Electoral District Boundary Effects in Canada

Abstract This article introduces the Markov Riding Position Sampler (MaRiPoSa) for simulating Canadian elections under alternative configurations of electoral district boundaries. Despite Canada’s strictly non-partisan system of boundary design, MaRiPoSa demonstrates that, for a single Canadian election, there are alternative, allowable and more equitable configurations of boundaries to return either Liberal or Conservative governments without changing the vote or location of a single voter. If the same votes in the same places produce different governments depending on which valid configuration of boundaries happens to be used, to what extent can we say that it is voters and their territorial distributions that determine the outcome of Canadian elections? Canada will need to adjust principles and institutions to prepare for the new technological environment brought about by advances in computing and artificial intelligence.