How much are UK households really saving?

16 Jun 2016

The household saving ratio measures the share of household income that is left unconsumed and is therefore available to fund consumption in the future. In the UK, the official household saving ratio reported by the Office for National Statistics has fallen significantly in recent years, from 11.6% in 2010 to just 4.2% in 2015. The reason is simple – over the last five years, household spending has been growing at around 3.75% per year while income growth has lagged behind at 2.1% per year. This suggests British households are squirreling away less for the future in order to spend more today.  But to what extent does this reflect the whole story?
The saving ratio is actually one of the least well-defined statistics across countries making international comparisons difficult. For example, most health spending in the US is attributed to the household sector while in Europe it is largely government spending – which will tend to push down the US saving rate relative to Europe. The point is – the saving ratio largely reflects definitions of household consumption and income.
The official UK measure is fairly standard. The saving ratio equals gross saving as a proportion of total household resources – which is disposable income plus an adjustment to reflect the change in assets of pension funds. However, would different definitions change the picture of household saving in the UK? Here I use some back of the envelope calculations to construct two alternative measures.
The first picks up on some interesting work recently published by the ONS which looks at the household saving ratio on a ‘cash’ basis. Not all elements of household spending and income which make up the ‘National Accounts’ measure of the saving ratio are directly observable to households. Excluding these would give an idea of cash savings made by the household sector – i.e. the difference between cash resources and cash spending.
The three biggest omissions in moving from the National Accounts to the cash measure are:
1. Imputed rents from housing – these are the implicit rents that owner-occupiers pay to themselves to live in their own houses. This is included so that National Accounts measures across countries, where rates of owner-occupiers are different, can be made more comparable. Although imputed rents are quantifiably important because they affect household incomes and spending in the same way the impact on the saving ratio is actually pretty neutral
2. Employers social contributions – these are payments made by employers into pension and insurance schemes held by their employees
3. Income payable on pension entitlements – these are the earnings arising from the investments made by pension funds that ultimately accrue to the household sector (e.g. defined contribution schemes)
As Figure 1 shows, on a cash basis the saving ratio follows a similar trend to the National Accounts measure – but is altogether lower. In fact, the interesting take away from the chart is that household saving on a cash basis has been negative for the last three years (-1.9% in 2015 compared to +4.2% in the National Accounts). So if it wasn’t for pension funds and employer social contributions the household sector would have saved very little over the last two decades.

 

Figure 1

Source: Office for National Statistics and author’s calculations
Low saving rates suggest UK households are accumulating wealth slowly. However, this is in stark contrast to what is actually happening. Between 1997 and 2015 total wealth increased from £3.1 trillion to £9.8 trillion, far in excess of what would have been achieved from the simple accumulation of household savings (see Figure 2). Household wealth has been propelled upward by the value of residential dwellings. According to ONS, average house prices rose at an average of 8% per year from 1997 to 2015 from £76K to £280K. In contrast, household resources, of which labour income makes up three-quarters, increased at an average rate of 4% per year. As a result, the multiple of household total net wealth relative to annual resources (incomes) increased from 5.6 in 1997 to 7.8 in 2015, although it should be noted that household wealth took a big hit in 2002 when the dot com bubble deflated and in the 2008 as a result of the global financial crisis.
Figure 2

Source: Office for National Statistics and author’s calculations
But what does this build up in wealth mean for the saving ratio. The National Accounts measure assumes the change in wealth of households equals gross saving or the accumulation of new financial assets. If changes in wealth also reflect capital gains or the revaluation of existing assets there is an argument that these capital gains should be added to both household resources and saving (unconsumed resources).
As Figure 3 shows this massively changes the arithmetic of the saving ratio. The accumulation of wealth over the last two decades from the rising value of housing and financial assets completely dominates the flow of saving from unconsumed income. The fall in stock markets and house prices connected with the global financial crisis meant 2008 was an annus horriblis for UK households, but these losses have since been regained and exceeded.
Figure 3

Source: Office for National Statistics and author’s calculations
Whether or not to include capital gains in measures of saving is a controversial issue. The current system of National Accounting rejects the idea – mainly because most capital gains are virtual or unrealised (i.e. capital gains or losses are only booked when an asset is sold). This does not fit well with the basic ethos of National Accounting which is based on a production framework – because capital gains and losses cannot be associated with the creation or disposal of productive assets. Some economists also reject the idea on the basis that capital gains and losses within a sector are simply transfers of wealth that net out to zero. For instance, higher house prices simply transfer wealth from existing future owners to existing home-owners, but leave the future supply of housing services available unchanged.
Those in favour of including capital gains typically point to three arguments. First, changes in wealth do seem to have an impact on household consumption behaviour in the aggregate, even if they are not realised gains. Most empirical models of consumer behaviour today find an important wealth effect, which has become more significant as financial innovation allows equity release or collateral to affect spending and saving. Second, in an open economy, higher domestic wealth increases command over foreign resources (imports) so can still raise a country’s future consumption possibilities. Third, many point to the possible contradiction in the existing National Accounts framework where capital gains are excluded from household resources, but capital gains taxes levied on the sale of financial assets are still deducted.
As a macroeconomist who spent seven years working on the National Accounts it’s an issue that has driven me around in circles – and one I still haven’t been able to make my mind up on. Any thoughts would be most welcome.

Share on LinkedIn:

Get in Touch

Your information will only be used to contact you about this enquiry, it will not be passed on to any third parties and will be stored in line with GDPR. You will not be added to an email marketing list.

Thank you for getting in touch.

We will respond to your enquiry as soon as we can.