Suggestions for Promoting a Better Public Service System under the Low Fertility Dilemma Based on VAR Modeling

This paper provides an in-depth analysis of the rapid decline in fertility and population aging. Using LASSO penalized Poisson regression and TVP-VAR model improved by MCMC algorithm based on Bayesian network, the factors affecting the fertility level and their interrelationships are studied. It is found that fertility and education levels affect each other with probabilities of 0.0806 and 0.0825 respectively at 10% significance level. The positive shock of social security level on fertility level increases the response value of fertility level from −1.5 to about 7.5 in 10 lags, showing a positive effect. In contrast, the response of fertility level to house prices gradually decreases from 5 in period 1 to −2.5 in period 10, indicating a negative impact. These analyses provide a scientific basis for constructing a public service fertility support system that can help mitigate the current challenges of low fertility levels.