A Theoretical Analysis of Unemployment and Inflation

There is a lot of debate over the direction of employment and inflation today and how policymakers should react to these developments. On one hand, the labor market seems to be in its best condition since the recession. In June, employers added 288,000 jobs, bringing the yearly total to 1.4 million. Unemployment has fallen to 6.1%, the lowest it has been in almost six years. Some predict that we could hit full employment within a year, leading to the argument that the Fed should raise interest rates sooner to prevent runaway inflation or “overheating.” Despite such promise in the labor market, growth has remained stagnant. Inflation is currently at 2.1 percent, slightly above the normal Fed target of 2 percent. Even so, this figure is much lower than expected, given the amount of stimulus and the amount of jobs that were added. Gross domestic product also fell by 2.9 percent at an annual rate, its worst since the recession. Overall growth in the first half of 2014 has been nearly zero.

            If economists expect inflation to rise as labor market conditions improve, there must be some relationship between unemployment and the rate of inflation that allows us to forecast future levels. This relationship is known as the Phillips curve, posited in 1958 by economist Alban Phillips in “The Relation between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom, 1861–1957.” In examining ninety-seven years of British data on unemployment and nominal wage growth, Phillips found that unemployment tended to be low when wages grew rapidly and high when wages grew slowly. The lower the unemployment rate the tighter the labor market, leading to firms’ raising wages to attract labor. At higher rates of unemployment, the pressure subsided. The modern-day Phillips curve relates the current rate of unemployment to future inflation, detailing a negative relationship between the two.


            In the 1960s, U.S. inflation and unemployment seemed to lie along a Phillips curve, with policymakers opting for higher levels of inflation to reduce unemployment. The close fit between the estimated curve and the data encouraged many economists, including Paul Samuelson and Robert Solow, to treat the Phillips curve as a menu of policy options. In the following decades however, the relationship between inflation and unemployment failed to hold. Edmund Phelps and Milton Friedman challenged the validity of the model. They argued that rational employers and workers would only consider real wages, which would adjust to make the supply of labor equal to the demand for labor, and the unemployment rate would then stand at a level uniquely associated with that real wage—the natural rate of unemployment. When unemployment falls below this rate, inflation tends to rise over time, and vice-versa. According to Friedman, monetary policy cannot peg interest rates or the unemployment rate for more than a short period. This limitation is a result of the timing of the policy decision and the expectations of the consumer. Let us apply Friedman’s logic to the present day, where the Fed has been maintaining low interest rates for years. It does this by buying securities, thus raising their prices and lowering their yields. This also increases banks’ available reserves and amount of credit, ultimately increasing the money supply. Although the initial increase in the money supply lowers interest rates, the process continues. The increased liquidity and lower interest rates stimulate spending, leading to rising income and eventually higher prices. Continually rising prices will lead the public into expecting further increases in prices. These latter effects place upward pressure on rates, which will eventually return to their initial levels, and possibly beyond that level in case the economy overreacts, leading to a cyclical adjustment process.

            Timing is also a factor in creating policy to control unemployment. Friedman believed in his interpretation of the Phillips curve wherein there should not be a negative relationship between inflation and unemployment, but rather between unanticipated inflation (the difference between actual and expected inflation) and cyclical unemployment (the difference between the actual and natural rates of unemployment). In the classical model, an increase in the money supply shifts aggregate demand from AD to AD. If the rate of growth of the money supply is constant for some time as well as the rate of inflation, the expected rate of inflation is the same as the rate of inflation, say 5 percent in this example. Thus, the expected price level grows by 5 percent per year. This causes short-run aggregate supply to shift up to SRAS’. In the meantime, because output remains at full-employment, unemployment remains at the natural rate and cyclical unemployment is zero. Our conclusion is zero unanticipated inflation and cyclical unemployment in this economy.


            Now assume that in the second year, money supply grows by 10 percent rather than 5 percent, which means a shift in aggregate demand to AD2 instead of AD. Also assume that the change is unanticipated. With the shift in short-run aggregate supply being the same as in the previous case, output level is now above full-employment. In addition, actual unemployment is below the natural rate and cyclical unemployment is below zero. Why is output in this case at this new level? The new rate of inflation is less than the 10 percent growth of the money supply but higher than the expected rate of inflation, which is still 5 percent. This causes the price level to increase by more than expected and producers believe that the relative prices of their goods have also increased. In the long run, the economy returns to full-employment and the inflation rate equals expected inflation. As the economy adjusts to this stage, output is still higher than full-employment output and actual unemployment is below the natural rate, which means the actual price level must exceed the expected price level. To summarize, our conclusion is exactly what Friedman and Phelps hypothesized. When consumers accurately predict aggregate demand growth and inflation, unanticipated inflation and cyclical unemployment are zero. However, an unexpected positive rate of aggregate demand growth causes positive unanticipated inflation and negative cyclical unemployment.


            You might be wondering whether the above analysis is of any use, considering unemployment and inflation data did not even follow a Phillips curve after 1970. Let us examine different factors in the 1960s, followed by the 1970s to determine why those inconsistencies exist. According to Friedman, the relationship between unemployment and inflation holds only when expected inflation and natural employment are approximately constant. In the 1960s, notice that the rate of natural employment changes slowly without much fluctuation. Although data for expected inflation in those years is not available, it is known that people were used to low and stable inflation at the time, meaning expected inflation was likely low and remained constant. Why did the level of stability change in the decades following? Two adverse supply shocks related to the price of oil caused dramatic fluctuations to expected inflation and the natural rate of unemployment. The natural rate of unemployment, which rose in the 60s, fell during this period as a result of changes in the makeup of the labor force. These fluctuations not only affected expected inflation but actual inflation, and the 70s were a time of high and volatile inflation, a problem exacerbated by the war in Vietnam, which incurred high levels of military expenditures and President Johnson’s Great Society programs. After the Fed pursued anti-inflationary policies from 1979-1982 inflation returned to low levels. This change from high expected inflation to low expected inflation suggests that the Phillips curve condition would not be satisfied.

4 5

            The instability of the Phillips curve during the period does not imply a lack of a systemic relationship between unemployment and inflation. Friedman and Phelps argued that the negative relationship between unanticipated inflation and cyclical unemployment exists and should appear in the data, regardless of changes in expected inflation and the natural employment rate.


            With this analysis in mind, what does this mean for future monetary policy? Can policymakers simply reduce unemployment by stimulating higher inflation and to what extent? Using the Friedman-Phelps model, we assume that unemployment will fall below the natural rate only when there is unanticipated inflation. Therefore, the question is whether unanticipated inflation can be generated to reduce unemployment. In the classical model used above, prices and wages adjust almost immediately to changes in information in the economy. In this model, consumers are rational and make intelligent choices regarding those changes. Because prices are flexible, the actual rate of inflation cannot be sustained at a higher level than expected inflation for more than a short time. This means that policies that stimulate inflation raise both actual and expected inflation, and do not sustain a reduction in unemployment.

            However, we just saw in our data above that there is some extent to which policymakers can affect inflation and unemployment, such as in the 1980s. There should be some time delay in which prices adjust to reflect new information, an assumption made in Keynesian economics. In this case, actual inflation can exceed expected inflation, which allows unemployment to remain below the natural rate. Let us return to our present-day situation. We expected that because of quantitative easing inflation would reach levels higher than the current 2.1 percent rate, meaning the expected rate is higher than the actual rate. As of now, actual unemployment is slightly higher than the natural rate. Policymakers are currently in a difficult situation because of poor growth, stagnant inflation levels, and a misleading unemployment rate. Our expectations-augmented Phillips curve analysis above indicates that because consumers have already expected increases in the money supply and therefore higher inflation, the Fed has been unable to keep actual inflation higher than expected inflation, thus failing to sustain a reduction in unemployment.