Modelling the spatiotemporal distribution of the incidence of resident foreign population
Article first published online: 24 NOV 2011
© 2011 The Authors. Statistica Neerlandica © 2011 VVS
Volume 66, Issue 2, pages 133–160, May 2012
How to Cite
Marra, G., Miller, D. L. and Zanin, L. (2012), Modelling the spatiotemporal distribution of the incidence of resident foreign population. Statistica Neerlandica, 66: 133–160. doi: 10.1111/j.1467-9574.2011.00500.x
- Issue published online: 12 APR 2012
- Article first published online: 24 NOV 2011
- Received: April 2011. Revised: August 2011.
- generalized additive model;
- incidence of resident foreigners;
- soap film spline basis;
- spatiotemporal maps;
- tensor product smoother
Many European countries have recently experienced a substantial increase in the proportion of immigrants in their populations. The incidence of resident foreigners calculated at a national level does not provide information on the local spatial and temporal distribution of the phenomenon. This information may be of crucial importance for planning local policies. In this article, we suggest a tool for practitioners to provide spatiotemporal maps representing the local distribution of the incidence of resident foreigners in the territory, and changes in spatial trends over time. We illustrate this with Italian data at a municipal level, for the period 2003–2008. To account for spatiotemporal interactions in the data, we propose using a generalized additive model incorporating a smoother of the time and space dimensions. Specifically, we set up a tensor product smoother combining a cubic regression spline basis for time and a soap film spline basis for space. This approach provides a consistent framework to produce spatiotemporal maps which could be effectively used by policy makers to decide the allocation of economic resources at a local level.