Social Trust and Commercial Insurance Participation in China: Distinguishing Within-Person Change from Between-Person Differences

This study examines how trust is associated with household commercial-insurance participation in China and separates two sources of variation that pooled estimates combine. Within-person change asks whether changes in the same respondent’s trust across survey waves are accompanied by changes in that respondent’s household insurance participation. Between-person differences compare respondents who maintain different average trust levels over time. Using six waves of the China Family Panel Studies (CFPS) from 2012 to 2022, we analyze 47,006 respondent-wave observations from 15,176 respondents. Pooled linear probability and Probit models show a positive association between the aggregate trust index and insurance participation. In contrast, respondent fixed-effects, conditional fixed-effects Logit, and city-by-wave fixed-effects models provide little evidence of a systematic within-person association. Correlated random-effects estimates locate most of the pooled relationship in persistent differences between respondents. The trust dimensions also differ: persistent trust in strangers is positively associated with participation, whereas trust in local government cadres is negatively associated. Expenditure models lead to a similar conclusion. The results show why generalized interpersonal trust and institutional trust should not be treated as interchangeable and why pooled trust–insurance associations need not describe changes within the same person over time.