What influences the perception of fairness in urban and rural China? An analysis using machine learning

A significant gap has long existed between urban and rural areas in China, leading to substantial differences in people’s perceptions of fairness. Analyzing the factors influencing fairness perceptions from an urban-rural dual perspective is crucial, as it can help enhance targeted strategies to improve fairness perceptions and promote social stability. This study uses CSS data from 2013 to 2021 to distinguish between senses of opportunity fairness and outcome fairness within urban and rural hukou population. Applying the Gradient Boosting Regression machine learning model and SHAP model, we identified key variables affecting these two types of fairness sense. The findings reveal that: First, rural residents have a stronger sense of fairness than urban residents, and fairness sense is higher in western regions compared to central regions. Second, urban residents’ fairness perceptions are mainly and stably influenced by personal characteristics and social environment. Lastly, rural residents’ fairness perceptions are influenced by more diverse factors, including personal characteristics, family environment, social environment, and internet use. This study is the first to use machine learning models to explain differences in factors affecting fairness perceptions between urban and rural hukou. These findings provide targeted insights for improving fairness perceptions among different groups and promoting social harmony.