Full text of Economic Commentary (Federal Reserve Bank of Cleveland) : Monetary Policy Rules and Stability : Inflation Targeting versus Price-Level Targeting, February 15, 2002

View original document

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

February 15, 2002*

Federal Reserve Bank of Cleveland

Monetary Policy Rules and Stability: Inflation
Targeting versus Price-Level Targeting
by Charles T. Carlstrom and Timothy S. Fuerst

ike Ulysses tied to the mast, most of
us have committed ourselves to some
extraordinary measure to overcome our
weakness in the face of temptation.
We may put our alarm clock across the
room to ensure that we get up promptly.
Even though we know we need to get
up at 6:00 a.m., when 6:00 rolls around
we are tempted to sleep just a “little”
more. Central banks can also gain from
commitment, which in their case means
adopting a rule that monetary policy
actions must follow. One of the earliest
and most famous proposals was Milton
Friedman’s constant-money-growth
rule. He argued that the monetary
authority should let the money supply
grow at a constant rate and ignore shortrun considerations, since attempts at
stabilizing inflation or output would
ultimately make matters worse.
With the unexpected shift in money
demand in the early 1990s, Friedman’s
proposal took a beating (for a good
description of how the rule became
ineffective after the technological innovations of the 80s and 90s, see
“Canada’s Money Targeting Experiment,” a 1998 Economic Commentary
by Paul Gomme). The relationship
between money and prices is no longer
as predictable as once thought. Most
policymakers now recognize that a
constant-growth rule will not prevent
inflation from varying substantially
over short and long time periods.
Central banks around the world have
recently turned to another type of rule,
namely, inflation targeting.

ISSN 0428-1276
*Printed April 2002

A consensus is growing around the
world that central banks should adopt
policies that achieve low and stable rates
of inflation. Inflation targeting appeals to
policymakers as a way of directly
accomplishing those goals. Inflation targeting also has the advantage of being an
easy rule to verify. People can tell if the
central bank is hitting the mark. But is
inflation targeting a good rule or might it
eventually break down like the money
growth rule? Unfortunately, some recent
research indicates that inflation-targeting
rules may have a weakness that makes
them unreliable tools for achieving stable inflation.
A closely related rule is a price-level target, in which the central bank promises
to keep the price level within a prespecified band. Although similar, the two
rules have a fundamental difference.
Inflation targeting is forward-looking, as
its goal is to stabilize the growth rate in
prices. In contrast, price-level targeting
builds in a backward-looking element
because the target is the level of prices.
This Economic Commentary first outlines the advantages of adopting a monetary policy rule. We then examine the
benefits of having a stable inflation rate
to help understand why several central
banks have a rule that promises to stabilize inflation. While there are some clear
benefits to adopting an inflation target,
we conclude that inflation targeting may
be destabilizing. We suggest that pricelevel targeting can avoid this potential
pitfall and still allow many of the benefits normally associated with inflation
targeting. While a common criticism of

Monetary policy rules help central
banks exercise the discipline necessary to achieve their long-term goals.
The type of rule many banks are
turning to these days is inflation
targeting, which has several advantages. But in following the rule, banks
usually base their actions on forecasts
of future inflation, and this can lead
to inflation-rate instability in some
cases. A price-level target offers many
of the same benefits as an inflation
target, but because it uses past inflation to guide the bank’s actions, it
avoids this vulnerability.

price-level targeting is its focus on the
past, we suggest that this feature may
actually be a virtue.

■

Rules versus Discretion:
Time Consistent Policy

Economists have long argued that the
best and surest way for a central bank to
do its job is to adopt a rule and stick to it.
One advantage of a rule is that it makes a
central bank’s actions more transparent.
Because a rule forces the bank to explicitly specify its long- and short-term
objectives, the public is less apt to misinterpret the bank’s actions, making inflation stability easier to achieve. For
example, a central bank may make a
temporary change in policy. If the bank’s
actions aren’t transparent, the public
may interpret this as a change in the
long-term objective of the bank. Because
the reasons for the bank’s actions are not
explicit, the public’s expectations about
future inflation may have no moorings.

But another, more sinister problem with
discretion is that a central bank’s shortterm objectives may be inconsistent with
its long-term objectives, preventing a
bank from realizing its long-run goals. A
rule prevents a central bank from undermining its long-range goals for shortterm results. This advantage of rules
arises because of what economists often
call the dynamic inconsistency problem.
An example can help clarify this problem. A central bank may want to keep
inflation low. But a little extra inflation,
if it’s not anticipated, can bring down
unemployment. So the central bank, like
a dieter longing for one more cookie, is
sometimes tempted to fool the public by
increasing inflation to reduce unemployment. While having one cookie or
two cookies a day is not necessarily
bad, ten can quickly become disastrous.
Similarly, if the central bank consistently tries to lower the unemployment
rate, the public would understand this
tendency. The extra inflation would be
expected, and unemployment would no
longer fall. The end result would simply
be higher inflation.
Rules are meant to avoid this tendency.
But for a rule to succeed it must be
transparent and easily verifiable so that
the public can monitor whether the
central bank is fulfilling its promise.
One such commitment mechanism is a
low-inflation target. We examine some
of the benefits associated with inflation
targeting to help understand why this is
becoming the rule of choice among
many central banks.

■

Benefits of Inflation
Targeting

Canada, England, Sweden, and New
Zealand have all adopted explicit inflation targets. Operationally, this means
that they have an objective to keep inflation within some prespecified band over
a period of usually one to three years.
Inflation targeting appeals to so many
central banks because variable inflation
is thought by many to be costly. One
cost of variable inflation is that if it is
unanticipated, it can lead to excessive
output variability. Another problem with
variable inflation is the misallocation of
resources caused by sticky prices. With
sticky prices, an increase in the money
supply will lead to an increase in the
demand for the goods of firms that don’t
raise their price relative to those that do.

The end result is that while real output
increases, resources (labor and capital)
are misallocated between firms that
adjust their prices and firms that cannot.
Eventually, these resources flow back to
where they are most valued, but adjustments are costly. Once again, inflation
variability may lead to excessive variability in output.
The presumed benefit of inflation
targeting is clear. If prices are predictable, firms can preset their prices
without risk. Prices will be the same
tomorrow whether or not firms can
adjust their prices. Inflation targeting
thus eliminates the inefficiencies associated with sticky prices.

■

Inflation Targeting: The
Problem with Looking Ahead

To achieve an inflation target, central
banks base their policy changes on
inflation projections. Without such
forward-looking behavior, the monetary
authority would repeatedly respond to
past inflation shocks, many of which are
temporary, with no bearing on future
inflation. The Bank of Canada sums it
up this way: “There are lags of a year to
18 months or more between monetary
policy changes and their effects on inflation and the economy. A chain of events
is set in motion that affects consumer
spending, sales, production, employment, and other economic indicators.
This means that monetary policy must
always be forward-looking.”
While the benefit of looking ahead is
clear, this approach may not provide an
anchor for inflation expectations. An
earlier Economic Commentary of ours
explains some recent research that indicates an inflation target can potentially
leave money growth vulnerable to what
economists call sunspot events—extraneous and unpredictable events that set
into motion self-fulfilling cycles of
inflation expectations and realized inflation (see “Sunspots and Monetary
Policy” in the recommended readings).
The term “sunspots” is a misnomer since
the events are not likely to be purely
extraneous as the name suggests, but
instead depend on some fundamental
economic variable.
In general, sunspots can arise because
the money supply is adjusted passively
by the monetary authority. It is supplied
at whatever level is necessary to achieve
the target. That is, the money supply is

endogenous. Changes in the money
supply can be self-fulfilling because
policy decisions depend on what the
public is expected to do, and the public,
in turn, bases its behavior on monetary
policy actions. This can lead to a wellknown problem of “infinite regress,” in
which the public’s behavior and monetary policy affect each other in turn, and
there is nothing objective on which to
“pin down” either. Outcomes are determined by each side’s beliefs about what
the other side will do.
Sunspots are endemic to a forwardlooking rule because current movements
in the money supply depend on
expected inflation, and expected inflation depends on current movements
in the money supply. Thus, an increase
in expected inflation sets in motion a
future increase in the money supply that
is ultimately inflationary.
Because sunspots are caused by the
endogeneity of the money supply, a
constant money growth rule as
advocated by Milton Friedman would
eliminate the possibility of sunspot
fluctuation. Yet, as discussed in the
introduction, this rule may leave
inflation unnecessarily volatile.
But inflation targeting may do the same.
Inflation targeting presents central
bankers with a paradox: to prevent
prices from rising over the short term,
they must react to changes in expected
inflation. But such action may lead to
self-fulfilling cycles of expected inflation, which can actually increase inflation variability.
Recent research has shown that another
way to avoid the possibility of sunspots
is for the monetary authority to change
interest rates aggressively in response to
inflation, and to base the bulk of the
response on past inflation. The problem
with a proactive agenda is that money
growth is responding to marketdetermined variables. A backwardlooking interest-rate rule, however,
commits the central bank to moving
future funds rates in response to today’s
price movements. This timing difference
mitigates the coordination problem
because the monetary authority does not
“move” until long after the public has
moved. Instead of responding to what
the public is expected to do, the monetary authority is responding to what the
public has already done.

■

Backward-Looking Rules:
Are they Time Consistent?

Does this guideline provide us with a
solution to the sunspot problem? On the
surface it appears to be a reasonable
candidate for a good monetary policy
rule. While it would result in more inflation variability than is optimal, inflation
expectations would be pinned down.
But an inflation target that is primarily
backward-looking has a major drawback. It doesn’t overcome the problem
that rules are meant to correct. A central
bank may still be tempted to go for
short-term results, which undermines
its long-term objective. The incentive to
cheat is always there because responding to yesterday’s inflation allows for
more short-term variability
of inflation than is desired, and looking
forward “just once” comes at no cost.
A central bank always has an incentive
to stabilize inflation changes (by looking ahead) because this provides good
results today. Occasionally reacting to
only expected inflation (as in doing so
today) is not a problem and indeed is
beneficial. But here is the rub: repeated
occasional movements quickly become
systematic, and monetary policy that
systematically looks forward is prone
to sunspots.
Because of this, it is difficult to see how
any inflation-targeting rule could be
operationalized to guarantee that monetary policy is based primarily on past
inflation. A central bank is always going
to have an incentive to cheat, while at
the same time a promise to look backward and not forward in setting monetary policy is not easily verifiable.

■

Price-level Targeting:
Looking Back to a Better
Future

The problem associated with forwardlooking interest-rate rules is quite general. To solve it, we need a rule that is
easily verifiable and naturally builds
into it a backward-looking element. A
price-level target would fill the bill.
Although the two are sometimes used
interchangeably, there is one crucial difference between an inflation target and a
price-level target. With an inflation target, past inflation misses do not affect
future policy actions. That is, there is
base drift. With a price-level target,
however, past misses must affect future
policy actions because the monetary
authority has to get the price level back

on track. To bring prices back down,
the central bank must respond to
increases in yesterday’s inflation by
increasing the funds rate today. A true
multiyear price-level target does not
have base drift.
An example will illustrate the point.
Suppose that last year’s inflation rate is
2 percent above normal. In the case of
inflation targeting, this fact is relevant
only if it helps in predicting the coming
inflation rate. In the case of price-level
targeting, this past inflation must be balanced by a central bank policy of lowering the future inflation rate to keep the
price level within the target range.
Price-level targeting necessarily builds
in a backward element that is crucial to
avoid sunspots and effectively pin down
short-term-inflation expectations and
real output. The backward element built
in by a price-level target minimizes
other problems usually associated
with inflation targets. An Economic
Commentary by Gavin and Stockman
pointed out that the base-drift problem
with inflation targeting leads to a great
deal of uncertainty about what the price
level 5, 10, or 30 years in the future will
be. The central bank may miss its inflation target by a very small percentage
in some years, but if these misses are
not offset, they will accumulate and
may become quite large after 30 years.
Therefore, a price-level target that
offsets these misses will reduce the
uncertainty associated with buying
and selling long-term fixed bonds.

■

Conclusion

This Economic Commentary has
explained why inflation targeting is
attractive to many central banks. Yet
despite the attractiveness, there is a
potential problem that central banks may
eventually run into if they adopt the rule.
While inflation-targeting rules based on
forecasts may work well some of the
time, perhaps even most of the time,
they may not always work well. Because
the consequences could be severe if the
rule fails, alternatives to inflation targeting are worth considering.
We suggest that a price-level target
may be better than an inflation target
because of its inherent backwardlooking feature. This may allow the
monetary authority to enjoy many of
the benefits of inflation targeting without exposing inflation to potentially
destabilizing shocks.

■

Full text of Economic Commentary (Federal Reserve Bank of Cleveland) : Monetary Policy Rules and Stability : Inflation Targeting versus Price-Level Targeting, February 15, 2002

FRASER