Premium
Net transition probabilities: an approach to subnational level projections of households and housing demand based on census data
Author(s) -
McDonald Peter,
Kippen Rebecca,
Temple Jeromey
Publication year - 2006
Publication title -
population, space and place
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.398
H-Index - 68
eISSN - 1544-8452
pISSN - 1544-8444
DOI - 10.1002/psp.424
Subject(s) - census , net migration rate , projection (relational algebra) , econometrics , geography , transition (genetics) , net (polyhedron) , statistics , economics , mathematics , demography , population , sociology , population growth , biochemistry , chemistry , geometry , algorithm , gene
The authors set out to make medium‐term projections of housing demand for subnational regions of Australia. Projections of the numbers of households by household type and age of a household reference person were required. The only possible source of input data at a subnational level was the quinquennial censuses. An innovative approach was developed specifically for the purpose of these projections. This involved the use of ‘net transition probabilities’ estimated from net changes in household classification types for individuals between censuses in the past, which are then projected into the future. The results of the projection for the period 1996–2001 were evaluated against 2001 Census results. Net transition probabilities were found to be robust, in that their age and sex specific profiles were very similar from one intercensal period to the next. The evaluation concluded that the model used is both practical and highly effective. Overall, the authors found that the projections have a high degree of reliability and produce very useful information concerning the future demand for housing at a regional level. Copyright © 2006 John Wiley & Sons, Ltd.