How do intertemporal preference and social signals drive parental choices? a discrete choice experiment on human papillomavirus vaccination

Parental vaccine hesitancy constrains human papillomavirus (HPV) vaccination uptake. However, how intertemporal preferences and social signals relate to this hesitancy remains underexplored. An online survey integrating a discrete choice experiment (DCE) and an intertemporal choice task was deployed to quantify attribute preferences, monetary discount rates, and the factors shaping parental decisions. Primary attribute preferences were estimated using a mixed logit model, while latent class analysis identified preference heterogeneity. Additionally, interaction terms were utilized to examine how intertemporal preferences modulate these choices. Empirical data reveal that protection duration and vaccine effectiveness were the primary determinants of parental choice, while social-reference information indicating that healthcare professionals’ children had been vaccinated was also valued. Vaccine origin showed no significant effect. Latent class analysis segmented the cohort into a receptive majority (79.4%) and a hesitant minority (20.6%). Interaction models indicated that monetary discounting was jointly associated with the structure of preferences, although no individual attribute interaction survived adjustment for multiple testing. Parental choices about HPV vaccination reflected clinical attributes, social reference information, and intertemporal preferences. Protection duration and vaccine effectiveness mattered most, and parents more often chose vaccination when told that healthcare professionals’ children had been vaccinated. Higher monetary discount rates were associated with membership in the receptive majority, although the attribute-level interactions were exploratory after adjustment for multiple testing. Reducing out-of-pocket costs and communicating credible social-reference information are worth testing as intervention strategies. Given the predominantly urban and highly educated sample, these findings should be generalized with caution.