The Phillips Curve and Commitment: Econ302, Fall 2004 Professor Lutz Hendricks
The Phillips Curve and Commitment: Econ302, Fall 2004 Professor Lutz Hendricks
Phillips Curves
We know that loose money raises output and employment. Should the Fed exploit this relationship? Why not drive the unemployment rate to zero by expanding the money supply? Do we see this relationship in the data?
1.1
A Phillips curve depicts the relationship between ination and unemployment. A. W. Phillips discovered in the 1950s that these are negatively related. In more recent data, this relationship has disappeared why?
1.1.1
Another way of looking at Phillips curves: Is there a positive relationship between ination and GDP relative to trend? Because GDP and unemployment are closely related, this captures Phillips s idea. We nd a Phillips curve in early data, but not in recent ones.
1960-69
1980-89
1.2
A one-time increase in the money supply results in: higher Y , N ! lower unemployment. higher P . Why does unemployment fall? Ination reduces the real wage. Firms hire more labor. But in the long-run prices are exible. The real eects of monetary policy disappear. All that is left is a higher P . Therefore, it takes repeated increases in M to keep unemployment low. The model therefore predicts a relationship between ination and unemployment, such as
Y Y T = (1=a) i
(1)
What happens if the Fed repeatedly raises M to inate away real wages?
1.2.1
Expectations matter
If the Fed expands M in every period, agents will update their ination expectations. Wage demands will reect this. Unions negotiate a wage sequence to target a particular real wage. Higher expected ination leads to higher nominal wage demands. This is an example of the general point: for money to have real eects, it must be a surprise. A stylized representation of the Phillips curve is then
Y Y T = (1=a) (i ie)
(2)
or
i ie = a Y YT
(3)
If ination is above the expected level (ie) then output rises above trend Y T .
1.2.2
Assumptions: 1. There is a Phillips curve, such as the one derived from the sticky price model. 2. The Fed cares about unemployment and ination. Represent Fed preferences by indierence curves.
Optimal Fed policy exploits the Phillips curve to attain the highest indierence curve.
Is point B sustainable?
1.2.3
With constant ination, the public eventually adjusts expectations (ie ") and workers demand higher nominal wages. The Phillips curve shifts up. Real eects vanish. The Fed must respond by raising the ination rate to stay ahead of ie.
10
1.2.4
NAIRU
This model suggests: there is only one unemployment rate that is consistent with constant ination. That rate is called NAIRU: Non-Accelerating Ination Rate of Unemployment. Any attempt to push unemployment below NAIRU leads to rising ination.
11
1.2.5
Evidence on NAIRU
Unanticipated ination is measured here as the change in ination, but one can be more sophisticated. NAIRU is gradually shifting over time. Historically, NAIRU was near 7%. In the Golden 90s is declined to about 5.5%.
12
1.2.6
What temporary factors have reduced NAIRU? 1. The investment boom has boosted labor demand. Additional capacity has kept ination low. 2. The demographic composition of the labor force has shifted towards lower unemployment (older and more skilled workers) 3. Low oil and other import prices kept ination low 4. Faster productivity growth tends to reduce NAIRU ("wage aspirations" story) Some part of the decline in NAIRU is likely permanent: 1. More e cient job matching 2. Shift towards temporary and ex time labor Globalization does not change NAIRU Trade/GDP has risen since 1950, but NAIRU does not trend down.
13
1.2.7
This makes it hard to know when to tighten or loosen monetary policy. Perhaps this is why the Fed moves very gradually.