Abstract
The research in new economic geography (NEG) suffers from discrepancies between its theory and empirics due to the different ways that the geography is treated. This paper proposes modelling the mechanisms of spatial distribution of economic activity on the microgeographic level, taking the location of each firm into account. It extends the literature on spatial point pattern analysis in economics by deriving the spatial laws of motion for the firms in geographical space. The basic NEG model is expressed as a spatial point process and is fit to a realworld pattern of firm locations. After accounting for the inhomogeneity of the underlying space, the NEG-based model produces patterns remarkably similar to the observed one and captures the amount of clustering on all spatial scales except the very local.