Fitting and testing a ‘predator‐prey’ model*

Abstract

Students of political competition and of social influence processes have long recognized the insights afforded by a particular nonlinear, dynamical formalism commonly known as a “predator‐prey” model. Yet, to date there remain relatively few research reports in which this model has been fit to actual observations and used to predict behavior. This is largely attributable to the fact that the predator‐prey cannot easily be fit to observations using such common estimation procedures as ordinary least‐squares. This paper presents a relatively simple method of fitting the predator‐prey to observations. And it illustrates the method using data on Finnish parliamentary elections.