Testing the Solow Hypothesis for Fiscal Convergence: a Dynamic Spatial Analysis

Abstract

The study of regional fiscal convergence is a recent extension of the neoclassical growth theory. Various studies have shown the existence of fiscal convergence across countries or states in federally governed countries. This paper tests the growth theory on income and fiscal variables differently in a centrally ruled country. Therefore, we estimate spatial and non-spatial panel models from 2004 to 2022 for Türkiye. A general-to-specific methodology is applied for selecting an appropriate model to determine spatial interactions of the variables by using the panel data at the level of 81 Turkish provinces. The Ordinary Least Squares (OLS) estimation results from the non-spatial model partially validate the growth theory as the study does not find evidence of absolute convergence for government expenditures. The results, however, confirm the conditional convergence for all variables. The Maximum Likelihood (ML) method is applied for the estimation of the dynamic and static spatial panel models to explain their spatial interactions. The ML findings are consistent with the OLS results. Moreover, unlike direct and total effects, it is not possible to define indirect effects explaining spatial spillover effects in the short and long terms.