Robert Chote: Never mind the rule, decisiveness would be golden: the Chancellor should go for tax rises

To get the public finances back on the right path, tax increases worth at least £11bn, or £430 per household, are needed

Ed Balls, until recently Gordon Brown's right-hand man at the Treasury, warned last week that forecasting the public finances is a "dangerous game". Alas, forecasting is unavoidable if policy is to be set or analysed sensibly; the danger is in making and explaining policy decisions without acknowledging that all forecasts are fallible.

Ed Balls, until recently Gordon Brown's right-hand man at the Treasury, warned last week that forecasting the public finances is a "dangerous game". Alas, forecasting is unavoidable if policy is to be set or analysed sensibly; the danger is in making and explaining policy decisions without acknowledging that all forecasts are fallible.

This is why the Institute for Fiscal Studies Green Budget, published last Wednesday in collaboration with Morgan Stanley, stresses the need to take greater account of the uncertainty that surrounds all forecasts for the public finances - even the Treasury's - in decisions about the future path of taxes and spending.

The idea is not radical. Indeed, it underpins the Government's own monetary policy regime. In explaining interest rate decisions, the Bank of England talks explicitly about the uncertainty surrounding its forecasts for the target measure of inflation. By helping people understand how rates are likely to respond if and when inflation diverges from the target, this bolsters the credibility of the Bank's actions while avoiding unrealistic expectations of what policy can achieve.

Forecasting the public finances matters in part because the Government's tax and spending decisions are constrained by a pledge to borrow only to finance investment. This "golden rule" implies that the current budget - revenues minus non-investment spending - should be kept in balance or surplus on average over the economic cycle, which the Treasury defines for now as the seven years from 1999-2000 to 2005-06.

Over the first five years, the current budget has swung into the red as tax revenues have weakened and borrowing has financed higher current spending. But the surpluses early in the cycle outweigh the more recent deficits. The Chancellor predicted in December that the current budget deficit would fall to £12.5bn this year and £6.9bn in 2005-06, meeting the rule with a cumulative £5bn to spare.

Mr Brown insists on this basis that there is no question of the rule being broken. But forecasting government borrowing is very hard to do with accuracy. The deficit is the small difference between two huge numbers: revenues and spending. A small error in predicting either can push the balance way off course.

To be more precise, the Treasury's average error predicting government borrowing just a year ahead is £12bn - larger than the margin by which it expects to meet the golden rule over the whole seven years. Based purely on its latest forecasts and past forecasting performance, the Treasury has about a 60 per cent chance of meeting the golden rule over the current cycle. Better than evens, but not quite the shoo-in Mr Brown implies.

Economists at the IFS are somewhat more pessimistic. Our best guess is that the current budget deficit will be £15.9bn this year and £13.4bn next, which means that the deficits over the cycle would outweigh the surpluses and the golden rule would be broken by around £5.5bn. But, like the Treasury's, our forecasts are fallible. Even if we are right about the size of the deficit this year, based on past forecast errors there would still be a one-in-three chance of meeting the rule with a year to go.

The bottom line is that the golden rule is likely to be met or missed by a relatively small margin. Whether the Chancellor succeeds or fails by a few billion either way should have little impact either on the economy or on the goal of fairness between the generations that the rule is designed to achieve. At the end of the day, the public finances are in pretty healthy shape.

But Mr Brown has staked more of his credibility on meeting the rule over an arbitrarily chosen period than either its analytical foundations or the Treasury's forecasting prowess would seem to justify.

Whether or not the rule is met in the short term is now largely in the lap of the gods. More important is whether today's tax rates and spending plans are consistent with meeting it further ahead. Mr Brown is confident: he expects the current budget to begin the next cycle in balance and then move into surplus. But we think revenues may be weaker and see a better-than-even chance that the budget will still be in deficit in 2008-09.

Whether this means the golden rule will be met or missed over the next cycle is hard to judge because of a further uncertainty: we do not know how long the next cycle will last. For that matter, we are not even sure if the current cycle is going to end this financial year or next.

Rather than try to predict the next cycle, we can say what we think the Chancellor would need to do to pay for his existing spending plans and get the public finances back on the path he was hoping for in last year's Budget. This would presumably give him the same confidence in meeting the fiscal rules in future that he was looking for last year. The answer, as we said last week, would be to announce new tax increases worth at least £11bn, or around £430 per household per year.

As the future is uncertain, the Chancellor has to weigh up the case for making such an adjustment now (or more plausibly after the election) against that of waiting to see whether we and other independent forecasters are being unduly pessimistic. If he waits and we are right, the eventual belt-tightening will have to be harder, more disruptive and politically costly. If he acts and we are wrong, he can cut taxes, increase spending or reduce the debt burden. This balance of risks merits careful thought.

Why raising debt need not be like raising hell

If the Green Budget is correct that borrowing is going to be higher over the next few years than the Treasury expects, the Debt Management Office (DMO) will need to sell more gilts (government IOUs) than in recent years to raise the money.

Fortunately, the amounts involved are likely to be modest in comparison to the financing needs of the big eurozone governments.

In addition, pension funds and insurance companies in the UK are likely to have a strong demand for low-risk, fixed-income, sterling assets.

Both these suggest that the Government is unlikely to have to offer higher yields to finance the extra debt, keeping down the cost to the taxpayer.

But this leaves open the question of what sort of debt the Government should issue. The DMO has adopted a simple and transparent strategy, issuing a range of debt of different maturities (or lifespans until repayment) with relatively little of it index-linked to provide insurance against future inflation.

Analysis in the Green Budget by David Miles and colleagues at Morgan Stanley suggests that the DMO should consider issu- ing a higher proportion of long-maturity and index-linked gilts than it does at the moment.

One reason is that this will reduce some of the risks the Government takes on when it issues debt. As it requires funding over the long term, issuing long-dated debt reduces the risk that it will have to be rolled over later at a time when real interest rates are temporarily high.

Meanwhile, issuing index-linked gilts fixes the real cost of servicing the national debt, which makes sense as the tax revenues used to service the debt are themselves linked to the strength of the real economy.

A second reason to issue more long-dated and index-linked debt is that this is where demand from institutional investors has been strongest and where yields have been driven down furthest - so it would be cheaper for the Government.

So, far from having to trade off cost against risk in its financing strategy, this looks like a win-win.

Another change to debt management strategy that has been mooted in recent months is the issuing of "longevity bonds", with payouts linked to life expectancy. This would help pension providers insure themselves against the possibility of unexpected increases in life expectancy and thus rises in the cost of pension commitments.

But whether the Government should seek to take on this risk is open to debate. After all, it already faces higher costs itself if life expectancy rises unexpectedly - for example, in running the health service and paying the pensions of public sector workers.

This suggests that it might be preferable for the Government to do what it can to encourage the private sector to issue longevity bonds, rather than to do so itself. Options here might include providing greater safeguards in the Pension Protection Fund for pension schemes that insure against some of their longevity risk.

Robert Chote is director of the IFS.