11:15 AM - 11:30 AM
[HTT15-08] Application of GIS to Local Public Finance: Focusing on Local Taxes
Keywords:local public finance, local tax, geographically weighted regression, local modeling
Background: GIS has been applied to various fields of municipal management. In particular, its usefulness in urban planning and disaster management is well known. In order for public actors to carry out such projects, it is necessary for local governments to collect financial resources and allocate budgets. However, there are no examples of GIS being used within the framework of local public finance. One of the reasons for this is the methodology of existing studies on local public finance. These studies have used analyses based on global modeling, which assumes a mechanism that works uniformly across the entire region. When making regional policies, it is important to pay attention to regional characteristics and regional differences, and to formulate specific policies for each region. However, conventional policy has focused on fiscal structure and size, overlooking regional characteristics. Therefore, there is potential for improvement.
Methods and Perspectives: This study focuses on local taxes, which account for a large share of municipal revenues. The analysis targets municipalities in the suburbs of the Tokyo metropolitan area, where local taxes account for a particularly large share of revenue. The analysis method is Geographically Weighted Regression analysis; GWR is a local modeling method that accounts for spatial non-stationarity and can detect relationships between variables that vary by location. This study compares the results of the GWR with a survey of municipalities. The questionnaire asked about the factors that cause fluctuations in local tax revenues and the measures to secure tax revenues that are important to the local governments. Similarly, for local tax collection rates, this study analyzed the GWR after detecting spatial concentration of collection rates using local Moran statistics.
Results: The results of the analysis showed that the explanatory power of the GWR was significantly higher than that of multiple regression analysis. The determinants of local tax revenues were variables such as working-age population and per capita manufactured goods shipments. The survey revealed that the perception of local tax revenue variables by local governments is spatially coherent and consistent with the spatial distribution of the determinants of local tax revenue. This implies that some municipalities have realized an increase in tax revenues through measures that respond to regional characteristics. The detection of the distribution of the coefficients also made it possible to design measures to secure tax revenues according to regional characteristics. Furthermore, the determinants of local tax collection rates were regional indicators related to poverty issues such as unemployment rates and the geographic location of local governments. Therefore, it is important to implement measures to secure tax revenues based on regional characteristics, cooperation among neighboring municipalities, and joint tax collection across prefectures. This study demonstrates the effectiveness of GWR, a local modeling method, for local taxes, which play a central role in revenue securing. The potential for the use of GIS for local public finance is great. It is expected that GIS centered on local modeling can be applied to public services and expenditures as well.
Methods and Perspectives: This study focuses on local taxes, which account for a large share of municipal revenues. The analysis targets municipalities in the suburbs of the Tokyo metropolitan area, where local taxes account for a particularly large share of revenue. The analysis method is Geographically Weighted Regression analysis; GWR is a local modeling method that accounts for spatial non-stationarity and can detect relationships between variables that vary by location. This study compares the results of the GWR with a survey of municipalities. The questionnaire asked about the factors that cause fluctuations in local tax revenues and the measures to secure tax revenues that are important to the local governments. Similarly, for local tax collection rates, this study analyzed the GWR after detecting spatial concentration of collection rates using local Moran statistics.
Results: The results of the analysis showed that the explanatory power of the GWR was significantly higher than that of multiple regression analysis. The determinants of local tax revenues were variables such as working-age population and per capita manufactured goods shipments. The survey revealed that the perception of local tax revenue variables by local governments is spatially coherent and consistent with the spatial distribution of the determinants of local tax revenue. This implies that some municipalities have realized an increase in tax revenues through measures that respond to regional characteristics. The detection of the distribution of the coefficients also made it possible to design measures to secure tax revenues according to regional characteristics. Furthermore, the determinants of local tax collection rates were regional indicators related to poverty issues such as unemployment rates and the geographic location of local governments. Therefore, it is important to implement measures to secure tax revenues based on regional characteristics, cooperation among neighboring municipalities, and joint tax collection across prefectures. This study demonstrates the effectiveness of GWR, a local modeling method, for local taxes, which play a central role in revenue securing. The potential for the use of GIS for local public finance is great. It is expected that GIS centered on local modeling can be applied to public services and expenditures as well.