A Research of the Regional Public Investments' Spatial Effect Between the Years 2002-2017 in Turkey

In this study, the spatial effects of public investments and the effect of the population density on public investments between the years 2002 and 2017 in level 2 regions are investigated. Both fixed and random spatial panel models, SAR (Spatial Autoregressive Model), SEM (Spatial Error Model), SDM (Spatial Durbin Model) and SAC (Spatial Autocorrelation Model) are used. For the selection of the fit model, AIC and BIC values are examined and the models with the lowest values are selected. By these values, it is concluded that the fixed effect spatial Durbin model (SDM) and the random effect spatial lag model (SAR) are more suitable models than the other spatial panel models. According to these models, it is determined that the public investment between the years 2002-2017 contains spatial effects (Also SEM and SAC models have similar results)