The Bank of England's Monetary Policy Committee (MPC) has described recent weakness in consumer spending as "a puzzle". We believe that far from being a puzzle, recent consumer behaviour has been a textbook example of the impact of a tax rise that was large, unanticipated and permanent. If we are right and consumer spending remains soft, the MPC will resume their "rapid reaction policy" and reduce interest rates further; we expect a 1 per cent cut by summer.
Last year saw the introduction of self-assessment for income tax, a system devised by the last Conservative government. This led to a massive jump in income-tax payments, far larger than anyone (including the Treasury) expected. In first quarter 1998, around pounds 28.7bn was paid in income tax, a rise of almost 40 per cent on the previous year. Some of this was improved compliance, as people who had not previously filled in tax returns owned up to undeclared income. Yet it was generally believed that most of the extra receipts resulted from speeding up tax collection from those who paid income tax outside the PAYE system, including Britain's 3 million self-employed; many people paid tax for two years at once. Distinguishing between these two explanations is important because the latter would be a one-off implying that tax receipts would fall to normal levels this year.
January and February are the two big months for these income-tax payments (the deadline for sending payments is 31 January). We now have data for those two months, which means we can make a fairly good guess at the total for the first three months of this year. As the chart shows, it is now clear that tax payments this year have been maintained at the new, much higher level.
As a result of last year's surge in income tax, household real disposable income fell by 1.4 per cent in first quarter 1998, despite rising wages and employment. It is very unusual for real household income to decline by this amount. Previous declines of this magnitude have only occurred in the midst of severe recessions and oil price hikes. For 1998 as a whole, real household income was unchanged, the worst performance since 1982 and only the seventh year in the past 50 when real incomes have not risen. Given the data on tax payments for January and February, another big decline in real household disposable income looks likely in the first three months of this year.
So what was the overall impact of self-assessment on the consumer last year, and what are the implications for 1999? Income tax is, of course, just one of many influences on consumer spending. When a consumer's income changes the impact on their spending depends on whether the change was anticipated and whether it is believed to be temporary or permanent. It would seem that self-assessment was largely unanticipated, initially believed to be temporary but now likely to be seen as permanent. If this is true, it fits the stylised facts of the consumer slowdown well. Because it was unanticipated the impact was delayed, with increased credit demand and lower saving taking up part of the slack. As the tax hit turned out to be larger and more enduring than first expected, it has taken time for the impact to build. Indeed it will probably take most of the rest of this year before the full effect on growth in consumer spending has worked through.
The impact of last year's unexpected drain on the consumer was exaggerated by the contrast with 1997, when households received windfalls amounting to pounds 30bn. The distinction between temporary and permanent is particularly important here. Standard economic theory suggests that consumers save the bulk of windfalls. But buying a car or other consumer durable counts as part-saving in this theoretical context because it leads to a stream of consumer "services". Having increased spending on durables in 1997, the natural counterpart would be less spending in 1998. The initial impact of self-assessment can be seen in this context as a negative windfall, exacerbating the downturn in spending. Indeed, spending on durables rose by almost 10 per cent in 1997; the growth rate peaked at 16 per cent in first quarter 1998, just as worried taxpayers were writing out huge cheques to the Inland Revenue, and has since collapsed to minus 0.7 per cent. We expect durable spending to fall further and overall consumer spending to remain soft.
Another way to look at this story focuses on the saving ratio; the fraction of their income consumers do not spend. This declined in 1998 from 9.5 to 7 per cent, the lowest since the Lawson boom of the late 1980s. While the decline then reflected excessive consumption associated with a housing boom, last year's fall in the saving ratio was a classic reaction to an income squeeze as consumers tried to maintain spending. Indeed, because consumers were caught out by the tax rise and took some time to respond, consumer spending grew by 2.8 per cent. The weakness the MPC described as puzzling could have been a good deal worse.
Looking forward, 1999 will see consumers trying to rebuild their saving ratio. Since the income tax rise has been sustained, this will be reflected in weaker consumer spending growth. Evidence that consumers remain reluctant to spend can be seen in the recent GfK/European Commission consumer confidence survey. Although confidence has improved as fear of recession has receded and interest rates fallen, consumers are reluctant to spend on big-ticket items; a net 6 per cent expect to reduce such purchases this year.
Where does this leave the MPC? One of the reasons they decided to keep rates on hold at this month's meeting was because they were puzzled by the consumer and may have felt that spending might start to pick up again. They will have been surprised by the data showing the scale of tax payments in February - their own internal estimate was for a much lower figure. With unemployment set to rise and wage inflation slowing, the renewed consumer squeeze should mean that they can cut rates by another percentage point by the summer.
Steven Bell is chief UK economist at Deutsche Bank