It speaks volumes for the difficulty of keeping the public finances under control that even Gordon Brown, once mocked for his emphasis on prudence, is struggling with the gap between what the Government is spending and its tax receipts. The Budget earlier this year estimated that net borrowing would be £27bn, up from a forecast of £24bn this time last year. The rate at which spending has outpaced receipts since the summer suggest the figure is likely to be bigger, perhaps over £30bn. This will represent a huge swing from a surplus of 1.7 per cent of GDP in 2000-01 to a deficit which could be more than 3 per cent this fiscal year.
Forecasting government borrowing is hard at the best of times. The net borrowing total is the gap between two very large numbers, and the different components of the tax and spending totals can be very volatile from month to month and year to year. So the average error in the Treasury's one-year-ahead forecast is £10bn - and the Treasury does better than most outside forecasters.
The recent revisions to the GDP figures have also complicated the Chancellor's plans, as he prepares for next month's Pre-Budget Report. The pattern of revisions means the level of GDP is higher than we first thought. With the benefit of hindsight, therefore, the Treasury should have been more worried about the increase in net borrowing during 2002 and 2003, because less of it has been due simply to the cyclical downturn in the economy. There is still enough leeway for the Chancellor to meet his "Golden Rule" - which says the current budget should balance over the course of the whole cycle - thanks to the huge surpluses his prudence delivered during the late 1990s. But it won't be met as comfortably as anticipated.
If the bigger-than-expected borrowing numbers have not been caused by economic slowdown alone, what's the explanation? Even after the revisions, slower growth than forecast this year does account for some of the shortfall, but the main culprit seems to be public sector pay. Average pay growth has shot up during the past 12 months, and is running at 5 to 6 per cent, almost twice the current rate of increase in the private sector.
A catch-up by public sector workers was inevitable once the Government decided on big increases in public spending. Deteriorating relative pay over many years had hit recruitment in key services and morale across the sector. But there are two problems. One is that the Treasury does not seem to have budgeted for it. The "Red Book" from last March's Budget spells out pretty clearly that the forecasts were based on similar low rates of increase in public and private sector earnings. It predicts a rise of around 70,000 a year in public sector employment which, against a background of "favourable" earnings trends, is described as "achievable within cash spending plans". But both the numbers employed and pay rises during the past 12 months have been greater than this implies.
The second problem is more fundamen- tal. Pay increases that outpace inflation are not as big a concern if productivity is rising too, though they would still have to be funded. Unfortunately, there is no sign public sector productivity is increasing. If anything, it seems to have deteriorated, although measuring productivity in non-market parts of the economy is very hard. The nightmare for the Chancellor must be that the big rises in spending on health and education will not only make the sums in his Pre-Budget Report very tricky, but will not have an electoral pay-off either.
If it's any comfort to Mr Brown, he is experiencing a problem common to the advanced economies. In all of them, for at least five decades, the cost of frontline public services has risen faster than the general level of prices. In all cases, that has meant pressure on government budgets. As long ago as 1966, and in a more recent paper in 1993, William Baumol, professor of economics at Princeton, said there was very little scope for standardisation or mechanisation in services such as nursing and teaching, so productivity was bound to grow slowly. Yet nurses and teachers very reasonably expect their standard of living to keep up with people working in the rest of the economy, where there is significant productivity growth.
It's hardly surprising that these competing pressures spill over into higher budget deficits, by far the easiest way out of the dilemma. Yet it would be very bad for the economy to jettison prudence. A lasting increase in the level of government borrowing would lead to interest rates being higher. Recent research confirms that the move to big US budget surpluses during the Clinton era made a substantial difference to the level of long-term interest rates and the US economy's growth. Here in the UK we have enjoyed the benefits of stable growth, low inflation and low interest rates for long enough to appreciate how unwelcome a reversal would be.
Professor Baumol calculated that, by 2040, spending on health and education could account for over half of US GDP - which would be fine, reflecting increasing productivity in other sectors. In the round, economic growth and productivity gains mean we can easily afford to pay for the services we want. The problem is that the slow-productivity-growth services lie in the public sector, and citizens are not satisfied with standards. Hence the thorniness of public service reform.
In the short term - before the next election, that is - the Chancellor looks likely to meet his rules for government borrowing. After that, the long-term problem will come back with a vengeance.
Diane Coyle runs the consultancy Enlightenment Economics and is visiting professor at the Institute for Political and Economic Governance, University of Manchester.