This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
The DCLG household projections are produced using a two-stage approach.
Stage 1 produces projections using long-term trends in household formation by age, gender and marital status drawing data from 1971, 1981, 1991, 2001 and (limited) 2011 Censuses.
The projections methodology uses time-series modelling which weights together simple and dampened logistics trends.
The assumptions about household formation are applied to trend-based projections of household population derived from ONS data.
► Assumptions about institutional population
► Marital status projections
Stage 2 splits the aggregate household projections from Stage 1 into household types using information from 2001 and 2011 Censuses.
► Estimated headship (and non-headship) rates are projected forward using a modified two-point exponential model - similar to the methodology used in other home countries.
► 2011-interim sub-national population projections
► Census 2011 key statistics and quick statistics
► Labour Force Survey data
Approach agreed with DCLG and Steering Group was to remain as methodologically consistent with the 2008-based projections as possible whilst making the most of the new data available.
DCLG 2011-based household projectionsData inputs and rationale
At the time of producing the 2011-based household projections Census 2011 data on households by age, gender and marital status was not available
The Stage 1 methodology used Labour Force Survey data to inform how household formation has changed by age band since the Census 2001
LFS data supported anecdotal evidence that household formation rates for younger age groups had increased more slowly than previously projected or declined more rapidly than previously projected.
Phase 2 splits the household projections from Phase 1 into detailed household types.
The projections are split by type and age (and gender for one person households) generated using headship rates generated from bespoke Census tables of which equivalents were not yet available for 2011.
Estimates for the each household type were generated from available Census 2011 data – largely for the totals for each household type not split by age.
Disaggregated using previous estimates for 2011 calibrated to 2011 household population data by age and gender.