The birth cohort born in 1967 turned twenty-two in 1989. Some were graduating from university, while others had been in the labour market a few years. Most aspired to starting families and owning their own homes. In the United Kingdom, these twenty-two year olds faced a housing market in which average prices had been rising for seven years, and had risen 70% in real terms in the last four years. The ratio of average house prices to average earnings was 5.5. By contrast, when the cohort of 1975 turned twenty-two in 1997, house prices were more than 20% lower than in 1989. Incomes had been catching up with prices, so that the house price to earnings ratio was 4. In short, this cohort faced a very different housing market than the cohort that turned twenty-two eight years earlier. Do these differences matter?
These differences may matter both in the short run and in the long run. In the short run, cohorts faced with difficult housing market conditions may, on average, be delayed in ‘getting on the property ladder.’ Perhaps even more seriously, these differences may also matter for the longer run home ownership rates of a cohort. Some members of a cohort that is delayed in its initial ownership transitions may find that they are never able to make the transition to owning their own home, and the ownership rate of the cohort may never ‘catch up’ to that of cohorts that faced more favourable initial conditions.
Recent theoretical modeling (Bottazzi, Low and Wakefield, 2007) suggests that housing market conditions early in a cohort’s housing career matter in the short run, but not in the long run. Simulations indicate that disadvantaged cohorts catch up, so that they have comparable home ownership rates as they approach their fifties. However, these simulation results are sensitive to modelling choices, and so an empirical assessment of these questions remains important.
In this paper we provide such an empirical assessment. We use the repeated cross sections of the Family Expenditure Survey/Expenditure and Food Survey (FES/EFS) from 1969 to 2007 to answer two questions. First, as each birth cohort reaches adulthood how are their transitions to home ownership affected by market conditions? Specifically, if we compare two cohorts, one that faced a property boom in their twenty’s and one facing a property slump, how different are their ownership at age twenty-five or thirty? Second, how persistent are the resulting differences? That is, do the home ownership rates of these two cohorts converge at older ages?
Our analysis employs cohort or “pseudo-panel” methods. The FES/EFS is not a true panel, in that individuals (or individual households) are not followed over time. However, because the FES/EFS provides us with a representative sample of the population in any survey year, it also gives us a representative sample of each birth cohort in any survey year. Thus we use the repeated cross-sections of the FES/EFS to track birth cohorts over time.
