Join our community of smart investors

The myth of the structural deficit

Sure, the thinking behind it is reasonable. The idea is that recessions naturally increase government borrowing, by increasing spending on unemployment benefits, and by reducing tax revenues. But how can we be sure that this is the only reason why government is borrowing? This is where the structural budget deficit comes in. It is the amount the government would borrow, if the economy were operating at its trend level.

To measure it, we first estimate how far output is below its trend level - the output gap. Then we estimate the sensitivity of taxes and spending to GDP. We then multiply these sensitivities by the output gap. The structural budget deficit drops out.

There are, though, four problems with this.

1. We don't know what the output gap is. Economists agree that the recession has permanently destroyed some potential output. This means the output gap is less than it would be if the recession were merely due to weak demand. But how great is this destruction? The Treasury puts it at 4.5 per cent of GDP*. If this estimate is too high, then the output gap is larger than the 6.5 per cent it estimates. If it's too low, the output gap is less than this. This matters, because the larger is the output gap, the larger is the cyclical component of the budget deficit, and so the smaller is the structural element. The problem is, it'll be years before we know what the output gap really is. (If indeed, the idea of an output gap is meaningful at all, but that's another story.)

2. There's an anchoring problem, such that it will often seem as if we have a structural deficit when the government borrows a lot.

Let's do a quick sum. The Treasury estimates that public sector net borrowing in 2009-10 will be £178bn, or 12.6 per cent of GDP which stands at £1409bn, with £498bn of receipts and £676bn of spending. It also estimates that output is 6.5 per cent below potential. So, what would borrowing be when output returns to trend?

Let's say taxes rise 1.5 times as fast as GDP, as growth brings some people into jobs, and others into higher tax brackets, and because some temporary fiscal measures, such as the cut in VAT, cease. Revenues then rise 9.8 per cent (1.5 x 6.5), to £547bn. And let's say this growth saves us £20bn in welfare benefits and temporary spending measures such as the car scrappage scheme, so spending drops to £656bn. The deficit is therefore £109bn, or 7.3 per cent of GDP (This, remember, has grown 6.5 per cent from £1409bn to £1500bn). This is our structural deficit.

Sounds big? Yes, but this is a simple arithmetic effect of the fact that we're starting from high spending and low taxes and so applying reasonable growth rates to both naturally leaves us with a big gap.

We get the same problem, in reverse, when the government runs a surplus. In 1989 and 2000, it looked as if the government had a structural surplus, because applying reasonable-looking growth rates to high receipts and low spending yielded continued surpluses. We know now that those calculations were mistaken. So why should we be confident about similar calculations today?

3. Policy and the cycle are not separable. The notion that they are is an ideological fiction; it assumes that the government is merely a wise manager, resistant to political pressures.

One problem here is that economic growth, far from reducing public spending, might increase it. It might increase pressure to raise spending. This would happen if a healthy labour market made it hard to for the public sector to retain and attract staff, in which case it would need to raise pay. Alternatively, prosperous voters might ask why they see "private affluence, public squalor", and so demand better schools and hospitals; remember, increased government spending was popular 10 years ago, for precisely this reason. If the government responds to such pressures as the economy grows, is the budget deficit increasing for structural reasons. or is it just that the cyclical improvement in the deficit is less than thought?

4. The government doesn't have full control over its borrowing. Its borrowing is instead the counterpart of private sector net lending.

Imagine two different ways in which the economy might be on trend. In one, there's a boom in investment and housing, so the private sector is borrowing a lot. The government therefore runs a surplus. In the other, the economy is doing well because cash-generative businesses are thriving. Companies are therefore saving out of high profits, so government is borrowing.

In our first economy, it'll look as if the government has a structural surplus. In the second, it'll look as if it has a structural deficit. But in fact - ex hypothesi - the difference between the two economies lies in what the private sector is doing, not the public sector.

And this is no mere theoretical construct. I've just described the difference between the late 90s and mid-00s.

For these reasons, I don't think the idea of a "structural" budget balance is at all helpful.

It's not even useful in helping us on the key issue: will bond markets continue to want to lend to the government?

It's quite possible for there to be a large structural deficit, and yet bond markets are relaxed - if, say, debt is low or if investors expect the government to take strong action to cut the deficit. Equally, it's possible that the deficit is cyclical, but that bond markets are scared by it - if, say, they expect the recession to last a long time, or if the deficit adds to already-mountainous debt.

I fear, then, that the idea of a structural deficit serves a political rather than analytical function. It's a pseudo-scientific concept which serves to legitimate what is in fact a pure judgment call - that borrowing needs cutting. By all means, make this call. Just don't think that talk of a structural deficit helps enlighten us.

* This - equivalent to £1000 for every man, woman or child in the UK - is the true cost of the banking crisis.