The broad objectives of the project were to advance longitudinal research on housing in the following ways:
- explore how individuals move within and between areas and change tenure in response to changes in fertility
- examine the extent to which the relationship between housing and fertility varies between areas and cohorts
- compare the housing-fertility relationship in Britain and Australia
- consider a number of methodological issues in the analysis of panel data (choice of the unit of analysis for individual-level data when decisions are taken at the household level, attrition and left-censoring, selection on omitted variables affecting both housing and fertility decisions)
More specifically, we:
- Explored the effects of births and the age and number of children on changes in residential mobility and housing tenure, and consider the extent to which these relationships vary over time and space (for example, due to temporal and spatial variation in house prices).
- Developed multilevel dynamic models that allow for area effects on housing transitions, and additionally allow for selection on shared unobserved determinants of housing and fertility that operate at the individual and area level.
- Assessed the robustness of substantive findings to adjustment for common features of panel data, such as attrition and left-censoring, and selection on omitted variables affecting both housing and fertility decisions.