“…Elhorst (2014), on the other hand, argues that the parameters of the other specific spatial models, such as the spatial Durbin model (SDM), are identifiable and free from the problem of overfitting. Consequently, we follow Harari and Ferrara (2018) and Hossain and Ahsan (2018) by controlling for spatial correlation in the regressors and dependent variable using the spatial Durbin model (SDM) and accounting for spatial dependence in the residuals via clustering standard errors by MPAs.…”