The Australian and US NGDP Gaps
How tight is Australian and US monetary policy?
|Stephen Kirchner||May 31, 2020|
Ahead of this week’s March quarter national accounts release, I thought it would be useful to take a look at the US and Australian nominal GDP gaps. The Mercatus Center has been promoting the concept of the NGDP gap as a measure of the effective stance of monetary policy, assuming the role of monetary policy is to stabilise nominal income around its trend growth path. Regardless of the monetary policy targeting regime actually in place, nominal income stabilisation is a reasonable benchmark for judging the effective stance of monetary policy. David Beckworth makes the case for NGDP level versus inflation rate targeting here.
The Mercatus gap is based on the deviation of NGDP from the long-term forecasts of professional forecasters. Here I suggest a different approach based on Kamber, Morley and Wong’s modified Beveridge-Nelson filter. Despite the different methodology, the underlying concept is similar in trying to measure deviations in the level of NGDP from a long-horizon conditional expectation.
The NGDP gap for the US is shown below. It suggests US monetary policy has been systematically tight since the financial crisis of 2008, but the trouble begins even earlier, coincident with Alan Greenspan’s departure from the Fed in January 2006. The Greenspan years are a Golden Age for US monetary policy, with the gap mostly very close to zero, with the obvious exceptions of the early 90s and early naughties recessions, but these gaps are closed reasonably quickly (for my narrative defence of Alan Greenspan, see here).
Australia sees a similar Golden Age from 1996 to 2016 under RBA Governors Macfarlane and Stevens, except that policy also errs on the side of being too tight after the peak in the terms of trade boom in 2011.
The signal-to-noise parameter (delta) KMW use to trade-off the fit and amplitude of the cycle in the case of the US and Australia is remarkably similar, with a delta of 0.32 for the US and 0.28 for Australia. In other words, about 32% of the shocks to US NGDP and 28% of the shocks to Australian NGDP are permanent. Small, open economies could be expected to experience more transitory shocks, but the US and Australia do not look dramatically different on this measure.
The real GDP gaps for the US and Australia have deltas of 0.24 and 0.11 based on Kamber, Morley and Wong’s research. The real output gap deltas are smaller than the nominal, presumably because many price level shocks are permanent. Australia differs from the US more in in terms of temporary real than nominal shocks. Note that changes in the terms of trade can be viewed as real income shocks. The RBA would be kept busier stabilising the real output gap than the nominal. The standard deviation in Australian NGDP gap since Q1 1993 (when the RBA moved in the direction of inflation targeting) is 1%, only slightly higher than the US at 0.9%. The estimated NGDP gap for Australia includes a structural break in Q1 1993.
One implication these measures is that the terms of trade volatility objection to Australia adopting nominal income targeting may not be as severe as many assume. It should be recalled that the floating exchange rate moderates the effect of movements in the terms of trade, so the burden of nominal income stabilisation does not fall entirely on monetary policy.
The RBA may have to work harder to stabilise nominal incomes than the Fed given a larger share of temporary shocks and somewhat greater volatility in the NGDP gap. Maybe that’s why RBA Governors are paid considerably more than their Fed counterparts (to be clear, I think Fed Governors are grossly underpaid). But it is not obvious that nominal income targeting is a considerably worse fit for Australia than for the US.
I had an op-ed in The Canberra Times on Australia-China trade frictions.
Donald J. Boudreaux defends financial markets against attacks from a conservative think-tank.
The Australian labour market is showing signs of stabilising: