Inflation refers to a sustained or continuous increase in the general (average) level of prices  within the economy, and its two causes are demand pull and cost push. As such, the Phillips Curve model can be used to distinguish the differences and interrelationship between demand pull and cost push causes of inflation.
The Phillips curve is named after the New Zealand born economist A.W. Phillips  . In 1958 he observed an empirical relationship between wage unemployment and wage inflation using UK data for 1861-1957  . In essence, the Phillips curve model portrays an inverse relationship between inflation and the level of aggregate activity (or demand). This level of aggregate demand has come to be represented by the unemployment rate. The logic behind such an assumption is that the greater the rate of growth of aggregate demand, the greater the resulting inflation and growth of real GDP  – and hence the lower the unemployment rate will be. While on the flip side, the slower aggregate demand grows, the smaller the resulting inflation and the slower the growth of real GDP  – and hence the higher the unemployment rate is expected to be. Hence we are faced with an economic dilemma – i.e. there is an apparent ‘trade off’  between high inflation and low unemployment, and between low inflation and high unemployment.
Firstly, one needs to understand the concept of demand pull as a cause of inflation. Demand pull inflation is the effect of excess demand pressures  in product and labour markets within the economy. This inflation is purely associated with shifts in aggregate demand (i.e. changes in aggregate demand expenditure), as the economy attempts to spend beyond its capacity to produce  .
In figure 1.1 we can use the model to analyse the effect of excess demand pressures on wage inflation. Within the model the rate of unemployment (u) can be used to represent the concept of excess demand pressures. The downward sloping curve represents the ‘trade off’ between wage inflation and unemployment in the short run due to changes in aggregate demand expenditures.
The unemployment rate Un is a situation characterised by no excess demand and in which the economy is at “full employment” output. At this level of unemployment there is only frictional or structural unemployment  – cyclical unemployment is therefore non existent. Initially at Un inflation is 0. Most sectors generally across the economy are already at their employment capacity. Now, if aggregate demand expenditure within the economy were to increase then unemployment would fall to U from Un. Aggregate demand expenditure in the short run exceeds the economy’s aggregate short term output. The more demand expenditure increases and the more the unemployment rate is pushed to U1 and below, then the greater the number of sectors in the economy that reach their capacity. With so many sectors at capacity, the greater the excess demand pressures, the greater the increase in wages  . Whenever unemployment (u) is above Un then excess supply of labour would put downward pressure on wages. The point Un thus represents the economy’s natural rate of unemployment.
We can thus convert this excess demand (Un-U) and wage inflation relationship to a relationship between excess demand and inflation in the general level of prices. We assume that labour costs are the only cause for change in prices, that there is no productivity growth and that there are no other changes in other input costs  . Hence the rate of price inflation can be set to the rate of wage inflation i.e. P*=w* – and our new Phillips curve model for price inflation is now represented by figure 1.2 below.
However, demand pull inflation only looks at the effects of excess demand or the aggregate demand (AD) expenditure effects on inflation – i.e. when AD> AS in the short run. We must also consider cost-push inflation which is inflation associated purely with shifts in aggregate supply  (AS). Cost push factors include increases in supply side costs that are not caused by excess demands pressures – such as wages (.e.g. market power of unions  ), other input costs or indirect taxes and importantly inflationary expectations.
Figure 1.3 represents cost push inflation. Within the model we have assumed some presence of initial cost push inflation at Un – our natural rate of unemployment (represented by the grey shading). This initial cost push inflation is equivalent to p*=3%. At point Un there is no demand pull inflation only cost push which is generated from the aggregate supply of the economy  . For example, this initial cost push inflation of 3% could be attributed to an increase in structural unemployment which increases Un. Any increases in autonomous factors e.g. monopoly power of unions or business  , indirect taxes or further structural unemployment shifts the Phillips curve upwards as represented by the arrows in the above model.
Most importantly, inflationary expectations or inflationary psychology is a major determinant of cost-push inflation.
In figure 1.4  we have included Inflationary expectations in our Phillips Curve (PC) model. The downward sloping Phillips curves represent the ‘trade off’ that occurs between unemployment and inflation due to changes in aggregate demand in the short run.
Inflationary expectations or psychology  18is important to understand in the overall context of inflation. If price levels change, workers would demand an increase in their nominal wages to compensate for the loss of real purchasing power  , and employers would similarly lift their product prices by the same rate to protect their profits. In this way inflation can actually become a self-fulfilling event.
In the figure above at the initial unemployment rate of Un (the natural rate of unemployment), workers are accustomed to 3% inflation (point A1) and expect that inflation rate to prevail into the future.
Now suppose the government believed the level of unemployment was too high or was unaware of the natural rate of unemployment being 6% – and sought to reduce unemployment with expansionary fiscal and monetary policy. Expansionary fiscal and monetary policy (i.e. lower taxes and/or increased government expenditure and/or cutting interest rates) would stimulate increased aggregate demand expenditure. As such lower credit costs and taxes would induce business investment and consumers would have greater disposable incomes and lower credit costs. Individual, business and government spending would increase from point A1 such that short run AD would exceed short run AS. Excess demand pressures pull up product prices leading to higher price inflation. Higher product prices increase business revenues and profits and firms then respond by hiring additional workers so that they can increase output at these higher product prices. Unemployment subsequently falls from Un to U1 (i.e. from 6% to 4%)
Now the economy’s transition to point B1 (4% unemployment and 6% inflation) was due to demand pull inflation created by excess demand pressures. This movement to point B1 is represented by the blue arrow.
Once workers realise that inflation is now 6% at B1 instead of 3% as it was previously at A1, they will demand and receive nominal wage increases to restore the purchasing power they had lost. Businesses will not necessarily be reluctant to pass these wage increases on, as they know they can simply increase their product prices by the same factor. As businesses pass on these wage demands, business profits will fall back to their A1 levels. This profit reduction means that the original motivation of businesses to increase output and employ more people disappears  . Once business profits are eroded or disappear, businesses cut back on labour to recoup some of their profit losses. Unemployment subsequently results, and the PC shifts upward to PC (2). Unemployment has returned back to 6%, yet inflation is still 6% – as people had adjusted their expectations of inflation to 6%. P*e=6. The red arrow represents this process of cost push inflation and the new subsequent fulfilling expectations of inflation.
If the government tries to move unemployment back down to U1=4%, then the same process happens once more. At B2, once workers realise inflation is actually 9% rather than 6% they will demand higher nominal wages to compensate for the loss of real purchasing power  . Businesses will pass on the wage increases, but profits will be reduced, as such the motivation to increase output and hire more staff disappears. Businesses cut back on their labour costs and unemployment results once more and the economy moves back to the full employment point Un=6%. Nevertheless expectations of inflation have adjusted once more – workers and employers build expectations of 9% inflation into their psyche (point A3, Phillips curve 3).
Unemployment, as represented in figure 1.4 by our downward sloping Phillips curves – can deviate only in the short term from the “full employment” output point. In the long run the economy returns to u=Un ((i.e. the non accelerating inflation rate of unemployment (NAIRU  )). In the long run the Phillips curve is vertical as any stable rate of inflation is consistent with the natural rate of unemployment.  Hence there is no long run trade off between inflation and unemployment  . In the long run expansionary policies to generate lower unemployment will only lead to accelerating inflation  .
In order to get inflation down, governments need to adopt contractionary policies (i.e. increase taxation, spending cuts and/or increased interest rates). That is they will need to create such contractionary measures that unemployment exceeds the natural rate of unemployment. Policy needs to attempt to break inflationary expectations and inflationary psychology – and that is often why inflation targeting is a good method of reducing expectations of inflation within an economy.
To conclude, demand pull as a cause of inflation is solely concerned with the excess demand (aggregate demand) that occurs when u