22 June 2010
A family preparing to move house
Professor Fiona Steele, in the Centre for Multilevel Modelling (CMM) at the Graduate School, will begin a new 3-year ESRC project in October that aims to advance longitudinal research on housing, working with Dr Michele Haynes (University of Queensland), Professor John Ermisch (University of Essex), Dr Paul Clarke (CMPO, Bristol) and Dr Hill Kulu (University of Liverpool).
Image by Mary R Vogt
Housing transitions - such as changes in housing tenure and residential mobility - are the outcomes of a complex history of other life-course events such as union formation and dissolution, birth of children, and changes in employment. Research on the effect of family events on housing changes is of particular importance because it provides valuable information on when and where families move, and identifies changing priorities for the provision of housing assistance and policy. The principal aim of the research is to examine the extent to which changes in housing tenure and residential mobility are triggered by fertility outcomes such as the birth of a (nother) child or a child reaching primary or secondary school age. They will also consider the effects of household structure (for example comparing lone parents and couples with children), union formation and dissolution and employment changes, as well as spatial and temporal variation in these relationships within Britain and cross-nationally between Britain and Australia.
We hope that the research will contribute to the evidence base on factors associated with individuals' ability to realise their housing goals, and provide information for policy-makers to help identify priorities for provision of housing assistance.
In order to investigate and quantify the relationships between housing transitions and stages in the life-course, longitudinal data are required to identify the timing of a life event such as pregnancy or the birth of a child relative to a residential move or change in housing tenure. The research will use data from the British Household Panel Study (BHPS) which provides detailed housing information collected annually since 1991. The cross-national comparison will be based on data from The Household, Income and Labour Dynamics in Australia (HILDA) survey, which began in 2001 and has a similar design to BHPS.
While prospectively collected event history data have undoubtedly advanced our understanding of the complex interrelationships between social processes, there are a number of important methodological issues to be addressed when analysing panel data, for example: (i) housing histories have a nested structure with repeated episodes (of living in the same tenure or house) nested within individuals; (ii) changes in housing and changes in fertility may be subject to shared or correlated influences, some of which are likely to be unmeasured; and (iii) life-course information is incomplete, for example because of events experienced before the start of data collection, attrition, and missing values on covariates. One of the aims of the proposed research is to investigate the robustness of empirical findings to changes in model specifications that adjust for these potential sources of bias.
Professor Fiona Steele said: “We hope that the research will contribute to the evidence base on factors associated with individuals' ability to realise their housing goals, and provide information for policy-makers to help identify priorities for provision of housing assistance. We will seek to publicise our research through the National Housing and Planning Advice Unit (NHPAU), a non-departmental public body sponsored by the Department of Communities and Local Government. The methodological aspects of the research will be disseminated to academic and government researchers engaged in longitudinal data analysis through training workshops and the development of on-line learning materials.“
Visit Fiona Steele's staff page
Find out more about the Centre for Multilevel Modelling
