Minding the gap: how we’re improving our measurement of income inequality

ONS publishes widely used estimates of household income, including how much people earn, what benefits they receive and how much they pay in taxes as well as looking at the gap between the richest and poorest households. Here, Dominic Webber writes about the important work to improve the quality of these statistics and provide more detailed information on this much debated topic.

Today we have published our latest figures showing how taxes – both direct and indirect – and benefits affect income inequality. Alongside this we’ve also published our plans on how we plan to transform these statistics, using both improved survey data and ‘administrative’ data sources from across government to create world leading estimates of changing trends in inequality.

Income inequality

Income inequality is an important and widely reported statistic. Often small changes or differences between our measure and that produced by the Department of Work and Pensions draw considerable comment. As with many statistics it is better to look at the longer-term perspective. The figures show that income inequality has been pretty flat over the past decades. However, while it may have fallen slightly over recent years, it is still significantly higher than it was before the steep increases seen in the late 1970s and 80s.

But, while income inequality itself may not have shifted much in recent years, the way we are measuring it is. This is all part of our wider programme to transform statistics on household finances. Over the past few decades we’ve seen a range of surveys evolve and develop, largely independent from each other. As a result we find ourselves with some excellent surveys that do the job they are meant to do in isolation, but together can, and will, do so much more. Take for instance, the ONS Living Costs and Food Survey (LCF) of around 5,000 households which asks about income and expenditure. This is the principal data source for today’s ‘Effects of Taxes and Benefits on Household Income’ publication. It allows us to examine not only the effects of direct taxes, such as Income Tax, on household income, but also indirect taxes – such VAT, and alcohol and tobacco duty – by asking about household expenditure on goods and services. The separate Survey of Living Conditions (SLC) of 12,000 households, also measures income and other metrics related to deprivation. Finally, the Wealth and Assets Survey of 18,000 households measures how much wealth households have stored away in the form of property, pensions and savings.

Transformation plans

As part of our transformation plans published today we have announced the harmonisation of key questions on income and living conditions on the LCF and SLC and merged their samples to form a new Household Finances Survey (HFS). This important improvement means we will now sample 17,000 households when compiling our estimates of income – more than three times as many as before. This will allow us to produce headline statistics with significantly greater precision. It also means we will be able to publish estimates of income inequality for local areas and various population sub-groups. This data is currently being processed by our statisticians, with results covering the financial year 2017/2018 out early next year. In addition, the HFS currently ‘in the field’ has also been partially integrated into our wealth survey, which will increase our sample size even more for some of our core inequality statistics.

However, it is the introduction of administrative data where the real step change will occur. We have said many times how greater use of administrative data will transform our statistics. And perhaps nowhere will the benefit of this data be seen more clearly than income statistics. A lot of research has shown that that the incomes of households at the very top and the bottom end of the income distribution are often not well recorded in our surveys. Administrative data such as HMRC tax returns and DWP benefits data will really help us develop a clear picture of these households’ incomes. We will be starting work on these data soon and plan to introduce better adjustments by the end of 2019.

There is a long road ahead of us to produce world leading statistics on households’ incomes and how they are impacted by taxes and benefits but today’s announcement marks an important step in that journey.

Dominic Webber is Head of Household Income and Expenditure Analysis for the ONS