What today’s household income figures tell us about 2016-17

Published on Incomes and Inequality

The best results can take time. We already have data for what happened to employment, prices and earnings in early 2018, but for a detailed look at household incomes the latest official data – out today – is for 2016-17. What does this release tell us about living standards, poverty and inequality back when The Donald first became President Trump and Article 50 was yet to be triggered?

1) Typical incomes grew healthily in 2016-17

Today’s ‘HBAI’ results show that real median household disposable income was 1.9 per cent higher in 2016-17 than in 2015-16. This is broadly in line with results from a separate ONS survey published earlier this year, showing growth of 2.3 per cent.

This strong real growth can partly be attributed to the fact that prices in the financial year as a whole were only 1.1 per cent higher than in 2015-16 (compare that to a recent inflation high of 3.1 per cent). Thanks to this low inflation and rising employment, 2016-17 can be seen as part of a ‘mini-boom’ for typical incomes, even though inflation rose over the course of the year itself.

Source: HBAI

2) Child poverty continues to fall in absolute terms and rise in relative terms

Due to the size of the survey on which this data is based, few of today’s statistics are considered significant when looking at single year to year changes, but they are important when seen as part of continued trends.

For instance, thanks to growing incomes child poverty in absolute terms (the proportion below a fixed real-terms income threshold) has fallen again.

But incomes at the bottom did not grow as fast as typical incomes, and as a result relative child poverty (the proportion below 60 per cent of median income) has risen for the fifth year in a row, from a low point in 2011-12.


Note: 2016-17 values shown here are estimates based on rounded changes since 2015-16. They represent the smallest changes that are consistent with the DWP figures. When more detailed data becomes available, relative poverty could turn out to have risen slightly further in 2016-17 than shown here and absolute poverty to have fallen slightly further. GB before 2002-03.
Source: HBAI and IFS

3) Inequality has been broadly flat since the financial crisis, but remains high

Although the change is not considered significant, inequality in the latest stats was lower in 2016-17 than in 2015-16, whether looking at incomes before housing costs have been deducted or after. The broad picture (on both measures) is one of fairly flat inequality between 2010-11 and 2016-17, though the Chancellor clearly remains wide of the mark with his earlier statement that “income inequality is at its lowest level in 30 years”.

Note: Values for 2016-17 are rounded.
Source: RF analysis of HBAI and IFS

4) The data needs further improvement

As we’ve said before, the HBAI statistics are currently the best resource we have for keeping track of household incomes. One advantage of these statistics is that they adjust the data of the highest income people to correct for known underestimation of their incomes. But we have also noted that this adjustment – while far better than nothing – also has its flaws. Due to the delay in filing and processing tax returns, in today’s 2016-17 results the incomes of the richest are in fact projections – based on outturn tax data from 2014-15 and various OBR estimates from March 2017 (two fiscal forecasts ago). This is not ideal.

We know top incomes were boosted in 2015-16 by people’s efforts to take dividends before an April 2016 dividend tax rise: and conversely their incomes in 2016-17 were somewhat depressed. This has a big effect on inequality and income figures for both 2015-16 and 2016-17 – in fact, today’s figures suggest that mean (rather than median) income fell by 0.8 per cent in 2016-17. However, neither year’s HBAI results are based on outturn tax data. And more recent estimates from the OBR suggest that self-assessed incomes (including dividends) performed better in 2016-17 than it had forecast in March 2017.

There is a trade-off between the timeliness and accuracy of data, so it’s understandable that even the best household income statistics are based partly on out-of-date projections. But, less justifiably, there are no plans to ever revise these top income figures once full tax data becomes available. And previous academic work has shown that this lack of revision has important implications for the accuracy of some results.

At the other end of the distribution, we also know that a lot of benefit and tax credit income is missing from the survey data – meaning that many people’s incomes are underestimated. Given these known problems at both ends of the income spectrum, the DWP and ONS should be looking to use the government’s own tax and benefit records to further improve their household income statistics over the next few years.

5) Income prospects for 2017-18 and beyond don’t look so good

On the whole, 2016-17 was part of a much-needed mini-boom for real household incomes. Unfortunately, everything we know from other economic statistics – including our detailed income modelling – suggests it was also the end of it. In 2017-18, average real wages fell and high inflation slashed the value of working-age benefits. And with most of the benefit cuts announced in 2015 still to bite, the living standards outlook for lower income working-age households remains weak.

Just as the political landscape has changed a lot since 2016-17, so too has the pace and distribution of living standards improvements.