Predicting China's Marriage Rate: Causal Inference Using Dual Machine Learning (DML) with XGBoost, LightGBM, CatBoost, and GBDT

After China's accession to the WTO and 20 years of rapid development, the marriage rate has shown a downward trend. This study aims to analyze the impact of socio-economic factors on the crude marriage rate (CMR) panel data in China from 2003 to 2022 using Dual Machine Learning (DML) for Causal Inference and machine learning models. DML was used to estimate ATE, CATE, and HTE for various factors. Four models—XGBoost, LightGBM, CatBoost, and GBDT—were employed for predictions, using 10-fold cross-validation for model evaluation. The results indicate that education (ATE 5.666) and birth rate (ATE 5.492) had the most significant positive impacts on CMR, while GDP (ATE 3.196) showed positive but varying effects, and the female proportion (ATE - 2.353) had a notable negative impact. CatBoost performed best in MSE (0.942) and RMSE (0.958), while LightGBM excelled in MAE (0.777). Education, GDP, and birth rate are key factors influencing CMR. CatBoost and LightGBM proved to be effective prediction models, though improvements are needed for regions with significant variability.