Wealth and Assets Survey (WAS) Wave 5: Validation of figures against external sources

Income data from Wave 5 has been compared with external sources such as the Department for Work and Pensions (DWP) Households Below Average Income (HBAI) and Family Resources Survey (FRS), The ONS’s Annual Survey of Hours and Earnings (ASHE) and HM Revenue and Customs’ (HMRC) income statistics. These covered financial year ending (FYE) 2014 and FYE 2016

The following table displays figures from the Households Below Average Income (HBAI), covering FYE 2016, and the equivalent (or closest matching) Wealth and Assets Survey (WAS) variables for Wave 5.

The following table displays figures from the FRS, covering the financial year ending (FYE) 2016, and the equivalent (or closest matching) WAS variables for Wave 5.

Gross and net employment and pensions income have been divided by the sum of all the income variables which are measured on a personal basis. Income is typically measured on a household basis, where the household measure includes equity release, but this is not included in the personal figure. Equity release makes up a very small proportion of household income.

Gross measures of personal income have been used where available, otherwise net measures have been used. However, the main contributors to total income (employment, self-employment, pensions and investment) are all on a gross basis. The value of elements where net figures have been used are comparatively small.

A possible contributing factor to differences between the median net equivalised household income figures in HBAI and WAS could be that “net” income in HBAI deducts council tax/domestic rate payments, student loan repayments and parental contributions to students living away from home. None of these factors are deducted from income to derive the “net” figure in the WAS. We would therefore expect the HBAI figure to be lower than the WAS figure.

Gross and net employment and pensions income have been divided by the sum of all the income variables which are measured on a personal basis. Income is typically measured on a household basis, where the household measure includes equity release, but this is not included in the personal figure. Equity release makes up a very small proportion of household income.

Gross measures of personal income have been used where available, otherwise net measures have been used. However, the main contributors to total income (employment, self-employment, pensions and investment) are all on a gross basis. The value of elements where net figures have been used are comparatively small.

A possible contributing factor to differences between the median net equivalised household income figures in HBAI and WAS could be that “net” income in HBAI deducts council tax/domestic rate payments, student loan repayments and parental contributions to students living away from home. None of these factors are deducted from income to derive the “net” figure in the WAS. We would therefore expect the HBAI figure to be lower than the WAS figure.

The Annual Survey of Hours and Earnings (ASHE) 2016 release gives a point estimate of a variety of variables for April 2016. It also covers all of the UK, whereas WAS Wave 5 covers Great Britain for the period July 2014 to June 2016. A comparison of the two sources is displayed in the following table:

With the exception of the median earnings of full-time employees in the South East, Wave 5 of the WAS displays lower values than ASHE than all variables compared here. However, because ASHE data is based on a snapshot of earnings taken in April of the reference year, it does not include some seasonal work which may only be available in winter or summer, and may be fairly low paid. We may therefore expect that median earnings reported in ASHE would be higher than those in WAS.

Notes: Income

1. HMRC figures measure gross income that is liable to income tax. This may therefore exclude income from those with low paid jobs who are not liable to income tax.

Property data from Wave 5 has been compared with several external sources. These include the housing indices from the ONS, Halifax and Nationwide, as well as other data from the English Housing Survey (EHS) and Family Resources Survey (FRS).

While figures from Nationwide and the Halifax are from time periods slightly different to Wave 5 of the Wealth and Assets Survey (WAS), they along with the ONS’s house price index display similar values, ranging from £195,391 to £208,135 for the average price of a house in the UK. However, WAS shows this figure as being £280,405. This is far higher than the other three estimates, although it should be remembered that WAS figures are based on self-valuation, and therefore subject to the respondent’s perception of the value of their property, which may be higher than the true figure. However, this is still a useful indicator as a household’s perception of their property value may influence their financial behaviour.

The EHS covers England and Wales in its estimates, therefore the WAS figures listed here for comparison purposes omit Scotland. The following table shows figures which cover this area from both sources.

Further figures on this subject are provided in the Family Resources Survey from the financial year (FYE) ending 2016. The following table displays figures from the FRS and WAS.

SA data has been compared with statistical releases from HMRC which cover the financial year ending 2015. This can be observed in the following table.

For ISA statistics, Wave 5 of the Wealth and Assets Survey (WAS) estimates show the number of respondents who have an ISA, whereas HMRC’s figures show the number of ISAs subscribed to over the financial year ending (FYE) 2015.

Pensions data from Wave 5 of the Wealth and Assets Survey (WAS) has been compared with variables from the Family Resources Survey covering the financial yearending (FYE) 2016.

We have also investigated the cause of a 19% rise in the reported value of cars, vans and motorbikes from Wave 4 to Wave 5. Data from SMMT (Society of Motor Manufacturers and Traders) indicated that there was a rise in new car registrations between 2014 and 2015, before declining slightly into 2016.

However, Wealth and Assets Survey (WAS) data indicates that the percentage of respondents who owned a car, van or motorbike only rose from 74% to 75% between the two waves. The difference has therefore been caused by an increase in the value of these vehicles, indicating the effect of price inflation.

Wave 5 is the first wave where the values of cars, vans and motorbikes have been collected as one, rather than the individual values of each vehicle type collected separately.

Wave 4 contained a very wealthy individual in the sample. The individual’s wealth was so large that it made a noticeable impact on the aggregates of some variables, particularly within financial wealth and income.

However, this individual was no longer in the sample in Wave 5. It would appear that falls in the aggregates of some variables can be (at least in part) explained by this. For example, net investment income (WAS variable DVNIINV) decreased by 25% (from £42.3bn to £32.0bn) between Waves 4 and 5. Omission of this individual from the sample meant that the weighted sum for Wave 4 decreased to £32.2bn.

The following table demonstrates the impact this individual had on Wave 4 aggregates.