Related issues to rural-to-urban migration in china

Abstract

Over the past two decades, China has witnessed unprecedented rural-to-urban migration. This thesis includes three self-contained empirical chapters, which are related to rural-to-urban migration in China. Migrants are vulnerable to mental health problems, so it is important to understand the factors that can mitigate their mental stress. Chapter 2 investigates the relationship between the social networks and mental health of Chinese rural-to-urban migrants. The empirical analysis is based on a unique migrant survey from the Rural-to-Urban Migration in China (RUMIC) project, which includes the most up-to-date information about Chinese rural migrants. Using OLS and fixed effect models, I find that larger networks are correlated with better mental health. I use the instrumental variable approach to mitigate endogeneity bias. Both IV and fixed effect IV estimates indicate that social networks significantly help reduce mental health problems. The heterogeneity analysis suggests that the effect is larger for migrants with smaller social networks or with limited access to social welfare. In addition, females benefit more from their social networks than males. Migrants are socially segregated and discriminated against in their destination cities in China. Chapter 3 uses a large representative survey to investigate whether interpersonal contact between urban locals and migrants could improve urban locals' attitudes towards migrants. The OLS estimates show that having previous contact experience with migrants is positively and significantly correlated with urban locals' willingness to interact with migrants. I adopt Lewbel (2012)'s heteroskedasticity identification approach to mitigate endogeneity bias between contact and attitudes. The estimates indicate that having previous contact experience with migrants could significantly improve willingness to engage in non-intimate interactions, but has no significant effect on willingness to engage in intimate interactions. The migrant household survey of the Rural-to-Urban Migration in China Project is the largest longitudinal survey documenting the city life of rural migrants in China. The survey has been widely used in migration studies. However, a large proportion of respondents have left the survey sample, because migrants tend to be very mobile. Chapter 4 studies whether attrition is random, and the extent to which attrition biases estimated results of the first six waves of the survey. The empirical analysis suggests that there are systematic differences between the non-attritors and attritors. The non-attritors tend to be socio-economically better off, enjoy larger income gains from migration, be more willing to stay in cities and are more likely to be self-employed than the attritors. This chapter finds that there is likely to be some attrition bias and that the non-attritor sample is unrepresentative of the general migrant population at the time of follow-up surveys. Nevertheless, the examples shown in this chapter suggest that, in some cases, attrition bias and sample (un)representativeness could have only limited impact on the regression coefficients of the individual-level variables which are most relevant to research and policy interests.