Assessing the Impact of Economic Variables on the Energy Demand in Malaysia: A Regression Approach

The Knowledge of long-term energy demand, is essential for sustainable energy planning, especially in emerging economies. This study examines the relationship between Final Energy Consumption (FEC) and eleven economic variables in Malaysia from 1980 to 2022. The analysis used univariate regressions, feature selection techniques (forward selection, lasso, and recursive feature elimination), and multiple regression with diagnostic checks to systematically evaluate the explanatory power of each predictor. The results reveal that Gross Domestic Product (GDP), manufacturing output, electricity consumption, and urban population are the first factors influencing FEC, while imports, exports, and Foreign Direct Investment (FDI) play secondary roles. Population and trade balance exhibit redundancy and limited predictive value. Diagnostic testing, including the Variance Inflation Factor (VIF), Durbin–Watson statistics, and residual analysis, confirms the reliability of the regression framework. The results show that GDP, manufacturing output, and urban population are the most significant predictors of long-term energy demand (R² = 0.94, p < 0.01), whereas trade-related variables have a moderate impact. These findings underscore the importance of aligning industrial growth, urbanization policies, and trade strategies with sustainable energy planning, highlighting the advantages of using regression with feature selection and diagnostics.