In this paper, throughout a standard Public Choice model for the demand of public goods, we intend to analyze the public expenditures in Turkey. In doing so, we employ a panel approach to test the median voter theorem at provincial level, over the period 1995-2001. To estimate the parameters in the model with panel data, we use fixed effect estimation specification with least squares method (LS). In addition, to compare the results and justify the reliability of our estimates, we also employ the generalization method of moments (GMM). For a further look, we also advance our study at regional level. Our findings strongly support the theoretical model. Furthermore, our investigation at the regional level suggests sharp differences across the regions.

In this paper, throughout a standard Public Choice model for the demand of public goods, we intend to analyze the public expenditures in Turkey. In doing so, we employ a panel approach to test the median voter theorem at provincial level, over the period 1995-2001. To estimate the parameters in the model with panel data, we use fixed effect estimation specification with least squares method (LS). In addition, to compare the results and justify the reliability of our estimates, we also employ the generalization method of moments (GMM). For a further look, we also advance our study at regional level. Our findings strongly support the theoretical model. Furthermore, our investigation at the regional level suggests sharp differences across the regions.