What drives generational differences in subjective well-being? A machine learning study of Chinese migrant workers in the Yangtze River Delta.

Abstract

The subjective well-being of migrant workers amid China's rapid urbanization has emerged as a critical issue, intersecting social policy and mental health research. This study employs Social Quality Theory's (SQT) four-dimensional framework-encompassing social security, inclusion, cohesion, and empowerment-to examine the associated factors of migrant workers' subjective well-being and generational differences. Utilizing machine learning and Shapley additive explanations (SHAP) analysis on Yangtze River Delta survey data, the research yields key findings that: (i) New-generation migrant workers report significantly lower subjective well-being levels than their first-generation counterparts; (ii) The subjective well-being of first-generation migrant workers is predominantly associated with social security factors (particularly work stability). For the new generation, while primarily related to these social security considerations as well, key associated factors extend to include family support, urban integration and belonging-collectively encompassing all dimensions of SQT; (iii) Migrant workers' subjective well-being exhibits positive linear associations with multiple factors, including work stability, economic status, and self-fulfillment; (iv) Nonlinear relationships emerge, including inverted U-shaped patterns for urban integration and perceptions of socio-economic status across both generations, as well as a distinct U-shaped association between career development and subjective well-being specific to new-generation workers; interestingly, the welfare system shows no positive association with subjective well-being for either generational cohort. These findings provide region-specific empirical insights for policymakers to design targeted interventions that address generational differences in mental health outcomes.