Analysis of Driving Forces of Energy Consumption in Rural Areas of China’s Henan Province Based on the STIRPAT Model and Ridge Regression

Abstract

The main objective of this paper was to analyze the status quo of energy consumption in rural Henan and identify the driving forces governing energy consumption based on the STIRPAT model. Potential driven factors of energy consumption including power of agricultural machinery, effective irrigated area, investment, income, total value, and per capital living space were selected to build the extended stochastic impacts by regression on population, affluence, and technology (STIRPAT) model, where ridge regression was applied to eliminate multicollinearity. The estimation results show that all the factors are positively related to rural energy consumption in Henan. The effective irrigated area is the maximal driven factor of energy consumption in rural Henan. Per capital living space, peasant household investment, power of agricultural machinery, total value of agricultural output, and per capita income all contribute to rural energy consumption growth in Henan. Finally, we propose policies that can be applied to mitigate future energy consumption in rural Henan.

Average ordinary least squares‐centeredpenalized regression: A more efficient way to address multicollinearity than ridge regression

“…The MSE is more useful than the bias in real‐world applications because the MSE and bias reflect the accuracy in a single sample and the average accuracy in an infinite number of samples, respectively. Therefore, a large MSE will result in an inaccurate estimator in the practical application and even an estimator that contradicts intuition, for example, when using the OLS regression, atmospheric pollutants are estimated to be protective of human health (Vedal, Brauer, White, & Petkau, 2003; Roberts, 2006), and the impact of the power of agricultural machinery on energy consumption is estimated to be negative (Tian, Li, Shao, Li, & Zheng, 2020).…”