The Phillips Curve

The Phillips Curve

In 1958, an economist called Phillips published a paper stating that there was a trade off between the rate of change of money wages and unemployment rate. But it was not some brilliant theory that he had invented, he had simply collected data on wage inflation and unemployment for every year between 1861 and 1957, plotted a scatter diagram and drawna line of best fit.

This might sound a bit useless, but when you think of how closely related the rate of change of money wages is to the rate of inflation, this relationship became very significant.

The Phillips Curve

The original data suggested that zero inflation was compatible with an unemployment rate of around 2.5%. If the government wanted to reduce unemployment below 2.5%, they could see the cost in terms of higher inflation and vice versa. Further research suggested that this relationship seemed to hold for most developed countries in the world.

This was a godsend for the government. Although they could not control unemployment and inflation at the same time, they now had a 'guide' that told them exactly how much inflation to expect for a given reduction in unemployment. Equally, if inflation was high, they could see how much unemployment to expect for a given cut in the rate of inflation.

Why would a reduction in unemployment cause the inflation rate to rise?

If the government decided to cut unemployment through an expansionary fiscal or monetary policy, the resulting shift in the aggregate demand curve would increase output, which could only be accommodated by employing more labour. Money wages had to be raised to attract these new workers (who were voluntarily unemployed at the lower wage rate). This, in turn would cause an inflationary rise in the price level. So unemployment fell, but inflation rose.

Most of the politicians, who took this relationship as gospel, did not appreciate fully one simple point. This 'Phillips curve' was not some theory that was cast in stone, like Newton's apples falling out of trees. It was simply a relationship between two variables that had held true for the past 100 years. Everyone hoped that the relationship would be true for the next 100 years, but that did not mean that it would.

It fact, the relationship started to break down as early as the late 60s. Inflation and unemployment started rising together. What was going on? The relationship had worked for the past ten years, so there must have been some sort of short-term connection. Perhaps this relationship did not survive over a long period of time.

The expectations - augmented Phillips curve

An economist called Friedman tried to explain what was going on in the context of shifting short run Phillips curves and one vertical long run Phillips curve.

The Phillips Curve

Exp. Inf. is short for Expected Inflation.

In the diagram above, you can see three short run 'normal' looking Phillips curves and a green vertical long run Phillips curve. Let us assume that the economy begins at point A, with 0% inflation and a rate of unemployment of U*. The government decides that it wants unemployment to be lower. It increases aggregate demand by, say, building some new roads. In order to attract the required workers, the money wage is increased. Previously unemployed workers take the bait (frictionally, or structurally unemployed), unemployment falls to U1 at a cost of increasing inflation (3%). The economy moves from point A to point B along the short run Phillips curve SRPC1.

Notice that this curve has been labelled 'exp. inf. = 0%'. This means that workers expect future inflation to be 0%. They took the job with increased money wages expecting inflation to stay at 0% so that the real wage is higher. Workers only increase their supply of labour in response to an increase in real wages.

But, the increase in money wages forced the inflation rate up to 3%. The new workers did not realise this straight away. They initially suffer from money illusion. Before long the workers realise that they have, in a sense, been duped. Real wages have not increased at all. The new workers were not prepared to work at this lower real wage rate, so they withdraw their labour services and become unemployed again. The economy moves from point B to point C. Unemployment is back up to U*, but inflation stays where it is because money wages have not fallen back to their original level. The economy is now on a different short run Phillips curve (SRPC2) with expected inflation equal to 3%.

If the government tries to reduce unemployment again, the same thing will happen. The economy will move from point C to D and then to E. In other words, every time the government tries to reduce unemployment below U*, it manages to do it in the short run, but as the workers suss out the money illusion, the economy goes back to the level of unemployment U*, but at a higher, and permanent, level of inflation. Hence, the long run Phillips curve is the green vertical line LRPC. Unemployment can be higher than U* in the long run, but not below U*.

The level of unemployment U* is often called the non-accelerating inflation rate of unemployment (NAIRU). Quite simply, at U*, inflation is non-accelerating. If unemployment is below U*, then inflation accelerates.

Once the government works out that reducing unemployment below U* in the long run is futile, it has to get inflation down to where it was before. It can do this, but only by getting workers' expectations of future inflation down to where they were before.

If the economy is at point C, the only way that the government can reduce inflation back down to 0% and convince workers that this is permanent is to move down SRPC2 from point C to point F, and stay there for a while. This will be at a cost of much higher unemployment ("the price worth paying for lower inflation" - Norman Lamont, ex-Chancellor, especially after that comment!). In fact, the cost in terms of increased unemployment is higher than the benefit the government thought they would receive in terms of lower unemployment originally (i.e. U2 - U* > U* - U1). Once workers' expectations of future inflation are back down to 0%, then the government can reduce unemployment to U* without inflation rising again. The short run Phillips curve shifts from SRPC2 to SRPC1 and the economy moves from point F back to point A.

The natural rate of unemployment

The implication of the above analysis is that there will always be some unemployment in an economy. This seems to be true looking back at past data. But who are these unemployed workers?

The U* level of unemployment consists of all the frictionally and structurally unemployed. In a sense, they are all voluntarily unemployed. Frictionally unemployed workers could get a job at any time. They are simply taking their time and finding the best job. It may seem a bit harsh to say that those who are structurally unemployed do so voluntarily (remember our shipbuilder?), but most of them could actually find a job, but it might involve a life-wrenching move from the family home.

Given that there will always be some voluntary unemployment in an economy, this level of unemployment is often referred to as the natural rate of unemployment, because it occurs, in a sense, naturally. It occurs 'naturally' even when the economy is at its long run equilibrium.

The Phillips Curve

In the diagram above, DL and SL are the usual labour demand and supply curves. Another supply curve is labelled 'workforce'. Remember that the workforce includes all those who are employed and unemployed but willing and able to work and looking for work. SL represents the supply of labour at the given real wage rate. Those who are structurally or frictionally unemployed may not want to work at the given wage. At real wage W1, therefore, L1 workers are employed and L2 - L1 workers are in the workforce, but decide not to supply their labour services for the given real wage rate W1. They are voluntarily unemployed. They represent the 'naturally' unemployed elements of the workforce.

Finally, note that the two supply curves get closer together at higher wages. This is to be expected. At higher real wage rates, fewer workers will stay frictionally and structurally unemployed. They will be tempted by the higher real wage and enter employment.