Social Capital and Housing Affordability: Evidence from China

ABSTRACT

This article attempts to examine the problem of housing affordability in China based on a set of household-level survey data. In contrast to the previous studies, our study focuses on the important implication of social capital for households’ house-purchasing decisions in this country. Our results show that household expenditures on the relations with parents and other relatives are important determinants for homeownership in China. We also find evidence that house-purchasing decisions are significantly affected by relatives-related variables such as number of immediate relatives in the same city, distance from parents, educational years of family head’s father, and whether parents are alive. Our research helps shed new light on the high homeownership rates in urban China.

KEY WORDS:

• Chinese household finance
• housing affordability
• housing demand
• social capital

JEL:

• R21
• R31
• P46

Notes

^1. For more details on the survey data of CHFS, please visit the CHFS website: www.chfsdata.org.

^2. Oorschot and Arts (2005) also suggest an analogous classification system of social capital, and highlight the importance of social relations with family members in social capital formation.

^3. Kutty (2005) also presents a better literature review on the measure issue of housing affordability.

^4. Although little work attempts to integrate these two strands of literature in order to examine the relationship between housing affordability and social capital in China, two noteworthy studies are Tu and Li (2011) and Liu, Wang, and Tao (2013). Tu and Li (2011) attempt to examine the effects of local social capital coupled with risk attitude on intra-city residential mobility in Tianjin, while Liu, Wang, and Tao (2013) pay special attention to the implication of connections with local residents for the housing conditions of migrant works in 12 Chinese cities.

^5. Since our research focus is on whether households that have the demand for owner-occupied housing have the affordability of housing, we do not attempt to investigate the house-purchasing behavior differences between the first-home buyers and second/third-home buyers. Thus, although our sample includes the second/third-home buyers in addition to the first-home buyers, we do not intend to distinguish between the first-home buyers and second/third-home buyers in probit regressions.

^6. Since there might be regional differences in economic development across provinces in China, we also control for the differences by adding 29 provincial dummies in our probit regression. Our regression results show that these provincial dummies almost have no impact on the statistical significance and influence directions of our key variables of interest. Due to the limitation of article length, however, we have omitted these results here.

^7. Similarly, we can also control for regional heterogeneity by adding 29 provincial dummies in our probit regression, while our results show that these provincial dummies almost have no impact on the statistical significance and influence directions of our key variables of interest.

^8. Since our article has not focused on the location choice of households but on the likelihood of house purchasing, we did not identify whether a household would like to pay more housing price premium for holding the emotional ties via staying in its familiar place. Nonetheless, this will be our future research direction.