‘The inflation controversy – Demand-Pull or Cost-Push?’
Keynesian economists were divided into two camps. Some believed that inflation was caused by too much demand for goods and services, or excess demand i.e. Demand Pull Inflation. Such economists were very enthusiastic about the Phillips curve because it seemed to provide strong evidence for their views. The second camp of Keynesian economists believed that inflation was caused by cost pressures arising from high wage settlements gained by strong trade unions and from increases in import prices i.e. Cost Push Inflation. They were doubtful about the usefulness of the Phillips curve. On the
Nicholas Perry
evidence of the Phillips curve they thought that wage inflation and unemployment could be traded off against each other, that policy makers could decide to have a bit less unemployment in exchange for accepting a bit more wage inflation or vice versa. The Phillips curve helped the ‘demand pull’ Keynesians in another way as in order to manage the level of demand properly they needed a measure of the capacity of the economy. They needed to know when any further increase in demand for goods would exceed the capacity of the economy to produce those goods. The Phillips curve suggested that unemployment itself was that measure because of the correlation with inflation. Inflationary pressure was the signal that capacity had been reached.
However it is by no means inevitable that the level of unemployment should be a good indicator of capacity; unemployment is the excess supply of labour. It seems that unemployment was only a good indicator in the early 1960s because when unemployment was low, there would be lots of unfilled vacancies. More people than were available were needed to produce additional goods and services, which showed there was an excess demand in goods and services. This correlation seems to have become much weaker.
‘The short run Phillips curve and its breakdown?’
Still it is important to note that in any one year the combination of wage inflation and unemployment might be some way from the cure. What was impressive was that over a number of years the discrepancies more or less cancelled each other out – the positive errors were matched by the negative errors. The years following the publication illustrated this perfectly, between 1959 and 1966 the results showed a remarkable confirmation of the Phillips curve. The evidence was so strong that most thought the ‘demand pull’ Keynesians economists were right and had won the argument. Yet this was soon shaded out as after 1966 wage inflation was much higher than predicted by the Phillips curve.
Economists began to wonder whether the ‘cost push’ school had the correct idea as they certainly seemed to offer the most convincing explanation that the Phillips curve relationship might only be true in the short term. This was particularly true because that old measure of capacity, the unemployment percentage, showed an absence of excess demand. Unemployment was significantly higher than before.
‘The Monetarist counter-attack and rational expectations?’
As an opponent of demand management policies, Professor Friedman argued that if the government tried to reduce unemployment, in exchange for accepting higher inflation, then the reduction in unemployment would only be temporary. The increase in the rate of inflation would be permanent. Friedman holds the view that there is a ‘natural rate of unemployment’. This natural rate is the rate of unemployment at which the rate of inflation (whatever it happens to be) will be stable. The cost of permanently trying to keep unemployment below the natural rate is ever-accelerating inflation. The natural rate is therefore also called the NAIRU. The real wage rate is determined by the interaction of the demand for labour and the supply of labour. Friedman stresses that the change in real wages should be on the vertical axis of the Phillips curve graph, this is so because the level of real wages determine workers’ willingness to accept employment. Friedman suggests there is an infinite number of short run Phillips curves, each one corresponding to a different expected rate of inflation. This is the
Nicholas Perry
expectations-augmented Phillips curve. Thus if we attempted to hold unemployment below its natural rate we would arrive at the position where we can get onto worse and worse short run Phillips curves i.e. each one above and to the right of the previous one.
However as we know there is only one long run Phillips curve which is vertical at the natural rate of unemployment. This suggests that in the long run no trade off between unemployment and inflation is possible. So in this case governments ought to accept higher unemployment because when it is above the natural rate it will reduce inflation and the higher unemployment will only be temporary.
It seems that Friedman’s theory has had considerable influence. Low inflation today could be ascribed, however, to quite different factors. It could be that the ‘cost-push’ theorists are right and that unemployment reduces the inflation by means of reducing wage pressures from trade unions. Or the ‘demand-pull’ school could be correct in which case unemployment has reduced inflation by reducing aggregate demand. This suggests that when unemployment falls the rate of inflation will rise again.
Advancing further there are theories which move away from the Phillips curve entirely, even in the short run. Such a theory is called rational expectations. It was argued that if wage negotiators knew that the government had adopted appropriate anti-inflation policies it would reduce wage settlements. Employers and unions would know that inflation was going to come down because of appropriate government policy. The pain of economic adjustment would be much less and unemployment would not have to rise so much. Yet when such a theory is applied to inflation and unemployment it seems too optimistic.