What’s in store for household incomes in 2015?

The 2015 election has been billed by many as the ‘living standards election’, meaning that we can expect much debate in the coming months on just how households are faring and whether economic growth is translating into income growth.

To date though, most attention has focused on wages instead of incomes. The six-year real-terms pay squeeze has been well documented, with wage increases failing to keep pace with inflation. There are signs that this squeeze has now ended, in large part because inflation has fallen well-below target, and official projections for 2015 – though subject to considerable uncertainty – point to a steady recovery in pay.

The picture for household incomes – the key measure of whether living standards are improving or not – is somewhat more nuanced, however. That’s because falling wages have been offset in part by increases in employment and by cuts in income tax. And wages form only part of the total. Non-wage income has been affected by investment returns for some and by welfare policies – such as the triple-lock on the state pension or cuts in working-age benefit generosity and eligibility – for others.

Crucially, the best direct sources of household income trends are significantly lagged. The DWP’s Family Resources Survey (FRS) captures directly reported incomes, but the data is released more than a year after it is collected. The latest figures show that average household incomes measured across the year to March 2013 were still falling in real terms, and were around six per cent lower than their 2008-09 peak. However, the period since March 2013 will remain a mystery until mid-2015. A similar but smaller household survey – the Living Costs and Food Survey (LCFS) – will be updated this side of the election, but even this will only provide details of what was happening early in 2014.

In the absence of a timelier version of this data, we can expect much pre-election attention to instead be given to the Real Household Disposable Income (RHDI) measure in the quarterly National Accounts. While statistically robust, the problem is that this indicator is potentially misleading when used as a measure of changing living standards.

Most obviously, it’s an aggregate measure that doesn’t account for population change. But even expressed on a per capita basis it’s not wholly appropriate for judging living standards, for two reasons.

First, the definition of ‘households’ includes several things that aren’t really households at all: non-profit institutions serving households such as universities, charities and trade unions. The second reason is that the deflator used to express incomes in real terms is a National Accounts ‘implied deflator’, which differs from the price indices – CPI and RPIJ – which we would normally use to deflate earnings and incomes. This implied deflator is entirely appropriate for the primary purpose of the National Accounts: measuring macro-economic trends. However, conventional inflation measures are more explicitly designed to capture price changes experienced by households, and therefore more appropriate for deriving trends in household living standards (though even these measures are flawed and fall short of being true ‘cost of living’ indicators).[i]

These distinctions are technical and little-understood but, given that RHDI per capita gets reported as reflecting changes in living standards, they are important. Having a timely picture of what’s happening to disposable household incomes takes on added importance in an election year. With this in mind, we’ve attempted to correct for these deficiencies in RHDI as best we can, to produce a single, regular indicator of ‘household income per capita’ deflated using CPI. Trends in this indicator and the other income measures discussed here are shown in Figure 1.

Figure 1:  Six years on, household incomes are still lagging their pre-crisis levels

Notes:  RHDI covers income of households and ‘non-profit institutions serving households’ (NPISH) including trade unions, charities and universities. It is deflated using the ‘households and NPISH expenditure implied deflator’. We derive the household income per capita from the non-deflated National Accounts ‘gross household disposable income’ measure. We make an attempt to isolate household income by looking at the split in expenditure between households and NPISH and applying this ratio to the overall income figure. We deflate using the CPI instead of the implied deflator, in order to better assess changes in living standards. The FRS household income measure captures income as reported in the DWP’s Family Resources Survey and is an average across households rather than per capita. The LCFS household income measure captures income as reported in the ONS’s Living Costs and Food Survey and is an average across households rather than per capita. The two deflators used have different weighting structures and price indices. CPI excludes some elements (such as the imputed rent of owner occupiers) that the National Account deflator includes. Equally, the National Account deflator uses some direct volume measures. Both are statistically robust, but CPI provides a closer picture of the actual changes in prices faced by households.

Source: RF analysis of ONS and DWP, Family Resources Survey

Stripping out non-households[ii] and changing the deflator suggests that average disposable household income fell from almost £17,760 per capita at the start of 2008 to a low of £16,940 at the end of 2013. At £17,070, the level in Q3 2014 remains 3.3 per cent below the pre-crisis peak, a significantly sharper reduction than the 1.8 per cent fall observed in the official RHDI per capita measure.

The chart also shows that trends in CPI-deflated ‘household income per capita’ are closer to those observed in the authoritative and directly reported, but less timely, survey measures of household incomes. However, even our adjusted measure appears to understate the size of the squeeze in 2013 relative to the surveys.

Some variance is natural given differences in coverage, but the magnitude of the gap in 2013 might imply that apparent improvements in the National Account figures in that year were driven by elements of income not readily recorded by households in survey data, such as ‘property’ (that is, dividends, interest and other investments) income or by employer social contributions (such as pension contributions and employer NICs). If so, then technical increases in incomes might not have translated into any feeling among households of being ‘better off’.

Turning to the future, we cast forward both National Account measures using OBR projections from the Autumn Statement. For our RF adjusted measure, annual growth amounts to a steady 1.7 per cent in 2015, back to the levels of growth recorded in the 2000s before the financial crisis.

Despite this growth however, average disposable household income would not return to its pre-crisis level until early-2016. This contrasts with a projection for recovery by Q2 2015 on the RHDI per capita measure. Given the political debate about living standards that we can expect in the run-up to the election, this distinction is likely to prove important (RHDI per capita is similarly projected to return to its 2010 election level in early-2015, while our adjusted measure would again not reach this threshold until early-2016).

Clearly though, all projections are subject to uncertainty. The OBR’s inflation projections in particular, despite being just a few weeks old, may quickly be overtaken by developments. And of course, these average measures will mask variation across households.

In general, the signs are that we can expect incomes to rise over the course of 2015, but the impact on voting intentions in the ‘living standards election’ will vary with individual circumstances. A key question is whether voters will compare their May 2015 incomes with their position in 2010 or with a more recent benchmark and prospective increases.

[i] For a discussion of the different coverage of the implied deflator and CPI, and the ways in which these two indices have diverged in recent years, see: Office for National Statistics, ‘The Reconciliation of the Differences Between the Consumer Price Index and the Implied Price Deflator,’ March 2012

[ii] There is no direct way of splitting household disposable income into separate household and NPISH components. Instead, we consider the split in final consumption expenditure between households and NPISH, which is available and apply the same ratio to the overall income measure. That is, household final consumption expenditure consistently accounts for between 95 per cent and 96 per cent of total final consumption expenditure. We therefore assume that households similarly account for between 95 per cent and 96 per cent of the total income measure. We acknowledge that this approach is imperfect, but the data does not allow for any other adjustment.