S. (Shuai) Fang, PhD

Research project: "Shelter"- Housing and Household Finance

Housing markets are crucial to a nation’s wellbeing. First, housing typically is a household’s largest financial asset that comes with considerable financial risks in terms of negative equity and foreclosure. Moreover, and perhaps more importantly, housing provides shelter and safety, being instrumental to school performance of children and workers’ job market performance. Third, housing (construction) provides a considerable share in GDP and jobs. Lastly, housing relates to land and (public) land assembly which crucially feed into urban planning. All these four dimensions lead to a multitude of societal questions, of which some need urgent consideration (see for instance the Dutch National Research Agenda, Nationale Wetenschapsagenda).

This research project aims to consider suchsocietal questions using academic research methods, and is geared towards providing policy-related insights regarding ‘the’ housing market. It explores household beliefs and decisions in relation to housing. This thesis reports on empirical studies conducted using nationwide administrative data from the Netherlands. It adds new insights to the field of household finance and housing by highlighting the roles that housing plays in households’ belief formation and decision-making. The main research question is How do households form beliefs and make decisions in relation to housing? Answers to this research question are provided by findings that are reported in four studies.

The first study examines the effect of house value growth on households' expectations regarding future house prices using nationally representative survey data combined with administrative registry data from the Netherlands. While household expectations can be shaped by market experience, the analysis demonstrates that households that havemore positive individual experience, measured by the annualized house value return since they moved in to their current residence, have greater expectations about the future housingmarket growth. The findings can be explained by the heterogeneity in the recency and the informativeness of information, and the ability to utilize information.

The second study addresses buyer households’ perceptions of the quality of housing given its listing history. This study aims to test for the existence of stigma effects in the sense that a house that remains unsold for a lengthy period may therefore be perceived as of low quality. Since relisting a property clears listing records and restarts the time on the market to zero, this analysis uses house relisting as a quasi-natural experiment. The findings show that a relisted house has a higher probability of a sale than a non-relisted house at any given time, ceteris paribus, which is interpreted as evidence that there are stigma effects in housing markets.

Given that housing generally makes up a large proportion of a household’s asset portfolio, the third study examines households’ portfolio choices, in terms of holding risky financial assets, in relation to housing value growth. Using administrative data covering the Dutch population, this study reports a positive effect of housing value in terms of both the likelihood of holding any risky financial assets and the proportion of risky financial assets in total financial wealth. While we find positive effects for both homeowners and renters, a further analysis of housingstock market correlations reveals heterogeneity in their motives: diversification among homeowners and hedging among renters for holding risky financial assets as a response to changes in housing values.

The fourth study investigates the effect of parents’ housing experience on their children’s home-purchasing decision. Using administrative data covering the Dutch population to establish parent-child pairs, the analysis shows that higher levels of parental housing returns are correlated with an increased likelihood of their children becoming homeowners. Further analysis shows that children’s expectations regarding future house prices could be shaped by their parents’ housing returns and that these findings hold for parents whose non-housing assets or net worth are in the lowest quartile in terms of value. These findings suggest that information sharingmight be an underlying mechanism, in addition to intergenerational transfer, in explaining the intergenerational persistence in housing wealth.