On 12 February the Assistant Governor of the Reserve Bank of Australia, Sarah Hunter, gave an important speech on the relationship between the labour market and inflation. This speech suggests the RBA has re-thought its position and has partially gone back to an earlier orthodoxy, but in so doing it now holds untenable positions on such issues as the natural rate of unemployment. We first need a quick potted history of post-war macroeconomic thinking in Australia.
In the early post war period, a version of Keynesian economics reigned supreme. This model recognized that capitalist economies are inherently unstable and policy involved the management of aggregate demand so as to maintain full employment with low inflation. For many years unemployment was under 2% in Australia and when it went over that in 1961 the long-lived Menzies Coalition Government was reduced to a majority of one seat at the subsequent federal election.
The Phillips Curve, named after its inventor, AW Phillips of the London School of Economics at the time, was influential in positing a stable tradeoff between inflation and unemployment. This provided policy makers with an apparent choice between lower inflation or lower unemployment. For example, suppose say 3% unemployment was associated with 5% inflation, policy makers would read that from the Phillips Curve and then use aggregate demand to achieve that outcome.
People like Professor Ted Wheelwright criticised the notion that an inherently unstable capitalist economy could be managed by consulting the equations, setting the desired outcomes in the model and so managing the macroeconomy. He referred to the process as dial twiddling. Joan Robinson referred to “Bastard Keynesianism” being the neoclassical synthesis in which Keynes’s economics was a special case where prices and wages were sticky and the economy could be described in a set of linear equations that lent themselves to manipulation using demand management tools.
The Keynesian neoclassical synthesis was blown apart by Milton Friedman’s critique in his 1967 Presidential address to the American Economic Association. The early monetarism had suggested that if the money supply increased, then the price level would also rise by the same proportion. Real effects were not considered in the model that Friedman wanted to promote. However, the Phillips Curve was an accepted fact and a policy tool that Friedman needed to demolish. So he argued that, if left alone, the economy would converge to an equilibrium at the natural rate of unemployment – that rate of unemployment that cleared the labour market. Friedman argued that short term movements of unemployment around the central tendency of the natural rate were only possible if people were fooled into thinking a nominal increase in their prices/wages represented a real increase, in which case workers would offer to work more and so the measured unemployment rate would be lower. Hence if demand increases (increases in the money supply) caused prices and wages to increase, workers might mistake higher money wages for higher real wages. However, as soon as workers woke up to the fact that their real wages had not increased, their extra work offers would be withdrawn pulling the labour market back to the natural rate of unemployment. In this way a reduction in unemployment suggested by the Phillips curve would be temporary but inflation would have increased. If the policy makers still wanted lower unemployment they would have to increase demand so that inflation would be higher still and more workers would be temporarily fooled into offering more labour. In Friedman’s view any trade off between inflation and unemployment would only be temporary and, in the long run, the Phillips curve would be vertical. Notice that in this thinking when you can only reduce unemployment by tricking people, any unemployment must be voluntary which, of course, is contradicted by the everyday facts of unemployment. That this sort of thinking entered mainstream economics says a lot about mainstream economists.
In the years following the Friedman critique the natural rate of unemployment became known by the awkward name of “non-accelerating inflation rate of unemployment” or NAIRU for short. That name expressed the view that if governments held unemployment below the NAIRU then inflation would accelerate. Figure 1 shows the sort of diagram used in the textbooks to illustrate the argument that the short run Phillips curve was unsustainable and that inflation and unemployment would converge to the vertical Phillips curve that intersects the unemployment axis at the natural rate of unemployment. So attempts by policy makers to move from A in the diagram to point B with lower unemployment but higher inflation were unsustainable and the economy would move to point C with the same unemployment but higher inflation.
Figure 1: Short run v long run Phillips curve
Source: RBA
The message to policy makers was that there was no tradeoff between inflation and unemployment except in the short run.
The RBA has drawn attention to results of its in-house studies that severely undermined the notion of a vertical Phillips curve. For example, in 2017 the then Governor of the RBA, Philip Lowe, raised the possibility that the Phillips curve is getting ‘flatter’ which is very significant meaning that lower unemployment means sacrificing a smaller amount of additional inflation. Other RBA researchers had also put forward the flattening of the Phillips curve as an explanation of a slowdown in wages growth at the time as well as the puzzling evidence that wages remained stagnant despite unemployment below that rate that was considered over “full employment”. Similar results have also been found in the US. Indeed, as the Phillips curve gets flatter, we can say that there is virtually no relationship at all between unemployment and inflation. That is also suggested by Australia’s recent experience. The implied Phillips curve would be a flat horizontal line.
Now we can examine Sarah Hunter’s contribution. She said:
we think of our NAIRU models as estimating the unemployment rate at which current inflation would converge to expected inflation…when the labour market is sustainably at full employment, inflation is expected to move from its current rate towards 2.5 per cent and then remain stable…when the labour market is tighter or looser than full employment on an ongoing basis, underlying inflation will hold above or below target but not necessarily accelerate away from it.
Notice that the NAIRU is the level of unemployment consistent with “full employment”, an obvious oxymoron that must irk many people. Hence this passage says that below the NAIRU inflation may be a bit higher and below, inflation may be a bit lower but the former will not accelerate nor the latter decelerate. The accelerating bit is out of date. More importantly, the unemployment at point A on Figure 1 only makes sense if the target inflation is that applying at point A. But if point A on the Phillips curve is arbitrary then so is the unemployment at point A.
Notice also that the 2.5% inflation, the mid-point of the RBA’s target range of 2 to 3%, is entirely arbitrary yet the RBA now defines the full-employment rate as that unemployment consistent with the target. The target was always arbitrary. Then-Governor Bernie Fraser said in a speech in 1993: “Ideally, we would like (to pick up Alan Greenspan’s definition of practical price stability) to see inflation kept low enough so as not to bias behaviour in these costly ways. Putting numbers on that definition is a matter of judgment… My own view is that if the average rate of underlying inflation could be held to 2 to 3 per cent over time, we would meet our test.” He made it clear that the target at the time was aspirational, a target they thought would be hard to achieve. Stiglitz was right when he said Australia’s inflation target, 2%, had been “pulled out of thin air” and there was “no theory or evidence that 2% was the right number [for the lower bound of the 2 to 3% target]”.
This is very significant. Friedman’s natural rate of unemployment and the subsequent NAIRU was supposed to be optimal in an important sense. Below it, inflation would worsen, and employment would be more than people would have worked had they not been tricked. So it could be said that unemployment below the natural rate was a worse outcome compared to unemployment at the natural rate. But if your inflation target is arbitrary then there is nothing unique at about any level of unemployment associated with that target. Public policy cannot abrogate its responsibility to the unemployed by saying there is no alternative.
What do we make of this? Clearly Hunter and the RBA are saying that we are not in the vertical Phillips curve world and the short-term curve like Figure 1 can persist. If the tradeoff is flatter than the RBA thought then we can have unemployment a good deal lower without inflation taking off. If now it is not inevitable that we return to a particular natural rate of unemployment then we can pick anywhere on the tradeoff and there is nothing magic about the natural rate of unemployment, whether or not the RBA says so. It is illogical to choose a point on the tradeoff and then claim unemployment at that point has unique characteristics. Rejecting the neoliberalism behind the NAIRU means we do not have to see unemployment and inflation as just demand management issues. We can get into the real world of price gauging and conflict between capital and labour, we should discuss employment programs, incomes policies, competition policies, skilled migration, vocational education without debate being denied when it is couched in the natural rate language.
Furthermore, if genuine choice exists then who decides? The RBA could only play god when it convinced everyone that the natural rate of unemployment was inevitable. But if there is a choice about the combination of unemployment and inflation and the mix of policies to get there then deciding policy should not be left to a technocratic elite; the new monetary policy board that was created following the review of the RBA. These are issues that concern all Australians and so should remain in the political sphere.
Comments