A Population-Employment Interaction Model as Labour Module in TIGRIS XL

Thomas de Graaff, Barry Zondag
(2013) In: Pagliara F., de Bok M., Simmonds D., Wilson A. (eds) Employment Location in Cities and Regions. Advances in Spatial Science (The Regional Science Series) . Springer, Berlin, Heidelberg, pp. 57-77

DOI Online

This chapter looks at the integration of a population-employment interaction model in the TIGRIS XL framework. TIGRIS XL model is an integrated land-use and transport model and is actually a system of sub-models (or modules) that allows for dynamic interaction between them. Currently, the population module responses to lagged changes in employment and the employment module responses to contemporaneous changes in population. Thus, the change in population mainly drives the change in employment—an assumption which, given the strict Dutch restrictions on population location, is not entirely unrealistic. We, however, argue that this assumption is too harsh, and that sectoral employment change might be more dependent upon employment change in other sectors than on population change. We, therefore, test and estimate such relations and conclude indeed that for some sectors the employment dynamics in other sectors are more important than population changes and propose an adapted version of the labour module in TIGRIS XL. The new methodology within TIGRIS XL is assessed by looking at a (stylized) case study. This study concerned the doubling of the size of the new town Almere located 20 km east of Amsterdam. About 60,000 dwellings are to be built and 100,000 jobs are to be added in the period up to 2030. The accessibility benefits of a particular land use planning variant, with a tailored public transport investment alternative, are examined for the new labour module in TIGRIS XL. Both models predict, in case of the construction of 60,000 dwellings, a number of additional jobs much lower than 100,000. The model results of the new module shows that the population-employment interaction model reacts slower and on shorter distances on population changes than the old model. Thus, employment does not seem to mold that easily to population changes as earlier Dutch employment models have predicted.