How can government expenditure effectively achieve energy poverty reduction? A non-linear perspective

Most existing literature assumes that the relationship between economic variables follows a linear model, and therefore uses linear models to investigate energy poverty such as panel fixed effects model, and generalized method of moments approach. In fact, economic cycles lead to non-stationary sequences of economic variables, and the nexus between economic variables are more likely to be nonlinear. To address this gap, the study employs the nonparametric additive model to explore the role of government spending in energy poverty. The primary empirical findings are as follows: (1) government spending has a positive U-shaped impact on energy poverty. This indicates that the poverty reduction effect of government spending is more prominent in the early stages, and this effect gradually diminishes in the later stages. (2) Heterogeneity analysis reveals that the nonlinear effect of resource exploration expenditure follows a U-shaped pattern. In contrast, the nonlinear impact of education spending demonstrates an inverted U-shaped pattern. The effect of social security expenditures is more intricate, exhibiting an inverted N-shaped mode. (3) Mechanism analysis reveals that green finance exerts an inverted N-shaped influence on energy poverty. This suggests that, in general, the implementation of green finance practices is beneficial for alleviating energy poverty. The influence of green technology innovation is characterized by an inverted U-shaped pattern, suggesting that its effectiveness in poverty reduction becomes increasingly prominent over time.