Most government surveys are designed to provide statistically reliable estimates at national or high-level geographies. So when statistics are required for more detailed geographical areas or small subgroups of the population, the sample sizes just aren’t big enough to make reliable estimates.
Increasing the size of samples would be prohibitively expensive – instead, estimation methods have been developed that combine data from administrative, census and survey sources to produce estimates for small areas or domains. There are many statistical techniques covered by small area estimation – a frequently used approach is a model-based one, where local area outcomes are estimated from the regression between survey data and auxiliary data from census and administrative data sources.
The “Small area model-based income estimates” pages on the Neighbourhood Statistics website provide the average household income for small areas within England and Wales. The user guide within these pages gives details of how ONS make model-based estimates of income for middle-layer super output areas.
The following Eurostat-funded projects are a good source of material on small area estimation:
Support can be provided via ONS’s Methodology Advisory Service (MAS) , who offer advice and guidance across the public sector. Work is undertaken on an individual project basis with services charged on a cost recovery basis. For more information, contact MAS.
The following training opportunities are available:
GSS Statistical Training Programme (1-day course)
MSc in Official Statistics (taught once every 2 years, also available as a short course) (small area estimation taught once every 2 years, also available as a short course)