Spatial Econometrics

Our objective was to empirically quantify the effects of spatial spillovers on outputand input-specific technical inefficiency in Dutch arable crop farms. First, a non-para-metric directional distance function was computed using DEA to estimate technicaland scale inefficiency scores for output, productive inputs and damage abatement inputs (pest control spending). Second, a spatial econometric model was estimated which incorporates regressors for spatial lags of farm characteristics alongside other non-lagged explanatory variables and time-period fixed effects. We use both the inverse distance weight matrix and the binary k-nearest-neighbours weight matrix and also estimate the distance cut-off and the optimal number of neighbours rather than imposing rules of thumb.