Cohort longitudinal estimation of housing careers

The housing life cycle is a time-honoured approach to longitudinal conception of people's housing careers, but one difficult to measure accurately. Cross-sectional data often provide a misleading 'snapshot' of changes over time; yet, panel data that follow individual observations over the life course also have several drawbacks. Construction of cohorts from cross-sectional data provides a useful third alternative, a quasi-panel approach, that captures major benefits of both the cross-sectional and panel approaches while avoiding their major faults. This paper describes three examples of fallacies of interpretation stemming from cross-sectional analysis of housing careers. One involves the stagnant home-ownership rate in the LIS and the apparent paradox of declining ownership rates of young households in contrast to rising ownership among the elderly. The second is the expected decline in housing consumption as the large baby boom generation ages past middle age. The third is the expected decline in residential overcrowding that could occur when immigrants become more settled in the US. The alternative, cohort longitudinal interpretation is presented to correct each of these fallacies.