Generalized Trust and Immigration Attitudes: The Moderating Role of Deliberative Engagement. Evidence from Latin America

Technical Summary: Generalized Trust and Immigration Attitudes in Latin America

Problem Statement and Research Gap
While generalized trust is widely associated with favorable attitudes toward immigrants, the specific mechanisms through which this relationship operates remain underspecified. Existing literature often treats trust as an independent predictor or links it broadly to social capital and cohesion, yet it fails to clarify the cognitive and social conditions under which trust translates into tolerance. Furthermore, much of the existing research focuses on European contexts with higher baseline levels of trust, leaving a gap in understanding these dynamics in regions where generalized trust is scarce, such as Latin America. This paper addresses the question of how and under what circumstances generalized trust explains attitudes toward migration, specifically investigating whether deliberative practices moderate this effect.

Methodology and Data
The study employs multilevel ordered logistic regression models (cumulative link mixed models) using survey data from three waves of the Latinobarometer (2020, 2023, and 2024) across 18 Latin American countries. The total sample comprises 58,623 respondents.

• Dependent Variable: Attitudes toward migration, measured by a three-level ordered variable (harmful, neutral, beneficial) and an index derived from five specific domains (economy, labor market, security, culture, public burden).
• Independent Variable: Generalized trust, measured by the standard item asking if "most people" can be trusted.
• Moderator Variables: The study distinguishes between different forms of political engagement:
  • Deliberative engagement: Political discussion with friends and community engagement (working on issues affecting the community).
  • Non-deliberative/Directional engagement: Directional discussion (discussing politics to convince others), participation in political party work, and social media use.
• Control Variables: The models account for individual-level factors (economic satisfaction, institutional trust, ideology, education, unemployment, age, gender) and country-level contextual factors (GDP per capita, V-Dem polyarchy index, net migration rate).
• Estimation Strategy: The analysis utilizes multilevel modeling to account for respondents nested within countries. Robustness checks include Ordinary Least Squares (OLS) estimations and alternative dependent variable operationalizations.

Key Results
The empirical analysis yields several critical findings regarding the conditional nature of generalized trust:

1. Conditional Effect of Trust: In baseline models without interaction terms, generalized trust is a significant positive predictor of favorable migration attitudes. However, when interaction terms are introduced, the main effect of generalized trust becomes statistically nonsignificant. This indicates that trust does not independently shape attitudes; its effect is entirely conditional on the presence of deliberative practices.
2. Specificity of Deliberative Engagement: The interaction between generalized trust and political discussion with friends is positive and highly significant. Similarly, the interaction between generalized trust and community engagement is significant. This suggests that the "social intelligence" associated with trust is activated only when individuals engage in communication aimed at solving social problems or reasoning with diverse others.
3. Ineffectiveness of Other Forms of Engagement:
   • Directional Discussion: Discussing politics with the intent to persuade does not moderate the effect of trust.
   • Party Work: Participation in political party activities does not produce a moderating effect.
   • Social Media: Mere exposure to social media content does not moderate the relationship, suggesting that passive information consumption lacks the epistemic benefits of interpersonal deliberation.
4. Magnitude of Effects: The predicted probabilities demonstrate that for individuals who never discuss politics with friends, the influence of generalized trust on migration attitudes is negligible (a flat slope). Conversely, for those who discuss politics frequently, higher levels of generalized trust substantially reduce the probability of viewing migration as harmful (a 13-percentage-point reduction) and increase the probability of viewing it as beneficial (a 10-percentage-point increase).

Contributions and Significance
The paper makes three primary contributions to the literature:

1. Mechanism Specification: It moves beyond correlational findings to identify the specific mechanism linking trust and migration attitudes. The author argues that the mechanism is epistemic: deliberative practices reduce misperceptions and activate the "social intelligence" associated with generalized trust, enabling individuals to evaluate immigration more accurately.
2. Contextual Relevance: By focusing on Latin America, a region characterized by low levels of generalized trust, the study tests the argument under conditions where the variable is scarce, demonstrating that even in low-trust environments, the combination of trust and deliberation is crucial for fostering tolerance.
3. Differentiation of Engagement Types: The research refines the understanding of political participation by distinguishing between deliberative engagement (accuracy-oriented, problem-solving) and other forms of communication (directional, partisan, or passive). It establishes that not all social interaction fosters tolerance; only those forms that facilitate the exchange of diverse viewpoints and reasoning amplify the positive effects of trust.

Limitations and Claims
The author acknowledges that the cross-sectional design limits causal claims regarding the directionality between trust, deliberation, and attitudes. While the findings are consistent with theoretical expectations and prior experimental work, the paper does not claim definitive causality. Additionally, the proxy variables for deliberation (frequency of discussion) do not capture the quality or diversity of the exchanges. Nevertheless, the study concludes that generalized trust alone is insufficient to foster positive attitudes toward migration; it requires the activation context of deliberative engagement to translate into social tolerance.