Fiscal multipliers and employment (wonkish)
With two weeks to go in the election campaign, we still haven’t seen anything resembling a budget proposal from Tony Abbott and the LNP. Various people have made estimates of the cost of his promises and the cuts likely to be needed to fund those promises and return to surplus. My main concern is that Abbott has locked himself so thoroughly into the rhetoric of surplus that, in the event of a downturn or recession, he will feel compelled to adopt the kinds of austerity measures that have had a disastrous impact in Europe and prevented any real recovery in the US. To make this point properly, we need some numbers. One way to get such numbers is with a macroeconomic model. That gives you some better precision, but often hides the key assumptions. Instead, I will give a very simple Keynesian analysis, yielding back-of-the-envelope estimates.
For illustration, I’ll assume a public expenditure cut of $10 billion a year – the calculation is linear so it can be scaled up or down as needed. In a recession, the fiscal multiplier is likely to be around 1.5 (that’s the value used by Christina Romer when she pushed for a larger fiscal stimulus in 2009, and consistent with recent estimates by the IMF). So, the impact of the cut, when multipliers are taken into account is $15 billion or around 1 per cent of national income (or GDP if you prefer that measure).
Now we can use Okun’s Law to estimate that the cut will raise the unemployment rate by around 0.5 percentage points. Taking participation rates into account, employment will also fall by around 0.5 per cent (about 50 000 jobs).
A bunch of qualifications and observations over the fold
This assumes monetary policy is unchanged. In practice, presumably there would be some relaxation of monetary policy. But in a recession and with interest rates already low, expansionary monetary policy is unlikely to have much effect. As my old teachers used to say, it’s ike “pushing on a string”.
This calculation assumes that the cuts aren’t offset by additional spending or tax cuts elsewhere. It deals with the idea of attempting to restore budget balance during a recession. When you are looking at a mixture of cuts and increases, what matters is composition effects. Changes in public sector employment are likely to have a bigger short-term impact than changes in taxes,particularly changes in corporate taxes.
Allowing for the scale of the economy, Newman’s cuts in Queensland are comparable to those used in the calculation above, and the increase in unemployment since then is consistent with the estimates. Of course there are lots of other things going on, but I conclude that Newman’s policies have increased unemployment both directly and indirectly through multiplier effects.
Finally, I haven’t had time to calibrate this for the 2008 & 2009 stimulus packages, but my preliminary guess is that they imply that GDP would have been 2-4 per cent lower and unemployment 1-2 percentage points higher in 2009 if not for the stimulus