What structural deficit?

Ralph Musgrave asks a good question: is the structural budget deficit a useful concept? I agree with him that it’s not.
The structural deficit is the one that would exist, if output were at its trend level. This means that its measurement depends upon three unknowns: the size of the output gap; the response of spending to moves in GDP; and the response of revenues thereto. All these are uncertain, as the OBR acknowledges:

Forecasts of cyclically-adjusted aggregates are subject to particular uncertainty, as they depend on projections of the current position of the economy relative to trend. They also rely on analysis of the effect of the economic cycle on borrowing from previous cycles, which may not hold in the future.

To get a rough idea of these uncertainties, let’s do some sums.
The OBR uses the Treasury’s procedure (pdf) for estimating the structural deficit. This says that a 1% output gap, after two years, causes current spending to rise by 0.5% of GDP and causes revenues to fall by 0.2 per cent of GDP.
Now, the OBR estimates that the output gap was 1% in 2008-09 and 4.1% in 2009-10, and that cyclical adjustment reduces the deficit from 7.6% to 5.3% of GDP.
But let’s say the output gap is one percentage point larger in both years, and that revenues and spending are each 0.1 percentage points more sensitive to the output gap.
These small tweaks reduce the deficit by a further 1.5% of GDP this year, to 3.8%.
This means that the fiscal tightening which Labour put in place is sufficient to cut the cyclically-adjusted deficit to just 0.1% of GDP by 2014-15, on the OBR‘s numbers. Which eliminates the need for a large fiscal tightening.
There’s another reason I don’t like the idea of a cyclically adjusted budget balance. It’s that, as Ralph says, the budget balance depends upon the savings of the private sector.
To see the problem, imagine output were on trend - that is, higher than it is now. This could be because the private sector  goes on an investment boom, and its investment exceeds savings. If so, the public sector will have a surplus, as the counterpart of this. By definition, this will be a cyclically adjusted surplus. However, it could also be that output is strong because of strong overseas demand and UK firms save the profits generated by this. In this case, the private and foreign sectors together might have a financial surplus, so the government will have a cyclically-adjusted deficit. But on both cases, ex hypothesi, fiscal policy is the same.
Cyclical adjustment, then, does not necessarily tell us whether fiscal policy is loose or tight.
So, could it be that the notion of a structural deficit is a pseudo-scientific concept, intended to give a cloak of respectability to what is, in essence, a narrowly political desire to cut public spending? For more on this, see this paper by Randall Wray and Yeva Nersisyan.

Comments

[Luis Enrique]

Are there any circumstances in which you would say the deficit is too big? In your view, is it ever possible for an economy to get into trouble by having its government (that issues debt in its own currency) spend too much, tax too little, and borrow too much? I mean, if not, reductio ad absurdum, why don't we set the deficit equal to infinity?

If there are some circumstances in which we may be borrowing to much, how do you describe those circumstances without reference to notions like a structural deficit?

You can try to shift it on to "the market" and talk about interest rates on gilts, but that just relocates the problem. How are market participants deciding the rate at which they are willing to lend to the government, without thinking about something akin to an expected structural deficit?

[Econoclast]

There are a couple of issues here.

First, you are talking about an accounting identity (private plus government equals foreign balance). The relationship between the sectoral financial balances implies no causal link. You say the budget deficit depends on the private sector, but you could equally say the private sector balance depends on the budget deficit (and the foreign balance). In your first example, the private sector might be running a deficit and so might the public sector, with the overseas sector filling the gap. This describes the situation for much of the decade up to 2007. It's a moot point whether the private sector and/or the public sector were borrowing too much and whether one caused the other or not.

Second, economists use concepts like the structural deficit knowing their shortcomings. But this doesn't invalidate them. For instance, the output gap (the ultimate unknowable) is a key input into any assessment of the appropriateness of monetary policy. We can't observe it directly, but we can make an estimate and draw some conclusions. Without these imperfect concepts, policymakers wouldn't be able to gauge the appropriate stance of policy at all.

[chris]

@Luis - there are loads of circumstances in which I'd support deficit reduction: if our debt were higher (its one of the lowest in the G7); if gilt yields were high; if the deficit coincided with a strong economy. Etc.
@ Econoclast - yes, there are times when the budget deficit causes a private surplus. But surely, now is not one of them. The private surplus reflects the credit crunch, not Ricardian-style worries about future taxes, surely?
And yes, economists are wary of precise estimates of the structural deficit. But discussion of this matter has spread outside of sensible economists. I'm not aware that Osborne or the deficit hawks have discussed the margins of error surrounding estimates of the structural deficit when they've talked about it.

[Luis Enrique]

Chris,

yes, but why would high debts matter, why would gilt yields be high etc. unless some concept similar to that of a "structural deficit" was meaningful?

As far as I can see, all possible answers to those questions smuggle in some notion of "it looks like the government is going to struggle to pay this back", which tantamount to coming up with some cyclically adjusted estimate of the deficit.

If you don't think it's meaningful to think about likely scenarios and conclude "hmm, I expect the government to still be running a stonking deficit even if the economy recovers" how on earth can you make any notion of "expected probability of repayment" operational? That's all estimates of the s.d. are doing.

You write that you'd support deficit reduction if debts where "higher", but "higher" only means anything relative to expected future deficits/surpluses. If the future surplus if high enough, no debt is too high, if expected future deficits are big enough, no debt is too low. As I wrote, you can't duck the question by pointing to gilt yields, because the question then becomes what determines gilt yields if not, among other things, the expected s.d.?

It's one thing to say that estimates of the structural deficit are very unreliable and that the deficit is not under the govt's control etc., but I don't think you can escape having to form some expectation that amounts to forecasting the structural deficit, and non-numeric informal judgments aren't any better than crude IFS style forecasts, probably worse. OK, I'll grant you pseudo scientific if you mean specious accuracy, but useless concept? No.

[Luis Enrique]

If you think the probability of the UK government being able to service its debt is >0, you have decided that an average deficit of 80% of GDP is unlikely over 2011-2021. That means you have formed some expectation of the structural deficit.

If you're a government minister, who when asked "can I have £20bn to build a network of nuclear powered rollerdiscos in every town?", answers "sure thing; money's not a problem" then you do think structural deficit is not a meaningful concept.

[Luis Enrique]

aargh! missing "not" in last para

[Econoclast]

Chris

My point was that I think we have to be wary about inferring a precise causal link between the private and public sectors as sensible people can disagree about this.

I share your concern that politicians have hijacked the 'structural' deficit for their own purposes without understanding/explaining what it means. But Keynes did warn us about this.

Gordon Brown was arguably the worst offender (by fiddling around with estimates of potential growth and the output gap). What estimates of the structural deficit can tell us is that Mr Brown allowed the fiscal position to deteriorate too much during the upswing, which limited the scope for discretionary easing in the recession. Had Mr Brown been more responsible five years ago, he would have had greater scope for truly Keynesian policies to counter the cycle and we wouldn't be in the mess we're in now with the prospect of arbitrary Tory cuts.