Self-employment or wage-employment? On the occupational choice of return migration in rural China

Purpose The purpose of this paper is to quantify the impact of return migration on the occupational choice in rural China. Design/methodology/approach - The authors' research uses the two-stage residuals inclusion estimation, 2SRI, to deal with the endogeneity problem, and then compares the occupational choice between returnees and stayers with multinomial logit estimation and counterfactual analysis. Findings - The authors mainly find that: the migration experience has a significant positive impact on wage-employment activities, but may be has a negative effect on the entrepreneurial activities. The workers engaged in non-agricultural activities (self-employment and wage-employment) have the same characteristics in the labor market (Le. younger, male, higher education levels, less average land and parents with little children) compared to the agricultural activities, but these characteristics show no significant affect on the occupation choice between self-employment and wage-employment Research limitations/implications - This paper extends the empirical analysis in internal migration, but it also has some drawbacks, such as not enough data can be obtained to distinguish the occupations between different types of self-employment as own account workers and as entrepreneurs. Further research needs more comprehensive data to support. Originality/value - The authors' research is the first study which uses self-selection model to examine the activity choice of return migrants in rural China. They also extend the existing studies in two directions: first, they use nationally-representative data from the general social survey of China carried out in 2006 to examine the relationship between the return rural migrants and their occupational choices. Second, they propose a more exact category for rural occupational choice including non-agricultural activities (self-employee, wage-employment) and agricultural activities (peasants).