Properties and Estimation of a Production-Constrained Alonso Model

Abstract

The Alonso spatial interaction framework encompasses a wide variety of movement phenomenon models; most notably, the ‘family’ of spatial interaction models. The production-constrained formulation of the Alonso framework incoporates a destination-related relative accessibility effect and therefore generalizes the traditional production-constrained interaction model. In this paper, the properties and estimation procedures for a production-constrained Alonso model are presented. Two related estimation approaches are developed: an ordinary least squares (OLS) procedure and a maximum likelihood (ML) procedure. The results of estimation experiments for these methods provide important insights for operationalizing the Alonso framework. It is demonstrated that the Alonso parameters can be recovered from a simulated data set if the exogenous weights are measured correctly. However, errors in the measurement of the ‘true’ exogenous weights in the data set result in incorrect estimates of parameters.