English Housing Survey - Household Data
(formerly called the Survey of English Housing)


Main uses

The household data of the English Housing Survey (EHS) has two main purposes.  First, as its name suggests, it is a survey about housing in England.  In this respect, its focus is mainly on the households’ use of the house rather than the condition of the house itself, the main English survey on the latter being the stock data part of the survey.

Second, it is general household survey for England asking a number of questions about households and individuals, with a particular focus on their neighbourhood.

EHS is the English equivalent to the Scottish Household SurveyNational Survey for Wales and Northern Irish Continuous Household Survey.


In summary:

  • Available from: UK data archive, with the datasets prior to 2008/09 being available from a different part of the archive.
  • Registration required: yes.
  • First survey available: 1993/94.
  • Frequency: annual.
  • Updated: November.
  • Scope: England.
  • Format: SPSS.
  • Files: 24 files per year, most of them comprising a record per household. 1
  • Documentation: user manual.
  • Weighted or unweighted: weighted.
  • Household income data: yes.  Equivalised from 2008/09 onwards but unequivalised only in the earlier datasets.

The three main files from 2008/09 onwards are as follows:

File Contains data about
General basic data about the household (weight, geographic location, etc)
Interview the main household data from the interviews
Attitudes the household’s views about their neighbourhood

Prior to 2008/09, the main file was:

File Contains data about
Household all the household data

General issues

Which software to use

As the annual dataset is around 50,000 records for individuals and 20,000 records for households, it can be exported into Excel.

Which tables to use

The simple answer to this is to use all the tables that collectively contain the data required for the desired analysis.  Because most of the tables have a record per household, and cover the same set of households in the same order, they can be linked by simply placing them next to each other.  Failing that, the household serial number can be used for linking.

What weights to use

There is only one weight, namely a household weight.  What this effectively means is that it is assumed that the answers given by the particular individual being interviewed are representative of the views of the household as a whole.  By contrast, most other surveys have two weights, one for use when counting households and the other for use when counting individuals.

Relevant graphs on this website

UK graphs

Because the dataset is for England only, the graphs below are also for England only.

Indicator Table Graphs Comments
Ability to travel households second
Homelessness individuals last Requires a complicated analysis.
Overcrowding individuals second
In mortgage arrears households second
Dissatisfaction with local area households first Requires the use of unequivalised gross incomes, as equivalised net incomes are only available from 2008/09 oinwards
third The better/worse question is only asked in some years
fourth The precise questions asked vary from year to year.

Scotland, Wales and Northern Ireland graphs

These graphs are similar, but not identical, to the English graphs above and come from the following surveys:

1. Prior to 2008/09, there were only 5 files per year, with the household file being the main one.