A comparison of four model specifications for describing small heterogeneous space†time datasets: Sugar cane production in Puerto Rico, 1958/59–1973/74

Data forming a short space†time series – too short to utilize a STARIMA model – can include a random effects term that is spatially structured in order to account for both serial and spatial autocorrelation. In this context, space†time heterogeneity can be accounted for in various ways, including specifications involving spatial filtering methodology. This paper summarizes comparisons of four model specifications – simple pooled space†time; sequential, comparative statics; temporally varying coefficients with a spatially unstructured random effects term; and, temporally varying coefficients with a spatially structured random effects term. Implementations are illustrated with annual sugar cane production data for Puerto Rico during 1958/59–1973/74.