A tutorial on fiscal surpluses

According to Ross Gittins in the weekend Herald,

There’s been a lot of debate – and confusion – over the right way to assess the degree of stimulus the budget will impart to the economy and how this may affect interest rates.

Well, he’s dead right about the confusion. But unfortunately this is one of the rare occasions when Gittins adds to the confusion rather than resolves it. He wants to make two points about the stance of fiscal policy, but gets them jumbled up in the exposition. The first is that we need to focus on the structural balance of the budget. The second is that, if we’re interested specifically in the stimulus imparted, it’s not the balance itself, but changes in the balance that matter.

The concept of the structural balance arises from the fact that, as Gittins notes, the budget balance in any given year is affected not just by the government’s decisions about how much to spend and what tax rates to set, but also by ‘cyclical’ factors, that is, the state of the economy. The cyclical deficit improves when the economy booms because we earn more, pay more tax, and collect less unemployment benefits; it deteriorates in a slump for the reverse reasons. The structural, or ‘cyclically adjusted’ balance is the surplus or deficit that would be recorded at some given level of economic activity — corresponding, say, to a particular rate of unemployment. In the absence of cyclical influences, any changes in the balance must be discretionary ones. When this balance improves, the government is reducing the level of demand; when it deteriorates, it is adding to the demand.

And this is where we need to stress that it’s the change in the balance that matters, not its level, that is, whether it’s in deficit or surplus. Suppose for argument’s sake that there is no business cycle – that all changes in the balance are discretionary. If the fiscal balance deteriorates from a big surplus to a smaller surplus, fiscal policy is still ‘looser’ in the sense that it’s increasing total spending from the current level.

Gittins redeems himself by making one indispensibe point about this particular budget:

Mr Costello’s official figures show an expected underlying cash surplus of $13.6 billion for 2006-07, and a planned surplus of $10.6 billion for 2007-08. The decline of $3 billion is equivalent to 0.3 per cent of GDP – which isn’t big enough for the Reserve to worry about. However, we should adjust these figures to take account of Mr Costello’s year-end fiddles, adding $3.7 billion back to the old surplus and taking $2.6 billion off the new surplus (the difference between the two gets the super co-contribution out of the comparison). And, since Mr Costello performed a similar fiddle in last year’s budget, we should remember to subtract $900 million from the old surplus.

Two other things are are worth pointing out. First, just because fiscal policy has been loosened doesn’t mean that total demand is necessarily in fact rising. The fiscal changes may be offset by changes in other types of spending that just happen to be occurring at the same time — a fall in investment, consumption or exports, or a rise in imports. The strong dollar is a reason why the latter might happen. The second point is probably obvious to most people: even if total spending is rising, this is going to be inflationary only if the economy is already operating at full capacity, so that the extra demand causes labour shortages bottle necks in particular markets. That does seem to be the case at present, but it can’t be assumed in all circumstances, and no-one knows exactly where the inflation threshold is.

Finally, we’ve only considered the direct or first order impact of the fiscal shock. The discretionary changes in government spending and tax may will almost certainly affect other components of expenditure. When the government spends the extra money and the households spend the income that they don’t have to fork out as tax, this may put more people in jobs, increase their income, and induce additional spending (the multplier effect). At the other extreme, when government spending increases, housholds might respond by cutting their consumption because they know they will have to pay for that at some future date and they have to budget for it now (a ‘new classical’ contention). But the point is that in any case, the amount of the changes in government spending and tax collection in themselves are a poor guide to the total change in spending that result from the fiscal policy.

The bottom line is that anyone claiming to have a formula that tells you how much the interest rate will need to rise based on the change in the fiscal stance, let alone the change in the actual budget balance, let alone the magnitude of the actual surplus or defict, is a fraud and a charlatan.

Finally, a terminiological question for anyone who happens to care about these things. I’ve been avoiding using the terms expansionary and contractionary, because there doesn’t seem to be enough agreement about how to use them.

Suppose the fiscal balance, governed by the last budget, is in structural surplus, and then a new budget is introduced, with a smaller but still significant surplus. Is this expansionary or contractionary fiscal policy?

Fiscal policy is supposed to be expansionary if it increases total demand. But, increases relative to what? If we mean relative to the previous state of affairs, then the fiscal policy in the example is expansionary, because it involves a discretionary increase in spending or decrease in taxes. By this interpretation, it doesn’t make sense to refer to a ‘contractionary fiscal stance’, because only changes in the stance can be contractionary or expansionary.

Alternatively, an expansionary fiscal policy might be one that increases the level of spending realtive to what would occur if the budget were balanced. By this interpretation, every structural deficit would be expansionary, even if the deficit was the same, year after year. Describing the situation in the example, we would say that the contractionary stance has been maintained.

I’d be interested to hear what people think, but in the meantime I favour the first terminology, for two reasons. The first relates to the context in which we are normally discussing these things. Most of the time, commentators are interested in anything that might affect the inflation rate, because that’s the RBA’s target in the setting interest rate. If the surplus is reduced but stays a surplus, then at full employment this will add to inflationary pressure. It would be awkward to say: fiscal policy has been loosened but it remains contractionary. Tight, yes, but contractionary, no.

Second, it’s actually not true that a balanced budget is necessarily neutral in its effect on total spending, as anyone will know who encountered the concept of the ‘balanced budget multiplier’ in undergraguate macroeconomics. If the government spends an extra billion and raises an extra billon in tax to cover it, this will most likely cause a net increase in aggregate spending, since households would not have been planning to spend the whole billion on consumption anyway — they would have saved some of it.