This paper introduces the R library estimateW to estimate spatial weight matrices for Bayesian spatial econometric panel models. The approach focuses on spatial weights that are binary prior to row-standardization. However, unlike recent literature our approach requires no strong a priori assumptions on (socio-)economic distances between the spatial units. The estimation approach relies on efficient Bayesian Gibbs sampling techniques and the library supports a variety of the most common spatial econometric panel specifications. estimateW moreover supports to elicit flexible shrinkage priors, which allow to estimate spatial spillovers even in settings where the number of time period is small relative to number of cross-sectional units. An empirical illustration for European NUTS-1 regions demonstrates that the method recovers plausible spatial dependence patterns, interpretable spillover effects, and meaningful clustering in the estimated network structure.
