China and some other Asian countries have experienced a large surplus of men of marriageable age. The existing literature studies the impact of sex imbalance using aggregate sex ratios, such as at the county, city, or province level. However, these studies may miss important impacts on health and behavior because the relevance of surplus sons to family decisions mainly stems from pressure conveyed through social interactions with the local reference group.

This paper draws from unique social network data, collected from households' long-term spontaneous gift exchange records (li dan), combined with household panel data from 18 Chinese villages to explore the prevalence of men's localized pressure to get married. The surveyed villages are home to Chinese ethnic minorities, which largely circumvents endogenous fertility decisions on the first-born child due to the implementation of One Child Policy and its associated relaxations afterwards. To identify the effect of pressure to find wives for their sons on parental risky behavior, we focus on comparing families with a first-born boy versus a first-born girl and distinguish the network spillover effect from the direct effect.

The spatial econometric decompositions suggest that the pressure mainly originates from a few friends with unmarried sons and unbalanced sex ratios in the friendship networks, though own village sex ratio and having an unmarried son also affects parental risk-taking behavior. The results are consistent across specifications allowing for long-run and short-run effects. We also find similar patterns for parental working hours, their likelihood to engage in entrepreneurial activities and decision to migrate. In contrast, parents with a daughter do not demonstrate this pattern. Since the sex ratio imbalance in China will probably worsen in the next decade, disentangling the real sources of marriage market pressure may help design policies to improve parental well-being.

Dr. Xi Chen's main research interests involve health economics and development economics in the developing contexts. He recently completed his PhD in applied economics at Cornell. His research seeks to better understand how social interactions affect health behavior and outcomes, how socioeconomic status drives social competition. Most of his current work draws on primary data from China and secondary data from India and Indonesia.

This research aims to better understand the impact of the Matlab health interventions by using panel data to control for unobservables and understand the dynamics and long-term effects of these programs. Heterogeneity in the fertility response to the family planning program is analyzed, using sequential fertility to isolate the family planning program from other interventions and examine heterogeneity based on time-varying characteristics. The link between childhood measles vaccination and school enrollment is examined using instrumental variables, and is motivated by the hypothesis that by avoiding the long-term health effects of a disease, vaccinated children are higher-achieving. Both analyses generate interesting findings that are not captured using the traditional methodologies and outcomes of program evaluation.

Julia Driessen, PhD, is an assistant professor of health policy and management in the Graduate School of Public Health at the University of Pittsburgh. She has a secondary appointment in the Department of Economics. In 2011 Dr. Driessen received her PhD in Economics from Johns Hopkins University. Her research interests include program evaluation and the links between health interventions and socioeconomic status, with an emphasis on heterogeneity of program effects as well as long-term outcomes. Recent research has analyzed the schooling effects of childhood measles vaccination and variation in the fertility response to a family planning program in Bangladesh. Her primary new interest since arriving at Pitt is the clinical and financial effects of electronic medical records in developing countries.