This paper investigates the estimation and inference of spatial panel data models in which the regression coefficient vector is a trending function. We use time differences to eliminate the individual effects and employ GMM estimations for regression coefficients with both linear and quadratic...