Conception and use of an individual-based model of residential choice in a planning decision process. Feedback from an experimental trial in the city of Besancon, France

In this paper, we present an experimental trial involving the use of a complex spatial simulation platform to support a planning decision process in the city of Besancon (eastern France). In medium-sized towns across France, households with children are leaving to settle in periurban areas, and Besancon is no exception. Under those circumstances, the political objective of the Besancon City authority is to keep middle- and high-income households with children within the city. The simulation platform MobiSim was used in this context to explore the possible outcomes of spatial planning policies, set by both the Greater Besancon authority (Besancon and the surrounding periurban communities) and the city of Besancon, on residential migrations over a 20 year period. MobiSim integrates several models representing demographic, social, economic, and spatial processes. Modelled entities are individuals and dwellings. Rules determine how individuals form households. Other rules locate dwellings in buildings. The experimental trial began in October 2012. It involved two researchers in geography and planning from the University of Franche-Comte (Besancon, France) and two planning practitioners working in Besancon city council's planning service: the Department of Planning, Projects and Forward Planning. The first stage of the trial was the collective definition and simulation of a baseline position simply extending existing trends from 2010 to 2030: the `Business as usual' scenario. The second stage was a collective reflection on possible modifications to some of the variables and parameters of this scenario in order to simulate a pro-active policy of housing construction in the medium term. Under the conditions represented in the model, the `Pro-active housing construction' scenario in Besancon allows an overall reduction of migration flows from Besangon to the surrounding periurban areas. More single parent families choose to reside in Besancon during the whole simulation time. Couples with children, however, choose preferentially to reside in Besancon until 2022 only. At this date, the controlled housing developments become less numerous. This experimental trial gave the researchers and the planning actors involved in it the occasion to exchange their scientific knowledge and their empirical knowledge. The simulation of the `Business as usual' scenario was, as it were, a means to thoroughly explore one possible future. This then led the group to discuss possible planning actions designed to reduce migration flows of households with children out of Besancon into periurban areas. The trial also shows some limitations: it highlights that this kind of PSS makes the planning actors heavily dependent on the expertise of the researchers for every stage of the process. (C) 2015 Elsevier Ltd. All rights reserved.