지방소비세 수입 예측 모델 개발에 관한 연구

Local finance is a key determinant of local governments’ capacity to provide public services. Ensuring fiscal stability and autonomy is an essential task for realizing decentralized governance and balanced regional development. With recent policies aimed at expanding fiscal decentralization, the Local Consumption Tax (LCT) has emerged as a core revenue source for local governments, highlighting the growing importance of accurate and sophisticated forecasting. This study analyzes the structural characteristics and determinants of LCT revenue and develops short- and medium-term forecasting models to enhance the predictability and reliability of revenue estimates. For short-term forecasting, cumulative monthly revenue data was used in combination with the progress rate estimation method and moving average method, and forecast accuracy was evaluated using RMSE and MAPE indicators. In the medium-term analysis, gamma regression models were applied with key explanatory variables such as private final consumption expenditure, the consumer price index, and the base interest rate. To control for tax rate changes, revenue data was adjusted to reflect collections per 1% tax rate. In addition, LCT was classified into four components—consumption-based allocation, acquisition tax compensation, transfer project compensation, and remaining consumption- based allocation. Each component’s predictability and the necessity for projection were reviewed. By presenting a forecasting framework grounded in quantitative analysis, this study advances beyond traditional empirical estimation methods and contributes to improved accuracy in LCT forecasting and financial planning by local governments. In the context of deepening fiscal decentralization and the increasing emphasis on responsible local fiscal management, the study offers meaningful implications for both policy and academic discussions.