View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

Hawks, Doves, Bubbles, and Inflation Targets
James Bullard
President and CEO
Federal Reserve Bank of St. Louis

George S. Eccles Distinguished Lecture
Jon M. Huntsman School of Business
Utah State University
16 April 2012
Any opinions expressed here are mine and do not necessarily reflect those of others on the Federal Open Market Committeee.

C ONCLUSION

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

Inflation targeting

C ONCLUSION

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

I NFLATION TARGETING AND THE DUAL MANDATE

The FOMC adopted an explicit, numerical inflation target at the
January 2012 meeting.
Some discussion has suggested that “inflation targeting” is
inconsistent with the Fed’s dual mandate.
The purpose of this talk is to argue that inflation targeting is
perfectly consistent with the Fed’s dual mandate.
Indeed, as we shall see, inflation targeting is consistent with
hawks, doves, and even bubbles.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

I NFLATION TARGETING

At the January 2012 meeting, the Federal Open Market
Committee (FOMC) named an explicit, numerical inflation target
of 2 percent.
The Fed joins many central banks around the world in adopting
an inflation target.
It has been Chairman Bernanke’s goal since joining the Fed.
Congratulations to the Chairman on this important
accomplishment.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

T HE DUAL MANDATE

The dual mandate is actually a triple mandate:
The Fed should conduct monetary policy to “... promote effectively
the goals of maximum employment, stable prices, and moderate
long-run interest rates.”

Most focus on two goals: “maximum employment” and “stable
prices.”

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

T HE F ED AND THE ECB

The ECB, by contrast, has a single mandate.
The goal is to promote “stable prices.”
In practice, monetary policy is viewed in the same way in
Europe as it is in the U.S., despite the differing mandates.
This talk may shed some light on why this occurs.

C ONCLUSION

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

D O CENTRAL BANKS BEHAVE DIFFERENTLY ?

C ONCLUSION

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

T HIS TALK

Use a simple “toy” model.
Use a simple policy rule.
The model plus the policy rule jointly determines an equilibrium.
The coefficients in the policy rule will affect the nature of the
equilibrium.
Key point: The choice of the inflation target is separate from the
choice of the coefficients in the policy rule.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

Consistency with the dual mandate

C ONCLUSION

I NFLATION TARGETING

A

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

TOY MODEL

Households maximize an index of material well-being by
making decisions on how much to consume and save, and how
much time to devote to market work.
The solution to their problem gives one equation.

Firms hire workers and produce output for sale to households.
The solution to the firms’ pricing problem gives a second equation.

The monetary authority controls a short-term nominal interest
rate.
We will use a Taylor-type policy rule to describe the decision on
interest rates.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

T HE CONTROVERSIAL STICKY PRICE ASSUMPTION
Models of this general form have been popularized by Michael
Woodford (2003, Interest and Prices, Princeton University Press).
A key assumption is that prices are “sticky” in a certain sense.
If prices are perfectly flexible, the model is like the “real business
cycle” model described by Edward Prescott (1986, FRB
Minneapolis Quarterly Review).
There would be no role for the central bank in that model.
The sticky price assumption is controversial: See, for instance,
Mark Bils and Peter Klenow (2004, Journal of Political Economy).
However, I will not challenge the sticky price assumption here.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

S OME EQUATIONS
The three equation system is
zt
πt
rt

= Et zt+1 θ (rt Et π t+1 ) + et
= κzt + βEt π t+1
= ϕπ π t + ϕz zt

(1)
(2)
(3)

Here zt is the output gap, π t is the deviation of inflation from
target, and rt is the deviation of the nominal interest rate from its
long-run value.
So, the steady state occurs where zt , π t , and rt are all equal to
zero.
The et term is a stochastic shock that keeps knocking the
economy away from steady state.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

S OME EQUATIONS
The three equation system is
zt
πt
rt

= Et zt+1 θ (rt Et π t+1 ) + et
= κzt + βEt π t+1
= ϕπ π t + ϕz zt

(1)
(2)
(3)

The model is forward-looking because the actors in the model
are forward-looking.
The parameters θ, κ, and β come from the structure of the model.
In calibrated versions, κ 0.024 and β
is mostly expected inflation.

0.99; thus inflation today

The “Taylor-type” policy rule has parameters ϕπ and ϕz .

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

W HAT IS THE OUTPUT GAP ?
The output gap in the model is not the output gap of common
parlance.
Instead, zt is the difference between the amount of output that
would be produced if prices were flexible versus the amount of
output actually produced when prices are sticky.
The flexible price level of output would fluctuate dramatically in
response to shocks in the economy.
The size of the output gap measured this way is likely to be
smaller than conventional measures once prices have a chance to
adjust.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

W HERE IS THE INFLATION TARGET ?

The inflation target is not explicitly specified in this system.
This is because it is inside the π t term, which is the deviation of
observed inflation from an inflation target.
The central bank controls the inflation rate in the medium and
long term.
The inflation target is simply the embodiment of this fact.
In effect, researchers “pencil in” the long-run rate of inflation
they think the central bank desires.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

N AMING AN INFLATION TARGET
Historically, central banks did not say explicitly what rate of
inflation they were trying to achieve in the medium to long run.
After the global inflation debacle of the 1970s, this practice was
called into question.
Since the central bank controls the inflation rate, there seems to
be little to be gained from “hiding” the inflation target.
Financial markets will “pencil in” their own perception of the
inflation target anyway, but with some uncertainty about the
true value.
That just adds unnecessary uncertainty to the macroeconomic
system.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

I S AN INFLATION TARGET HAWKISH OR DOVISH ?
Naming an explicit numerical inflation target is neither hawkish
nor dovish.
It is simply a recognition that the central bank controls the
medium- to long-run rate of inflation, and that in order to
minimize uncertainty the central bank may as well say what it is
trying to achieve.
The subject of which actual value of long-run inflation is best for
society is the subject of an entire literature.
The literature generally supports low rates of inflation.
As a practical matter, many central banks have adopted 2
percent.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

T HE DUAL MANDATE AND THE STEADY STATE
In the steady state, the central bank achieves its inflation target.
The other equations, representing the private sector, then churn
out the steady state values of the other variables.
Chief among these are the steady state level of consumption and
the steady state level of labor supply.
The steady state level of labor supply could be interpreted as the
“maximum employment” of the dual mandate.
It is the amount of time households desire to work given wages
and all other variables in the economy.

The shock et keeps knocking the economy off of this level of
employment (either above or below).

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

T HE BEST POLICY ACHIEVES THE DUAL MANDATE

In the three equation system above, the central bank can move
the nominal interest rate to offset incoming shocks exactly.
The other variables never leave their steady state values.
Inflation remains at the target rate of inflation.
Employment remains at the maximum level.
The dual mandate is achieved exactly at every point in time.
Real-world policy cannot fully offset incoming shocks, but this
toy model provides a conceptual benchmark.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

W HAT ABOUT A SINGLE MANDATE ?
In a single price stability mandate system, the essential story
would not change.
The central bank still controls inflation over the medium to long
run.
It still makes sense to explicitly name the inflation target as
opposed to “hiding” it.
And, achieving the single mandate is still consistent with the
maximum level of employment of households.
The single mandate may be a clearer way to describe the
essential story, but it would not change the story.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

Hawks and doves

C ONCLUSION

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

T HE RHETORIC ON THE DUAL MANDATE

If inflation targeting is consistent with the dual mandate, why all
the discussion?
Answer: There are more aspects to policy than just the inflation
target.
In particular, there are the policy parameters ϕπ and ϕz in the
Taylor-type policy rule.
These parameters describe how aggressively the central bank
reacts to inflation ( ϕπ ) and to the output gap ( ϕz ) when setting
the nominal interest rate in the Taylor-type rule.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

M ORE RHETORIC

Relatively large values of ϕπ might be viewed as “hawkish”,
while relatively large values of ϕz might be viewed as “dovish.”
But whatever values are chosen for these parameters, the system
operates within the context of an inflation target.
In other words, the nature of the policy rule is separate from the
issue of naming an inflation target.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

T HE TWO PARTS OF THE TAYLOR - TYPE RULE

I think most of the discussion about the dual mandate is really a
discussion about how much emphasis should be put on each of
the two parts of the Taylor-type policy rule.
If shocks can be offset completely each period, then the
policymaker should choose values to accomplish that.
But that can only be done in the toy model.
In reality, both inflation and output are going to deviate from
their steady state values.
So, how should these values be set?

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

W HAT DO THE HOUSEHOLDS WANT ?
One advantage of starting the analysis with households is that
policy can be chosen to maximize the well-being of the
households.
This is superior to allowing policymakers to impose their own
judgements on the macroeconomy.
The idea is to choose ϕπ and ϕz to maximize household utility
when the system is simulated over a long period of time.
There is a large literature on this topic which provides ample
fodder for both hawks and doves.
In general, the answer will depend on additional assumptions
made in the underlying model

C ONCLUSION

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

Beyond interest rate adjustment

C ONCLUSION

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

O THER CONSIDERATIONS

There are many important monetary policy considerations other
than the nature of interest rate adjustment.
Some of these can be discussed even with just the toy model.
Others involve missing elements from the model.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

B UBBLES
There has been a lot of discussion concerning the possibility that
current Fed policy may lead to “bubbles” in the economy.
I interpret bubbles to mean that there are multiple equilibria, that
is, situations where two or more sets of prices and expectations
can clear markets.
In that situation, the model cannot tell us which equilibrium will
be achieved.
Some equilibria can be very volatile.
Interestingly, the toy model has a clear condition for such a
situation to exist.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

T HE TAYLOR PRINCIPLE

The condition for multiple equilibria in the toy model is that the
policymaker violates the “Taylor principle.”
The Taylor principle is, in its simplest form, that nominal interest
rates should be adjusted more than one-for-one with deviations
of inflation from target.
The principle is violated when ϕπ and ϕz are “too small.”
In effect, the policymaker must be sufficiently aggressive in
responding to shocks, otherwise the economy will have multiple
equilibria, some of which may be very unpleasant.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

T HE WORST POLICY
In the toy model, one of the worst policies is to set both ϕπ and
ϕz to zero.
This guarantees that the Taylor principle is violated, and that
multiple equilibria exist.
This is also known as the “interest rate peg” policy, because
interest rates never change.
Actual policy rates in the U.S. have been near zero since
December 2008 and are projected to remain there until late 2014.
This could be viewed as an approximation to the “interest rate
peg” policy, and thus conducive to multiple equilibria.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

U NEMPLOYMENT
The toy model is missing important variables.
One of these is unemployment.
Existing models of unemployment in macroeconomics have a
search-theoretic nature.
Leading authors: Peter Diamond, Dale Mortensen, Chrisopher
Pissarides.
It is difficult, but possible, to merge the search-theoretic models
with macroeconomics.
See, for instance, Mark Gertler, Luca Sala, and Antonella Trigari,
Journal of Money, Credit, and Banking.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

U NCONVENTIONAL POLICY

Another issue is that most monetary policy is not currently about
interest rate adjustment.
Instead, so-called “unconventional” policy such as quantitative
easing has come to the fore.
For an assessment of the effectiveness of quantitative easing
programs, see the Federal Reserve Bank of St. Louis conference,
“Quantitative Easing,” June 30, 2011.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

M ISSING FACTORS
Other factors have been especially important during the last five
years, in ways that were unanticipated before 2007.
One critically important area is housing and real estate more
generally.
Another critically important area is financial market stability.
Runs on non-bank financial institutions were very important
during the recent crisis.

Without these elements in the models in a well-understood way,
we cannot be completely sure that any particular monetary
policy is the appropriate one.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

G LOBALIZATION

I would be remiss if I did not mention globalization and
international monetary policy arrangements as another missing
aspect of this model.
For some discussion of this issue, see Bullard and Singh (2008,
Journal of Monetary Economics).

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

R OBUST POLICY

These considerations suggest that policy should be conducted in
a way that is robust to possible errors of omission or comission
in the baseline model.
There is a “robust control” literature that adopts techniques from
engineering to the forward-looking systems of macroeconomics.
The findings there have been interesting but are not sufficient at
this time to address the many issues in macroeconomic models.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

Conclusion

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION

C ONCLUSION
The FOMC recently set an explicit numerical inflation target.
I have argued that such a target is consistent with the Fed’s dual
mandate.
Much of the discussion about the dual mandate is, in my view,
really about the nature of the Fed’s reaction function to economic
events.
However, that issue is separate from setting an inflation target.
I have stressed that heavy focus on the nature of the Fed’s
interest rate reaction function in the current environment is
questionable.
There are many issues at least as important, and resolution of
any of those issues could change the argument for a particular
reaction function.

I NFLATION TARGETING

C ONSISTENCY WITH THE DUAL MANDATE

H AWKS AND DOVES

Federal Reserve Bank of St. Louis
stlouisfed.org
Federal Reserve Economic Data (FRED)
research.stlouisfed.org/fred2/
James Bullard
research.stlouisfed.org/econ/bullard/

B EYOND INTEREST RATE ADJUSTMENT

C ONCLUSION