Abstract
The role of household projections as a basis for forecasts of households at national and sub-national level is discussed and a number of criteria for such projections are outlined. The projection method used by the Department of the Environment is examined in the context of these criteria and it is concluded that it is both practical and robust. However, it is open to criticism, first because of its failure to make the best use of the available data and of theoretical knowledge, and secondly because of its ‘black box’ nature. An alternative two-stage strategy is developed. The first stage involves constructing projections using a new curve-fitting method which takes account of within cohort life-cycle headship rate changes. The second is a method of analysing the resulting projections by modelling transition rates between different household states. Worked examples of both methods are presented.

This publication has 2 references indexed in Scilit: