Abstract
Household projections produced by official agencies usually use the headship rate method, based on the proportion of people designated as ‘head of household’ (or ‘reference person’) in various sex, age, and family categories. Reasons are discussed why this method has been criticised in recent years: the headship method does not incorporate any behavioural assumptions or dynamic aspects; only the characteristics of the head are included explicitly; and the concept of ‘head’ is thought to be incompatible with more egalitarian household roles.
The advantages of dynamic models are discussed, and a number of alternative dynamic models of household formation and dissolution are presented. The data source used is the Office of Population Censuses and Surveys 1% longitudinal study, which contains linked information on 500000 people in the 1971 and 1981 England and Wales Censuses of Population.
Get full access to this article
View all access options for this article.
