Does the housing provident fund crowd out the households’ non-housing consumption? Evidence from Chinese micro data

This study investigates the causal effect of mandatory housing provident fund (HPF) contributions on non-housing consumption patterns among urban Chinese households. Leveraging cross-sectional data from the 2017 China Household Finance Survey (CHFS), we employ instrumental variable estimation to address endogeneity arising from reverse causality and omitted variable bias. Results reveal a statistically significant negative association between HPF contributions and discretionary spending: An RMB 1 increase in per capita HPF outlays corresponds to an RMB 0.843 decline in non-housing consumption. Strikingly, this crowd-out effect intensifies among households that refrain from utilizing HPF for residential purposes—exhibiting a marginal propensity to consume reduction of RMB 1.492 per yuan contributed. Heterogeneity analyses further demonstrate that low-income families, who exhibit lower HPF withdrawal probabilities, bear the inhibitory effect: Their non-housing expenditures decrease more relative to higher-income counterparts under equivalent contribution levels. In both the theoretical and empirical models, the results suggest that whether a household has an HPF does matter, but it is more important for the household to take advantage of its lending capacity to mitigate liquidity constraints, rather than merely participate in it.