Trust and needles: how perceptions of inequality shape vaccination in South Korea

The study examines how individuals’ subjective perceptions of economic inequality shape trust in science and society and, in turn, influence vaccine-related attitudes and behaviors in South Korea. Despite high national COVID-19 vaccination rates, notable individual- and regional-level disparities persist, suggesting demand-side determinants beyond vaccine supply constraints. Prior work links socio-economic, demographic, political, religious, health, and informational contexts to vaccine uptake, yet the role of perceived inequality remains underexplored. Given widening global and Korean inequality and Korea’s reliance on voluntary compliance for vaccination, the research asks whether perceived inequality erodes trust in science and social institutions and thereby reduces trust in vaccines and willingness to vaccinate. The study adopts a dual approach: individual-level path analysis of COVID-19 vaccine attitudes and behaviors, and district-level panel analysis of influenza vaccination (2015–2021) to assess how local economic inequality relates to vaccine uptake.

The paper situates its inquiry within several strands of literature. Prior research identifies socio-economic status, education, age, religion, political ideology, race/ethnicity, and health conditions as determinants of vaccine uptake. For COVID-19, trust in healthcare and institutions, perceptions of efficacy and safety, misinformation, social media exposure, and government responses are central influences. Theoretically, the work draws on social context and inequality scholarship (Bourdieu and Wacquant; Wilkinson and Pickett), emphasizing that subjective perceptions of inequality—shaped by reference groups and local context—can diverge from objective status and affect behavior. Status anxiety and reduced self-efficacy in unequal settings may diminish engagement with science and technology, lowering trust in scientific knowledge. Social trust, influenced by experiences of deprivation and inequality, underpins social capital and compliance with collective health measures; low social trust can depress expectations of others’ cooperation and reduce one’s own vaccination. Based on this, the study posits hypotheses: H1 perceived inequality lowers trust in science; H2 trust in science increases trust in vaccines and willingness to vaccinate; H3 perceived inequality reduces social trust; H4 social trust increases trust in vaccines and willingness to vaccinate; H5 perceived inequality directly reduces trust in vaccines and willingness to vaccinate.

The study comprises two analyses. Analysis 1 (individual level) uses the 2021 Korean Happiness Survey (National Assembly Futures Institute), a nationally representative face-to-face survey (Aug 21–Oct 27, 2021) with over 17,000 respondents. Key variables: perceived inequality (1–5; higher = more inequality) from “How equal or unequal do you think income and wealth are in our society?”; trust in science (1–5) from agreement that science and technology make life more comfortable; social trust (1–5) from agreement that most people are trustworthy; vaccine trust measured via reverse-coded agreement with “It is difficult to believe the information provided about a vaccine for COVID-19” (1 = highly agree to 5 = not at all, higher = more trust); vaccination behavior via binary indicator (1 if vaccinated at least once or intends to vaccinate soon; 0 otherwise). Controls include income (1–12), education (1–5), age (1–6), political orientation (1 = left to 10 = right), and dummies for sex (female), homeownership, managerial/professional occupation, marital status (living with spouse/partner), and religion. Path analysis (a subset of SEM) with maximum likelihood estimation assesses direct and indirect effects along hypothesized paths: perceived inequality → science trust and social trust; science trust and social trust → vaccine trust (and alternatively vaccination); and direct paths from perceived inequality to vaccine trust (and vaccination). Model fit and robustness checks are reported in supplementary materials. An alternative model includes a path from vaccine trust to vaccination.

Analysis 2 (district level) examines influenza vaccination turnout across 252 Korean administrative districts (si-gun-gu) from 2015–2021 (panel N = 1764). The primary explanatory variable is district economic inequality measured by a Palma ratio constructed from national health insurance premiums (mandatory coverage for ~97% of the population), using average premiums of the top 10% relative to the bottom 40% at the district level; higher ratios indicate higher inequality. District average income/wealth level is proxied by the log of average insurance premiums. Controls include district-level education, average age, sex ratio, degree of urbanization, and healthcare access (population per hospital). Estimation strategies include OLS, fractional probit (for fractional turnout), random effects, fixed effects with district FE, and two-way fixed effects (district and year FE) to account for time-invariant heterogeneity and common shocks. Robustness checks and additional specifications are documented in the supplementary appendix.

Analysis 1 (path analysis):

• Perceived inequality significantly reduces trust in science (estimate ≈ −0.037, p ≈ 0.0), supporting H1.
• Trust in science significantly increases trust in vaccines (estimate ≈ 0.157, p ≈ 0.0), supporting H2.
• Perceived inequality significantly reduces social trust (estimate ≈ −0.102, p ≈ 0.0), supporting H3.
• Social trust significantly increases trust in vaccines (estimate ≈ 0.012, p ≈ 0.0), supporting H4.
• Perceived inequality directly and negatively affects trust in vaccines (estimate ≈ −0.036, p ≈ 0.0), supporting H5.
• Using vaccination (vaccinated or intends) as the outcome yields qualitatively similar pathways: perceived inequality lowers science and social trust, which in turn are positively associated with vaccination (estimates ≈ 0.008 and 0.009; p ≈ 0.0). An alternative model including a path from vaccine trust to vaccination finds this path statistically insignificant, likely due to conceptual overlap and multicollinearity between vaccine trust and vaccination (Pearson r ≈ 0.7, p ≈ 0.03).

Analysis 2 (district-level panel of influenza vaccination, 2015–2021):

• District inequality (Palma ratio) is consistently negatively associated with vaccination turnout across specifications: OLS ≈ −0.011 (SE 0.001, p < 0.01); Fractional Probit ≈ −0.032 (SE 0.003, p < 0.01); Random Effects ≈ −0.014 (SE 0.001, p < 0.01); District FE ≈ −0.014 (SE 0.006, p < 0.05); Two-way FE ≈ −0.011 (SE 0.005, p < 0.05). Higher average district income/wealth (log premiums) is generally positively associated with vaccination, though the association attenuates in the most conservative two-way FE model.
• At the metropolitan/provincial level, COVID-19 and influenza vaccination rates are positively correlated (Pearson r ≈ 0.72, p = 0.04).

Overall, perceived and contextual economic inequality diminish trust in science and society and are associated with lower trust in vaccines and lower vaccine uptake.

Findings indicate that perceived inequality undermines trust in science and social institutions, which are pivotal for accepting vaccine information and engaging in vaccination. These trust deficits appear to be key pathways by which inequality translates into lower vaccine confidence and uptake. The district-level results corroborate the individual-level mechanisms: residents in more unequal districts exhibit lower influenza vaccination turnout, consistent with the idea that people use local inequality as a heuristic about institutional reliability and collective cooperation. Together, the multilevel evidence suggests that efforts to bolster vaccine uptake must address not only informational deficits but also the broader social conditions—particularly inequality—that erode trust and collective action necessary for successful immunization programs. The results are relevant beyond COVID-19, as they generalize to routine influenza vaccination behavior over multiple years and across diverse local contexts.

The study demonstrates that perceived economic inequality reduces trust in science and society and is associated with lower trust in vaccines and reduced vaccination, both directly and via trust pathways. Complementary district-level analyses show that higher local economic inequality correlates with lower influenza vaccination turnout across a range of econometric models. Methodologically, the paper integrates individual-level path analysis with district-level panel models, offering a versatile framework for studying social determinants of vaccination. Policy implications include addressing economic disparities to rebuild trust and tailoring vaccination campaigns to the needs of communities experiencing inequality. Future research should extend beyond Korea to diverse contexts, employ longitudinal and multi-item measures to strengthen causal inference and construct validity, and investigate interactions between objective socio-economic status and perceived inequality in shaping trust and vaccine behavior.

Key limitations include reliance on single-item measures for core constructs (inequality perception, science trust, social trust, vaccine trust), which constrains reliability assessment; cross-sectional survey data in Analysis 1, creating temporal ambiguity and susceptibility to unobserved confounding; and, in Analysis 2, while fixed effects address time-invariant heterogeneity, time-varying confounders may remain. Measurement overlap between vaccine trust and vaccination likely contributed to an insignificant trust-to-vaccination path in the alternative model. The study does not fully disentangle how objective socio-economic status shapes perceived inequality and interacts with it to influence vaccine-related attitudes and behaviors. Data constraints precluded richer multi-item scales and stronger causal designs.