Building in Inflation

When Keynes wrote his General Theory, inflation was not an issue facing policy makers searching for ways out of the Great Depression.  In fact the problem during the 1930s was more likely to be falling prices (deflation) than rising prices (inflation). It is not surprising, therefore, that in the Keynesian model we looked at in the 1903s there was no explicit treatment of inflation.

By the mid 1960s, though, inflation was an issue. After falling through the latter half of the 1950s, the inflation rate began its climb in the early 1960s, and by the end of the decade the rate had reached 7 percent - a rate which would cause prices to double in ten years. There was no avoiding the fact that inflation was real, and it needed to be addressed by policy makers.  For this to happen, however, the prevailing economic theory / model needed to be modified to include an analysis of inflation.  This was one of the challenges facing Keynesians as they attempted to take their anti-depression model and transform it into an anti-recession model.

wpe6.gif (2717 bytes)

The key concept in the extension of Keynesian analysis to accommodate inflation was the Phillips curve - a concept named after a professor at the London School of Economics who wrote a very influential article in 1958 entitled "The Relation between Unemployment and the Rate of Change of Money Wages in the United Kingdom, 1861-1957."  Based on nearly 100 years of data, Phillips found an inverse relationship between the rate of unemployment and the rate of increase in wages - the higher the rate of unemployment the lower the rate of wage growth. The logic was straightforward.  As unemployment rates decreased, the balance of power between workers and employers shifted to the workers who would use this new found power to raise their wages.  Similarly, when business was bad, when the economy was in a recession, the balance of power shifted to the employers and there would be little, if any, power to raise wages. The results of his work looked similar to what we see in Diagram 1.  

Diagram 1

Phillips1.gif (2154 bytes)

In the translation from England to the US the relationship was modified slightly - wage growth was replaced by growth in the price level - but the logic remained the same.  In this case, an increase in wages associated with low unemployment rates would translate into higher prices and greater inflation. 

The nature of the relationship between the price level and unemployment can be seen by returning to the AS-AD diagram. On the left-side we have a diagram you should now be familiar with - an AS-AD diagram where an increase in aggregate demand is depicted with the outward shift in the AD curve. This increase in demand moves the economy to the northeast, from point A to point B.  This move would produce a higher price level and higher output, which would translate into more jobs and lower unemployment. If the demand curve shifts further to the right, we would find an even higher price level (higher inflation rate) and higher output level (lower unemployment rate).  If we plotted the relationship between the inflation rate and the unemployment rate, we would have the curve on the right - the Phillips curve.

Diagram 2
AS-AD Diagram Phillips Curve

Phillips_ASAD.gif (3918 bytes)Phillips_ASAD1.gif (3062 bytes)

The evidence from the 1960s in the US seemed to confirm what Phillips had found in England. There was a very definite, although not perfect, negative relationship between the inflation rate and unemployment rate.  This curve became viewed as a constraint on policy makers - just another example of the "no-free lunch" concept.  If there was a desire among policy makers to reduce the unemployment rate, it would come at a cost of higher inflation. This was a time when you could find academics telling their students that Republicans tended to favor point A while Democrats tended to favor point B in Diagram 2.

It was also a time where discussions focused on the definition of full employment and on policies to shift the Phillips curve.  In Diagram 3, let's assume that the economy is currently experiencing unemployment and inflation rates of 7 and 4 percent [point b]. Using traditional macroeconomic policies the economy could be moved towards point a where we have been able to "buy" a lower unemployment rate (4 percent) with higher inflation (6 percent).

As you know from your earlier discussion of the production possibility curves, there is another option for policy makers concerned with reducing unemployment.  A second option would be to alter the tradeoff, to move the Phillips curve inward so the economy could attain a 5 percent unemployment rate with inflation of 4 percent.

Diagram 3
A Shift in the Phillips Curve

Phil_shift.gif (2569 bytes)

To understand the discussions of policies needed to shift the curve we need to look a bit closer at unemployment. A good place to start is by recognizing that all unemployed are not equal - a point raised by Throop in his testimony against the Kennedy tax cut.  A very popular classification scheme decomposed all unemployment into three components - frictional, cyclical, and structural unemployment.

We can now return to the original question: How do we shift the curve inward? To see how this might happen, consider the situation in the mid 1980s when thousands of Americans were moving to Texas following the dreams of wealth and opportunity sold in the popular television show "Dallas." Unfortunately, these migrants found little opportunity as Texas suffered through a severe recession.  The nightly news was filled with stories of broken dreams, unemployment, and families living in their cars. The question facing policy makers was: how could we lower unemployment without increasing aggregate demand which would bring with it higher inflation?

One possibility would have been better information for the decision makers. Under this view the unemployment in Texas was the result of people making poorly informed decisions. If they had only known the situation, they would have never left for Texas: if they had better labor market information, they may not have given up their jobs, or they may have known California was actually experiencing more rapid growth than Texas at that time. A national clearinghouse for employment opportunities would certainly improve the tradeoff as people would be less likely to make the wrong choices that resulted in unemployment.

A second possibility would be to raise the costs / lower the benefits of unemployment. Returning to our Texas example, if local officials put these people up in the local Holiday Inn as they searched for work, they would be likely to engage in a lengthy search and remain unemployed for a longer time. Unemployment rates would be reduced if people spent less time unemployed, which they would do with lower unemployment benefits. The increase in unemployment rates attributable to the widespread availability of benefits is an example of where there can be substantial unintended side effects associated with public policies.

An example of where the unintended side-effects of a policy could alter the 'equilibrium' level of unemployment would be the proposal in 1998 to extend Medicare payments to those over 55. Because of unintended side effects, the cost of the Medicare expansion is likely to be far larger than the simple addition of the costs of those currently in need of the coverage. Costs would be substantially higher since this proposal could prompt some individuals to retire early and some employers to drop their health coverage because it would be picked up by the government. In both instances this would increase the number of individuals on Medicare.

For the structurally unemployed there is a third possible approach to reducing structural unemployment - worker retraining. Moving the structurally unemployed workers back into the labor market can be viewed as the equivalent of trying to fit square pegs into round holes. Firms are looking for workers with a certain set of characteristics - good computer skills for example - and the structurally unemployed possess a very different array of skills. If we provided these individuals the opportunity to retrain, then their stay on unemployment would be shorter and the unemployment rate lower. When confronted with a limited budget, however, the policy makers face a difficult choice.  Do they use the limited funds to help the really needy, those that are beyond rehabilitation, or should they use the funds to help those living on the margin pass above the margin.  Unfortunately, there is no easy answer and this problem will always plague retraining policies.

We have one last question: what should be the target rate of unemployment? One concept that came to be associated with this was full employment - which the Employment Act of 1946 set as the target for government policies. But what was full employment? It certainly did not mean a zero percent unemployment rate. As we saw earlier, some unemployment was a good thing, a necessary feature of a dynamic, vibrant economy that was creating new jobs and destroying older jobs.  By the early 1960s there was a general acceptance of the fact that 2 percent was unrealistic and in the early 1960s John Kennedy set 4 percent as the level of full employment. In the 1980s the full employment rate of unemployment became a hotly debated issue, but more about that in the 1980s unit. 

The 1960s taught us that inflation could not be ignored, and it prompted Keynesian economists to extend their analysis. What the extensions made clear was that lower unemployment came at a cost, but it kept intact the faith in the ability to fine-tune the economy. While we might now have a problem determining an appropriate target rate of unemployment, there should be no problem utilizing traditional monetary and fiscal policies to hit the target. There may, however, be problems with timing.