| Summary: | <p>This thesis contributes to the sociological literature about how social capital influences individual labour market outcomes by using and combining a constellation of identification strategies to alleviate endogeneity bias. The principal proposition of social capital theory has been that social resources embedded in social networks have significant influences on individual labour market outcomes. Although a vast amount of empirical literature has accumulated, previous sociological studies of this line have not paid enough attention to the endogeneity problem rising from various sources such as omitted variables, self-selection, sample-selection, and simultaneity bias. Unless the endogeneity problem is explicitly pointed out and well-corrected for, one cannot arrive at any convincing causal conclusions about whether social capital matters or not. </p> <p>Building on the existing literature, this thesis offers a systematic review in Chapter 2, paying close attention to the identification problem and empirical strategies. To identify causality under the counterfactual framework, four empirical chapters test for the effects of social capital using different identification strategies. Chapter 3 combines the instrumental variable method with the Heckman model to estimate network effects on labour market outcomes among rural-to-urban migrants in China. Chapter 4 presents evidence of the network effects on individuals' employment and income in China employing the instrumental variable method to control for the fact that friendships are not randomly formed. Using panel data and Heckman model from the former East Germany and contemporary China, Chapter 5 and Chapter 6 address the direct effects of using contacts to find jobs and the effects of contact’s prestige in labour markets under state socialism. </p>
|
|---|