Measuring change in household and individual income is a key aim of many social surveys and social policy analyses. The latest official data shows a small rise in income for the latest available year, after a number of years of stagnating or even declining incomes.
The ‘average’ individual in England has a weekly net income of £491 before housing costs and £418 after housing costs (accounting for different household sizes - a process known as ‘equivalisation’).
However, what do we know about the picture at the local level? Social survey data is available for regions and, unsurprisingly, shows higher incomes in regions such as London and the South East, and lower incomes in the East Midlands and the North East. Until recently, there has been very little data held at lower geographical levels.
The Census has never included income questions, because of fears about both accuracy and potential negative implications for response rates. A number of solutions have been used to overcome this constraint.
The ONS has produced some ‘model-based’ estimates by trying to estimate what local income distribution would be given the information we do have in the Census and other sources. Receipt of social security benefits has been particular important in this, as it is counted at the local level. For this reason, benefit receipt is used in the calculation of ‘income deprivation’ for the Index of Multiple Deprivation.
Perhaps most significantly, the ONS has now started to experiment with using national data held by HMRC on employment and occupational pensions. The first local area income estimates are being produced, though they don’t take account of tax and housing costs (see the two examples below):
Proportion of households in Greater London with equivalised gross household incomes of less than £15,000 per annum (Data: Office of National Statistics)
Proportion of households in the South West with equivalised gross household incomes of less than £15,000 per annum (Data: Office of National Statistics)
However, while this has been happening, the private sector has been busy in the collation of modelling of data about local income – this is done to help clients design marketing and retail strategies.
The National Centre for Social Research (NatCen) is working with the Consumer Data Research centre to explore the potential of some of this commercial data on local income for academic research. The CDRC is one of the ESRC Big Data centres and already holds a treasure trove of data – such as ‘Fitbit’ data and GPS data from bike hire schemes.
Our research will be used to compare the various measures of income in the UK at the local level, including that held by retail and consumer data consultancy organisations. The aim is to produce a more comprehensive picture of geographical patterns of income in England.
We will also be exploring how a better understanding of the local income distribution can inform other areas of social policy; for example, to build a model of house prices and to assess housing affordability in England.
‘Big data’ is the new frontier for social science, and it is important to be open to new data sources and approaches – from private as well as public sectors.
NatCen is hoping to play an important role in this agenda in the coming years, both in exploring new data sources, but also generating new research questions and theories.