Relief Satisfaction and Post-Disaster Community Engagement: Asymmetric Mediation through Community Cohesion and Social Support

Backgrounds
Disasters lead to different trajectories of civic engagement among survivors. Some withdraw from participation, while others become more actively involved through recovery efforts and community advocacy. Although satisfaction with disaster relief has been linked to recovery outcomes, the mechanisms through which it influences community participation remain unclear.
This study examines whether community cohesion and social support mediate the relationship between relief satisfaction and changes in community engagement. We propose an asymmetric mediation pattern: relational resources may help prevent civic withdrawal through identity-based and buffering mechanisms, whereas increases in civic engagement may operate more directly through institutional trust and personal efficacy.

Methods
We use national data from the Long-term survey on the change of life of disaster victims (NDMI, Republic of Korea), Wave 3 and Wave 4 (N = 2,440). The sample includes disaster survivors affected by typhoons, floods, earthquakes, and fires who received government relief pensions.
The outcome, community participation change, was a three-category variable (Decrease / Same / Increase), reflecting self-reported changes compared to the pre-disaster period. Relief satisfaction captured evaluations of services received (e.g., relief supplies, temporary housing, medical and psychological services). Two mediators were included: community cohesion and social support. Controls variables included age, gender, education, marital status, perceived damage severity, health status, and years since the disaster.
We used GSEM within a counterfactual mediation framework in Stata. Mediator models were estimated using linear regression, and the outcome model using multinomial logistic regression.

Results
The results show an asymmetric mediation pattern. For decreased participation (vs. no change), the total effect of relief satisfaction was significant (TE = −.028, CI = −.053, −.005). Both the natural direct effect (NDE = −.023, CI = −.047, −.001) and indirect effect (NIE = −.005, CI = −.010, −.002) were significant. The indirect pathway was driven by community cohesion, as social support was not significant, accounting for 19.2% of the total effect.
For increased participation, the total effect was significant (TE = .045, CI = .028, .063). The effect was mainly direct (NDE = .044, CI = .027, .061), with no significant indirect effect (NIE = .001, CI = −.001, .004). These results indicate that relief satisfaction promotes increased participation primarily through a direct pathway.
Overall, relief satisfaction influences participation asymmetrically. It reduces decline through both direct and indirect pathways via community cohesion, but increases participation mainly through a direct pathway.

Conclusion and Implications
This asymmetric pattern highlights the importance of distinguishing between prevention- and promotion-oriented processes in post-disaster community participation.
For policy, disaster assistance should be timely, adequate, and perceived as fair, as relief satisfaction shapes community engagement trajectories. Strengthening community cohesion is important for preventing declines in participation.
In contrast, increased participation appears to operate through direct pathways, suggesting a role for institutional trust and individual efficacy.