Consensus pessimism about UK consumer spending prospects partly reflects the low level of the official household saving ratio measure. The accuracy of this measure, however, is questioned by other official data suggesting that saving has increased over the past year and is at a respectable level by historical standards.

The official measure of the household saving ratio fell further to 5.6% in the third quarter, the lowest since 2008 and well below an average of 8.4% over 1996-2015 (20 years). According to the consensus view, the ratio is unlikely to decline much further, so an expected inflation squeeze on real income growth in 2017 should result in a significant slowdown in consumer spending.

The official measure of the ratio is derived from the household income account, with saving defined as the difference between disposable income, adjusted for the change in pension entitlements, and consumer spending.

An alternative measure, however, can be calculated from the financial and capital accounts. This measure defines saving as new investment in financial or tangible assets minus borrowing. The two definitions of the ratio are theoretically equivalent but differ in practice because of measurement errors.

The first chart shows the two saving ratio measures plotted as four-quarter moving averages, to reduce quarterly volatility. The alternative measure based on the financial / capital accounts is usually higher than the official measure derived from the income account but the current divergence is unusually wide. In contrast to the official series, the alternative measure has risen strongly since the first half of 2015 and is close to its 20-year average – households, that is, have been adding to their net assets at a faster pace recently even while increasing consumer spending solidly.


The financial / capital account detail shows that the increase in the alternative saving ratio measure has been driven by stronger growth of household money holdings and faster accumulation of life insurance and annuity entitlements.

The Office for National Statistics prefers the income-based saving measure because it has greater confidence in the income account data. The recent rise in the alternative measure, however, accords with expectations that households would increase precautionary saving before and after the Brexit vote. The healthier financial position implied by the alternative measure, meanwhile, tallies with solid consumer spending growth. The rise suggested by the alternative measure is also supported by a recent fall in the net percentage of households planning to increase saving in the EU Commission monthly consumer survey: this percentage has been negatively correlated with the saving ratio historically (i.e. a higher current level of the ratio is usually associated with lower savings intentions) – see second chart.


If it is assumed that the alternative measure of the saving ratio is more accurate at present, the implication is that either consumer spending has been over-recorded or disposable income under-recorded in the income account. Supporting the latter possibility, a sizeable divergence has opened up recently between the net percentage of households reporting an improvement in their financial position over the past 12 months in the EU Commission survey and the annual rate of change of real disposable income – third chart.


If household saving has been higher than officially reported in recent quarters, saving of other sectors must have been correspondingly lower, since aggregate saving (including the financial surplus of the overseas sector) must equal aggregate investment. An examination of the financial accounts of other sectors suggests that the savings positions of the overseas sector and financial corporations are weaker than reported in the income accounts. According to the financial account data, for example, the overseas sector financial surplus (i.e. saving minus investment) was £71.9 billion in the year to the third quarter – significantly lower than the reported current account deficit of £93.4 billion.