Pathos or logos? How governance legitimacy perception influences individual privacy trade-offs during COVID-19 pandemic

The study addresses the tension between protecting individual privacy and enabling effective public health governance during COVID-19. Governments’ reliance on personal data (e.g., contact tracing, geolocation, biometric data) raises privacy and rights concerns, yet many individuals still disclose information for pandemic control. Prior explanations emphasize functional benefits (perceived effectiveness, ease of use, risk–benefit trade-offs) and trust or values (trust in authorities, social norms, greater good). The authors identify a gap: most value-based explanations focus on trust in institutions and social influence but overlook individuals’ evaluations of the legitimacy of governance actions themselves. The research therefore examines how perceived governance legitimacy—distinguishing cognitive legitimacy (logos; the necessity and taken-for-granted appropriateness of actions) and moral legitimacy (pathos; alignment with social norms and public interest)—shapes willingness to disclose health information during crisis. The study tests whether these legitimacy perceptions directly increase privacy disclosure and whether altruism mediates these relationships, as well as whether education moderates the effects on altruism.

Legitimacy is defined as public acceptance of authorities’ actions as appropriate within social norms, values, and beliefs. In crises, legitimacy underpins voluntary compliance with public health measures. The authors distinguish: (1) cognitive legitimacy—passive acceptance that actions are necessary, comprehensible, or taken for granted; and (2) moral legitimacy—active evaluation that actions align with ethical standards and promote public interest. Hypotheses: H1a/H1b posit positive effects of cognitive and moral legitimacy on privacy disclosure. Altruism—prosocial acts without expectation of reward—can motivate information sharing but may be challenged by privacy risks and stigma; perceived legitimacy may sustain altruistic behavior by framing compliance as the right course of action. Thus, H2a/H2b propose altruism mediates the effects of cognitive and moral legitimacy on disclosure. Education may moderate legitimacy’s effects on altruism by shaping critical evaluation skills, moral values, and understanding of trade-offs; hence H3a/H3b posit education moderates the cognitive/moral legitimacy-to-altruism relationships.

Design and setting: Cross-sectional online survey during the Shanghai lockdown context. Data were collected from July 15 to August 14, 2022, via wenjuan.com, a large Chinese online research platform with quality controls. Quota sampling by district matched the 7th Shanghai population survey: Pudong 22.8%, Puxi 46.6% (Minhang, Jingan, Huangpu, Putuo, Hongkou, Yangpu, Baoshan), suburban 30.6% (Jinshan, Chongming, Jiading, Songjiang, Qingpu). Sampling and sample: 2,583 questionnaires distributed; 1,074 returned (41.58% response). After excluding incomplete, too-fast (<10 min) and incomplete open-ended responses, N=1,008 valid cases. Measures: - Privacy disclosure (4 items, 5-point Likert 1–5; e.g., willingness to provide health/travel info to authorities). Reliability α=0.83; M=4.01, SD=0.75. - Cognitive legitimacy (6 items adapted from Alexiou & Wiggins, 2019; 1–5 Likert; necessity and purpose of information collection for pandemic control). α=0.90; M=3.89, SD=0.79. - Moral legitimacy (9 items; adherence to ethical standards, public endorsement, regulation following, public interest). α=0.92; M=3.82, SD=0.74. - Altruism (3 items from Rushton et al., 1981; Feng et al., 2020; helping others during pandemic; 1–5 Likert). α=0.71; M=3.88, SD=0.71. - Education: seven categories collapsed into low (primary, middle, high school) vs high (junior college, undergraduate, master’s, doctoral); dummy-coded for moderation. Controls: gender, age groups (18–29; 30–39; ≥40), annual household income (<200k RMB; 200–500k; >500k), and media consumption frequency during lockdown (nine media categories; composite α=0.80). Analysis: Structural equation modeling (SEM) in R tested: (1) main effects of cognitive and moral legitimacy on privacy disclosure; (2) altruism as mediator between legitimacy and disclosure; (3) education as moderator on paths from legitimacy to altruism. Control variables entered on cross-stage paths. Model fit for full model with interactions indicated excellent fit (e.g., CFI=1.000; RMSEA=0.00; SRMR=0.00).

• Main effects: Both cognitive legitimacy (β=0.33, SE=0.03, p<0.001) and moral legitimacy (β=0.48, SE=0.04, p<0.001) positively predicted privacy disclosure. - Mediation via altruism: Cognitive legitimacy (β=0.16, SE=0.07, p<0.001) and moral legitimacy (β=0.18, SE=0.06, p<0.001) positively predicted altruism; altruism positively predicted disclosure (β=0.08, SE=0.02, p<0.01). The total indirect effect was significant (β=0.03, SE=0.01, p<0.001), supporting mediation for both legitimacy dimensions. - Moderation by education: Interaction terms indicated education moderated legitimacy→altruism paths: Education × Cognitive legitimacy β=0.44 (SE=0.15, significant), Education × Moral legitimacy β=−0.43 (SE=0.15, significant). Interpretation: higher education strengthens the effect of cognitive legitimacy on altruism and weakens the effect of moral legitimacy on altruism. - Model fit indices indicated excellent fit (CFI=1.000; RMSEA=0.00; SRMR=0.00). Overall, moral legitimacy exhibited a larger direct effect on privacy disclosure than cognitive legitimacy; altruism partially mediated both relationships; and education differentially moderated the legitimacy–altruism pathways.

The findings show that perceived governance legitimacy is a pivotal driver of citizens’ willingness to disclose sensitive health information during crises. Moral legitimacy—perceptions that authorities’ actions align with ethical standards, public interest, and social norms—exerts a stronger influence than cognitive legitimacy on disclosure, especially in cultural contexts (e.g., China) where collective moral norms and social obligations are salient. Legitimacy promotes altruism, which in turn increases willingness to disclose personal information for the common good, highlighting altruism’s mediating role in the legitimacy–disclosure pathway. Education differentiates how legitimacy translates into altruistic motivation: higher education appears to enhance responsiveness to cognitive (reasoned, rule- and evidence-based) justifications for governance, while lower education shows greater sensitivity to moral appeals rooted in traditional norms. These results suggest that legitimacy-building strategies tailored to cultural and educational contexts can more effectively encourage voluntary participation in public health data practices. The study also points to broader implications for privacy management as governance increasingly relies on personal data: trust and legitimacy narratives can shape citizen cooperation beyond purely functional incentives.

The study advances understanding of how governance legitimacy—distinguished into cognitive and moral dimensions—shapes individuals’ privacy trade-offs during crises. Moral legitimacy showed stronger effects on privacy disclosure, altruism mediated both legitimacy paths to disclosure, and education moderated legitimacy’s impact on altruism (strengthening cognitive pathways and weakening moral ones at higher education levels). Conceptually, the work introduces and motivates a governance-oriented perspective on privacy, emphasizing legitimacy-building and culturally informed strategies to promote cooperative data sharing in public health. Future work should examine cross-cultural generalizability, integrate macro-level modernization indicators with micro-level legitimacy and privacy behavior, and further unpack how moral versus cognitive frames shape disclosure intentions in varying societal contexts.

• Generalizability: Data were collected in the specific context of Shanghai’s 2022 lockdown; results may not extend to other regions with different cultural backgrounds or policy approaches. - Self-report bias: Measures of disclosure intentions, legitimacy perceptions, and altruism relied on self-reports, which may be affected by social desirability or recall bias. - Cross-sectional design: The study’s design limits causal inference and the ability to observe changes over time; longitudinal research is needed to track dynamics across pandemic phases. - Sampling via online panel and quota methods, while controlled for quality, may still introduce selection biases relative to the broader population.