAnalytic deviations from preregistration.\u0026nbsp;\u003c/strong\u003eFor consistency with Hypothesis 2 in the GSS data, we retained the original coding of dichotomized attitude variables in the NZAVS rather than reverse-coding when the argument advantage was negative. This decision does not affect H1, and H2 remains supported when analyzed according to the preregistered plan (see the Online Resource). In addition, inspection of model output revealed that the preregistered multilevel logistic regression models produced non-normal random intercepts for several issues, with distributions that were highly skewed or bimodal. As a robustness check, we re-estimated each hypothesis using models that do not assume normally distributed random intercepts. Results were consistent across specifications (see the Online Resource).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMeasures.\u003c/strong\u003e In addition to the variables in Study 2, this study also includes respondents\u0026apos; preference for individualizing arguments.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eMoral Attitudes.\u003c/em\u003e We identified 36 moral attitude items from the NZAVS that met the inclusion criteria. Of these, 29 exhibited a significant argument advantage and were retained for analysis, as specified in the preregistration. All items were dichotomized by removing neutral responses and collapsing graded options into binary categories (1 = agree, 0 = disagree). In line with the preregistered plan, the liberal position on each issue was determined by its individualizing argument advantage: if the advantage was negative, the liberal stance was coded as \u0026ldquo;disagree\u0026rdquo;; if positive, as \u0026ldquo;agree.\u0026rdquo; Notably, this approach produced results highly consistent with the ideology\u0026ndash;attitudes association used in Study 1, with 26 of the 29 issues yielding the same directional sign.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eEducation.\u003c/em\u003e Educational attainment was measured on an 11-point scale, ranging from 0 (no qualification) to 10 (doctorate degree).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePolitical Ideology.\u003c/em\u003e In the NZAVS, political ideology was measured with a 7-point item ranging from \u003cem\u003eExtremely liberal\u003c/em\u003e (1) to \u003cem\u003eExtremely conservative\u003c/em\u003e (7), with a neutral midpoint at 4. Because this item has been included in multiple survey waves, we calculated each participant\u0026rsquo;s mean response across all available waves. We classified respondents as liberal if their mean score was \u0026le; 3 and as conservative if their mean score was \u0026ge; 5. The binary scale was used in preregistered analyses, whereas figures displayed the full 7-point scale for consistency with Studies 1 and 2.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eIndividualizing Argument Advantage.\u003c/em\u003e Argument advantage scores for the 36 NZAVS items were collected in a new survey of 537 New Zealand participants, following the same procedure used for the GSS items. (See the Online Resource for a detailed description of this data collection and the full list of items and their advantage scores).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePreference for Individualizing Arguments.\u003c/em\u003e In one wave of the survey, participants completed the 30-item Moral Foundations Questionnaire (MFQ; Graham et al. 2011). From this, we calculated each participant\u0026apos;s preference for individualizing concerns by subtracting their average score on the binding foundations (Loyalty, Authority, Purity) from their average score on the individualizing foundations (Harm, Fairness).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAnalysis Plan.\u003c/strong\u003e First, to test our theoretical assumption, we examined the correlation between participants\u0026apos; political ideology and their preference for individualizing arguments. Second, to replicate Study 1, we tested for the Education \u0026times; Ideology interaction (H1). Third, to replicate Study 2, we tested for the Education \u0026times; Ideology \u0026times; Argument Advantage interaction (H2). As a robustness check of H2, we also ran a linear regression of the interaction effect obtained in H1 on the individualizing argument advantage to assess whether the advantage positively predicted the size of the interaction, as we expected in the preregistration. Due to the panel nature of the NZAVS data, all models were multilevel, accounting for repeated observations nested within individuals.\u003c/p\u003e\n\u003ch2\u003eResults\u003c/h2\u003e\n\u003cp\u003eThe data provided strong support for our core theoretical assumption of a positive relationship between political liberalism and preference for individualizing arguments (\u003cem\u003er\u003c/em\u003e = .54, 95% CI [.51, .56], \u003cem\u003ep\u003c/em\u003e \u0026lt; .001). As shown in Fig. 5A, the preference for individualizing arguments was captured by ideological liberalism: while no significant preference was found among the most conservative respondents, the preference increased stepwise with greater liberalism, becoming very strong among the most liberal respondents.\u003c/p\u003e\n\u003cp\u003eSecond, the results replicated the findings from Study 1, providing further support for H1. Across the 29 moral issues, the interaction term between education and ideology was in the expected direction for 26 of 29 issues (89.6%, 95% CI from 72.6% to 97.8%, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001). As shown in Fig. 5B, the positive association between education and liberal attitudes was strongest among the most liberal New Zealanders and disappeared among the most conservative.\u003c/p\u003e\n\u003cp\u003eFinally, we successfully replicated the findings from Study 2. The three-way interaction between education, ideology, and argument advantage was positive and significant, supporting H2 (see Table S1 in the Online Resource). In addition, the robustness check confirmed that argument advantages were positively associated with the Education \u0026times; Ideology interaction obtained in H1 (\u0026beta; = .376, 95% CI [.240, .511], p \u0026lt; .001). As shown in Fig. 5C, the relationship between education and attitudes was highly sensitive to the argument advantage of the issue for liberals, but this sensitivity diminished and vanished for conservatives.\u003c/p\u003e\n\u003ch2\u003eDiscussion\u003c/h2\u003e\n\u003cp\u003eThe results of Study 3 provide strong, preregistered support for our theoretical account. We successfully replicated the core findings from the U.S. data in a different Western democracy, demonstrating the robustness of both the asymmetric link between education and attitudes and its moderation by the moral content of the issues. Furthermore, we confirmed a key assumption of our model: that the liberal-conservative ideological dimension maps directly onto the preference for individualizing moral foundations. This confirmatory evidence provides a solid foundation for the generalizability of our theory.\u003c/p\u003e"},{"header":"General Discussion","content":"\u003cp\u003eFor decades, social scientists have sought to understand why higher education is associated with more liberal moral attitudes. Across three studies, we identified a novel empirical phenomenon—that the education-liberal attitudes link is profoundly asymmetric—and provided a theoretical model to account for it. Our findings in the United States, replicated in a preregistered study in New Zealand, demonstrate that the effect of education is conditional not only on who is being educated but also on the moral arguments surrounding an issue.\u003c/p\u003e\n\u003ch2\u003eRevisiting the Education-Attitudes Link: The Value Clarification Hypothesis\u003c/h2\u003e\n\u003cp\u003eOur central theoretical contribution is the Value Clarification Hypothesis, which posits that education enhances individuals' ability to connect their underlying moral values to specific attitudes through a process of value-argument alignment. The striking asymmetry arises because the moral value systems of liberals and conservatives are structured differently. This explanation was supported by our finding that the effects were strongest for issues with a clear individualizing argument advantage, and that ideological liberalism itself maps strongly onto a preference for these same arguments. Rather than operating as a uniform liberalizing force, higher education appears to function as a value-clarifying mechanism.\u003c/p\u003e\n\u003cp\u003eThe empirical patterns we document present significant challenges to existing theoretical accounts. The strength of the Value Clarification Hypothesis lies not only in fitting these data, but in generating specific, cross-level predictions that were confirmed across multiple national contexts through preregistered replication. While we cannot claim definitive causal proof, the convergent evidence provides substantial support for our proposed mechanism over existing alternatives.\u003c/p\u003e\n\u003ch2\u003eImplications for Understanding Political Bias in Higher Education\u003c/h2\u003e\n\u003cp\u003eThese findings reframe the public debate about ideological bias in academia. The fact that the education-attitudes link was essentially zero for strong conservatives suggests that education is not a monolithic liberalizing force. This suggests the link is driven less by indoctrination and more by a process of intellectual exploration that allows students to build more coherent belief systems based on their pre-existing values.\u003c/p\u003e\n\u003ch2\u003eBroader Implications for Political Socialization\u003c/h2\u003e\n\u003cp\u003eOur research contributes to broader understanding of how education shapes political belief systems. The findings help explain previously observed inconsistencies in the education-attitudes link across different contexts. For example, the finding that the link is weaker in some countries (Weakliem 2002; Weil 1985) may be partly explained by those countries having different ideological compositions in their populations. Our model suggests that future research should account for the ideological distribution of samples when estimating the effects of education.\u003c/p\u003e\n\u003cp\u003eThe pattern also has implications for understanding political polarization. If higher education primarily serves to clarify and strengthen existing ideological commitments rather than fostering convergence toward liberal positions, this could contribute to the increasing polarization observed in many democracies. Educated liberals and conservatives may emerge from university with more crystallized and internally consistent—but divergent—worldviews.\u003c/p\u003e\n\u003ch2\u003eLimitations and Future Directions\u003c/h2\u003e\n\u003cp\u003eOur research has several important limitations. The fact that we obtained similar findings in the United States and New Zealand suggest some degree of cross-cultural generalizability. However, as both societies are Western democracies, replication in culturally more distant contexts is required to answer how far the findings can be generalized.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAnother limitation is that we did not measure whether education in fact enhances argument exposure and comprehension, as we assume it does. While beyond the scope of this study, this could be examined in future research.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThe link between education and moral attitudes is more complex and systematic than previously understood. It is not a uniform liberalizing tide that lifts all boats. Instead, it is a highly specific current, powerful for those with a liberal compass and nearly still for those with a conservative one. By identifying this asymmetry and providing a robust theoretical explanation for it, this research advances our understanding of political socialization and offers a new lens through which to view the enduring relationship between what we know and what we believe.\u003c/p\u003e\n\u003cp\u003eIn an era of increasing concern about political polarization and the role of educational institutions in democratic societies, these findings suggest that universities may be less politically transformative than either critics or advocates claim. Rather than systematically converting students to liberal positions, higher education appears to help individuals—particularly those with liberal inclinations—develop more coherent belief systems aligned with their pre-existing values.\u003c/p\u003e\n\u003cp\u003eThis nuanced understanding has important implications for ongoing debates about academic freedom and intellectual diversity in higher education. Our findings suggest that concerns about uniform liberal indoctrination may be overstated, while also indicating that educational experiences do have differential effects across the political spectrum. The key insight is that education functions less as a liberalizing force and more as a value-clarifying mechanism that helps students connect their values to specific positions through enhanced exposure to and comprehension of moral arguments.\u003c/p\u003e\n\u003cp\u003eBy identifying these previously unrecognized patterns and providing a theoretical framework to explain them, this research offers new directions for understanding how educational experiences shape political belief systems. While we do not claim definitive proof of our proposed mechanism, the empirical patterns we document present clear challenges to existing theories and point toward more sophisticated models of educational influence on political attitudes.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eData from the General Social Survey (GSS; http://doi.org/10.5255/UKDA-Series-200006) are publicly available. Data from the New Zealand Attitudes and Values Study (NZAVS) are not publicly available but can be requested from the study administrators. Argument advantage scores for all items included in the analyses are provided in the\u0026nbsp;Online Resource.\u003c/p\u003e\n\u003ch2\u003eEthical Statement\u003c/h2\u003e\n\u003cp\u003eThis study relied on secondary analysis of publicly available, anonymized survey data from the General Social Survey (GSS) and data collected under ethical oversight from the New Zealand Attitudes and Values Study (NZAVS). No additional ethics approval was required.\u003c/p\u003e\n\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eThis work was supported by the Knut and Alice Wallenberg Foundation [grant number 2022.0191] and the Swedish Research Council [grant number\u0026nbsp;2021–01654].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors have no acknowledgements to declare.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBritt, M. Anne, Kristopher J. Kopp, Amanda M. Durik, Dylan Blaum, and Peter Hastings. 2016. \u0026ldquo;Identifying General Cognitive Abilities Involved in Argument Comprehension and Evaluation.\u0026rdquo; \u003cem\u003eZeitschrift F\u0026uuml;r P\u0026auml;dagogische Psychologie / German Journal of Educational Psychology\u003c/em\u003e 30(2\u0026ndash;3):79\u0026ndash;95. doi: 10.1024/1010-0652/a000173.\u003c/li\u003e\n\u003cli\u003eCampbell, Colin, and Jonathan Horowitz. 2016. \u0026ldquo;Does College Influence Sociopolitical Attitudes?\u0026rdquo; \u003cem\u003eSociology of Education\u003c/em\u003e 89(1):40\u0026ndash;58. doi: 10.1177/0038040715617224.\u003c/li\u003e\n\u003cli\u003eDavern, Michael, Rene Bautista, Jeremy Freese, Pamela Herd, and Stephen L. Morgan. 2024. \u0026ldquo;General Social Survey 1972-2024. [Machine-Readable Data File]. Principal Investigator, Michael Davern; Co-Principal Investigators, Rene Bautista, Jeremy Freese, Pamela Herd, and Stephen L. Morgan. Sponsored by National Science Foundation. NORC Ed. Chicago: NORC, 2024: NORC at the University of Chicago [Producer and Distributor].\u0026rdquo;\u003c/li\u003e\n\u003cli\u003eDeary, Ian J., and Wendy Johnson. 2010. \u0026ldquo;Intelligence and Education: Causal Perceptions Drive Analytic Processes and Therefore Conclusions.\u0026rdquo; \u003cem\u003eInternational Journal of Epidemiology\u003c/em\u003e 39(5):1362\u0026ndash;69. doi: 10.1093/ije/dyq072.\u003c/li\u003e\n\u003cli\u003eEriksson, Kimmo, and Pontus Strimling. 2015. \u0026ldquo;Group Differences in Broadness of Values May Drive Dynamics of Public Opinion on Moral Issues.\u0026rdquo; \u003cem\u003eMathematical Social Sciences\u003c/em\u003e 77:1\u0026ndash;8. doi: 10.1016/j.mathsocsci.2015.06.004.\u003c/li\u003e\n\u003cli\u003eEriksson, Kimmo, Irina Vartanova, Isabela Hazin, and Pontus Strimling. 2023. \u0026ldquo;Cognitive Ability and Ideology Join Forces in the Culture War: A Model of Opinion Formation\u0026rdquo; edited by J. Druckman. \u003cem\u003ePNAS Nexus\u003c/em\u003e 2(6):pgad205. doi: 10.1093/pnasnexus/pgad205.\u003c/li\u003e\n\u003cli\u003eEriksson, Kimmo, Irina Vartanova, and Pontus Strimling. 2022a. \u0026ldquo;How Does Political Discussion Frequency Impact Political Moral Opinions? The Moral Argument Theory of Opinion Dynamics.\u0026rdquo; \u003cem\u003eFrontiers in Psychology\u003c/em\u003e 13.\u003c/li\u003e\n\u003cli\u003eEriksson, Kimmo, Irina Vartanova, and Pontus Strimling. 2022b. \u0026ldquo;Opinion Trends on Moral Issues in the United States and the United Kingdom Explained by the Applicability of Generally Accepted Arguments.\u0026rdquo; \u003cem\u003eInternational Journal of Public Opinion Research\u003c/em\u003e 34(1):edac001. doi: 10.1093/ijpor/edac001.\u003c/li\u003e\n\u003cli\u003eFitzmaurice, Garrett M., Nan M. Laird, and James H. Ware. 2011. \u003cem\u003eApplied Longitudinal Analysis\u003c/em\u003e. 2nd edition. Hoboken, New Jersey: Wiley.\u003c/li\u003e\n\u003cli\u003eGraham, Jesse, Jonathan Haidt, and Brian A. Nosek. 2009. \u0026ldquo;Liberals and Conservatives Rely on Different Sets of Moral Foundations.\u0026rdquo; \u003cem\u003eJournal of Personality and Social Psychology\u003c/em\u003e 96(5):1029\u0026ndash;46. doi: 10.1037/a0015141.\u003c/li\u003e\n\u003cli\u003eGraham, Jesse, Brian Nosek, Jonathan Haidt, Ravi Iyer, Sena Koleva, and Peter Ditto. 2011. \u0026ldquo;Mapping the Moral Domain.\u0026rdquo; \u003cem\u003eJournal of Personality and Social Psychology\u003c/em\u003e 101:366\u0026ndash;85. doi: 10.1037/a0021847.\u003c/li\u003e\n\u003cli\u003eGross, Neil. 2013. \u003cem\u003eWhy Are Professors Liberal and Why Do Conservatives Care?\u003c/em\u003e Harvard University Press.\u003c/li\u003e\n\u003cli\u003eHaidt, Jonathan, and Jesse Graham. 2007. \u0026ldquo;When Morality Opposes Justice: Conservatives Have Moral Intuitions That Liberals May Not Recognize.\u0026rdquo; \u003cem\u003eSocial Justice Research\u003c/em\u003e 20(1):98\u0026ndash;116. doi: 10.1007/s11211-007-0034-z.\u003c/li\u003e\n\u003cli\u003eHyman, Herbert Hiram, and Charles Robert Wright. 1979. \u003cem\u003eEducation\u0026rsquo;s Lasting Influence on Values\u003c/em\u003e. University of Chicago Press.\u003c/li\u003e\n\u003cli\u003eLipton, Peter (2000). Inference to the Best Explanation. In W. Newton-Smith, A companion to the philosophy of science. Malden, Mass.: Blackwell. pp. 184\u0026ndash;193.\u003c/li\u003e\n\u003cli\u003eLipset, Seymour Martin. 1959. \u0026ldquo;Some Social Requisites of Democracy: Economic Development and Political Legitimacy.\u0026rdquo; \u003cem\u003eAmerican Political Science Review\u003c/em\u003e 53(1):69\u0026ndash;105. doi: 10.2307/1951731.\u003c/li\u003e\n\u003cli\u003eMa-Kellams, Christine, Aida Rocci Ruiz, Jacqueline Lee, and Andrea Madu. \u0026ldquo;Not All Education Is Equally Liberal: The Effects of Science Education on Political Attitudes.\u0026rdquo; Journal of Social and Political Psychology 2, no. 1 (2014): 1. https://doi.org/10.5964/jspp.v2i1.259.\u003c/li\u003e\n\u003cli\u003eSibley, C. G. 2023. \u0026ldquo;Sampling Procedure and Sample Details for the New Zealand Attitudes and Values Study.\u0026rdquo; doi: https://doi.org/10.31234/osf.io/wgqvy.\u003c/li\u003e\n\u003cli\u003eSimon, Elizabeth. 2022. \u0026ldquo;Demystifying the Link between Higher Education and Liberal Values: A within-Sibship Analysis of British Individuals\u0026rsquo; Attitudes from 1994\u0026ndash;2020.\u0026rdquo; \u003cem\u003eThe British Journal of Sociology\u003c/em\u003e 73(5):967\u0026ndash;84. doi: 10.1111/1468-4446.12972.\u003c/li\u003e\n\u003cli\u003eStrimling, Pontus, Irina Vartanova, Fredrik Jansson, and Kimmo Eriksson. 2019. \u0026ldquo;The Connection between Moral Positions and Moral Arguments Drives Opinion Change.\u0026rdquo; \u003cem\u003eNature Human Behaviour\u003c/em\u003e 3(9):922\u0026ndash;30. doi: 10.1038/s41562-019-0647-x.\u003c/li\u003e\n\u003cli\u003eStubager, Rune. 2008. \u0026ldquo;Education Effects on Authoritarian\u0026ndash;Libertarian Values: A Question of Socialization1.\u0026rdquo; \u003cem\u003eThe British Journal of Sociology\u003c/em\u003e 59(2):327\u0026ndash;50. doi: 10.1111/j.1468-4446.2008.00196.x.\u003c/li\u003e\n\u003cli\u003eSurridge, Paula. 2016. \u0026ldquo;Education and Liberalism: Pursuing the Link.\u0026rdquo; \u003cem\u003eOxford Review of Education\u003c/em\u003e 42(2):146\u0026ndash;64. doi: 10.1080/03054985.2016.1151408.\u003c/li\u003e\n\u003cli\u003eVartanova, Irina, Kimmo Eriksson, Isabela Hazin, and Pontus Strimling. 2021. \u0026ldquo;Different Populations Agree on Which Moral Arguments Underlie Which Opinions.\u0026rdquo; \u003cem\u003eFrontiers in Psychology\u003c/em\u003e 12.\u003c/li\u003e\n\u003cli\u003eWeakliem, D. L. 2002. \u0026ldquo;The Effects of Education on Political Opinions: An International Study.\u0026rdquo; \u003cem\u003eInternational Journal of Public Opinion Research\u003c/em\u003e 14(2):141\u0026ndash;57. doi: 10.1093/ijpor/14.2.141.\u003c/li\u003e\n\u003cli\u003eWeil, Frederick D. 1985. \u0026ldquo;The Variable Effects of Education on Liberal Attitudes: A Comparative- Historical Analysis of Anti-Semitism Using Public Opinion Survey Data.\u0026rdquo; \u003cem\u003eAmerican Sociological Review\u003c/em\u003e 50(4):458. doi: 10.2307/2095433.\u003c/li\u003e\n\u003cli\u003eWerfhorst, Herman G. Van de, and Nan Dirk de Graaf. 2004. \u0026ldquo;The Sources of Political Orientations in Post-Industrial Society: Social Class and Education Revisited.\u0026rdquo; \u003cem\u003eThe British Journal of Sociology\u003c/em\u003e 55(2):211\u0026ndash;35. doi: 10.1111/j.1468-4446.2004.00016.x.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"higher education, political attitudes, moral psychology, political socialization, ideology","lastPublishedDoi":"10.21203/rs.3.rs-7658969/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7658969/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSocial scientists have long documented a robust association between higher education and liberal moral attitudes. However, this relationship may be more complex than previously understood. Using large-scale survey data from the United States (General Social Survey), we demonstrate that the education-liberalism link is profoundly asymmetric across political ideology. Education strongly predicts liberal attitudes among self-identified liberals but has virtually no association with attitudes among strong conservatives. We explain this striking pattern by introducing and testing the Value Clarification Hypothesis, which proposes that education acts as a value-clarifying rather than uniformly liberalizing force. Higher education enhances individuals' ability to align their attitudes with their more fundamental moral values, but this process produces different outcomes depending on one's ideological starting point. For liberals, whose values center on preventing harm and promoting fairness, exposure to moral arguments consistently reinforces positions justified by such arguments. For conservatives, whose broader moral framework creates cross-cutting considerations, the net effect is negligible. A preregistered replication using similar survey data from New Zealand (the New Zealand Attitudes and Values Survey) confirms both the asymmetric pattern and our theoretical explanation. These findings reframe debates about political bias in higher education and offer new insights into how educational experiences shape political belief systems.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e","manuscriptTitle":"The Value Clarification Hypothesis: Explaining a Striking Asymmetry in the Link Between Education and Moral Attitudes","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-14 05:37:58","doi":"10.21203/rs.3.rs-7658969/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"6bc21e33-85c0-437b-9334-0a01b87b73cb","owner":[],"postedDate":"October 14th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-18T20:23:51+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-14 05:37:58","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7658969","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7658969","identity":"rs-7658969","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}