“…Following previous studies (Chen, 2013; Chen, Liu, Zhu, & Li, 2017; Knight & Gunatilaka, 2010; Liu, Zhang, Wu, Liu, & Li, 2017), we included a set of covariates in the models, including age (continuous variable), gender (categorical variable), hukou (categorical variable), marital status (categorical variable), the length of residence in the host city (continuous variable), general self‐rated health (categorical variable), average working hours per day (continuous variable), participation in the Urban Basic Medical Insurance (categorical variable, including Urban Employee Medical Insurance and Urban Resident Medical Insurance), the logarithm of the area of cultivated land per household member (continuous variable) and self‐built housing area per household member (continuous variable). Additionally, it has been demonstrated that age has a U‐shaped relationship with SWB, so we included the square of age (centred on the grand mean) in the model (Clark, Yi, & Huang, 2019; Wang et al, 2019). Note that no concept was included in both the dependent variable (i.e., SWB) and the variables on the right‐hand side.In addition, we employed a directed acyclic graph (DAG) with the help of DAGitty v3.0 software (www.dagitty.net) to illustrate the relationship among variables (see Figure A1).…”