View original document

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

A Policymaker Confronts Uncertainty
St. Louis Gateway Chapter of the National Association for Business Economics
St. Louis, Missouri
September 16, 1998
Published in the Federal Reserve Bank of St. Louis Review, September/October 1998, 80(5), pp. 3-8


hen, in August, I chose a title for
these remarks I thought my
approach would be academic in
tone. Little did I guess that the
topic would be immediate and obvious because
of the press of events, both abroad and at home.
Uncertainties abound in today’s environment. I
must say, though, that the uncertainties we face
have been in the background all along. What is
different at the moment, and worth remembering,
is that the uncertainties are simply very obvious
right now.
I have long been interested in the analysis of
monetary policy under uncertainty. The problems
arise from what we do not know; we must deal
with the uncertainty from the base of what we
do know. The general public seems to have a
vastly unrealistic view of how much the Fed
knows, and sometimes assumes that the problem
is simply that the issues are so complicated that
only the experts can understand them. In fact,
the basic issues of uncertainty can be explained
quite readily.
I will, shortly, talk more precisely about the
current policy situation facing the Fed, but before
getting into current issues, I do want to indulge
my academic proclivities by providing an outline
of the sources of uncertainty we face.

I divide the sources of uncertainty facing the
monetary policymaker into five categories: 1) the
data; 2) future events, shocks and disturbances;
3) how the economy works; 4) market reactions

to Fed policy; and 5) market anticipations of Fed
policy. Let me talk briefly about each of these and
then turn to a policy framework for dealing with

Economics is an empirical science, and the
lifeblood of every empirical science is the data
necessary to confirm theoretical propositions and
to establish the regularities upon which scientific
inference depends. U.S. data are pretty good, but
we are constantly finding gaps and questionable
areas in our base of empirical knowledge. Day in
and day out we struggle to interpret the flow of
current information and must always be alert to
various possible anomalies and biases. Let me
provide just one example that has been of interest
in recent quarters.
Measuring wage inflation is of tremendous
importance in trying to understand whether
unexpectedly robust growth in nominal demand
is feeding through not only to lower unemployment but also to higher than desired increases in
wages. For tracking wages, I rely heavily on the
Employment Cost Index (ECI), for that index is
constructed to provide the best possible measure
of increases in labor compensation. A key feature
of the index is that it tracks compensation changes
occupation by occupation, position by position,
and is meant to be free of distortions that arise
from changes in the industry composition of
employment and the percentage of overtime in
total employment. The question that analysts
must deal with now is whether wage inflation
may be taking the form of promotions to higher


grades in firms’ employment ladders. That type
of wage inflation would not show up in the ECI,
which tracks changes in compensation grade by

Future Events
The course of the economy will obviously
depend on things that happen in the future.
Because of the lags in the effects of policy changes,
I want the Fed to act ahead of future events that
would require policy adjustments. Yet, my crystal
ball is little better than anyone else’s. The private
market did not foresee the severity of developments in Asia, nor did I. Obviously, every forecaster does the very best he can at forecasting
the future, but, without question, the difference
between what actually happens and the forecast
can be very large indeed.

How the Economy Works
The interplay of developments in economic
theory and empirical observation has enormously
improved our understanding of the macro economy over the last 50 years. Yet, there are extremely
important areas subject to wide dispute. A key
area today, and much in the news, is the magnitude of the wealth effect on household spending.
The recent decline in the stock market may
depress consumption spending substantially,
or maybe it won’t. In recent years, consumption
spending has been high relative to the current
flow of disposable income. It is certainly consistent
with the hypothesis of a substantial wealth effect
to observe this low saving rate out of current
income. Households have been relying on the
appreciation in the stock market to increase
wealth, and have, therefore, felt comfortable in
spending a large fraction of current disposable
income. Under this argument, the decline in the
stock market should indeed be expected to reduce
consumption. However, the effects may come
about with a long lag, and there are other possible
hypotheses to explain the high level of consumption in recent years. The fact of the matter is that
I really am not confident about the size of the
wealth effect.

Market Reactions to Fed Policy
The economy’s response to a Fed policy
adjustment will depend importantly on whether
the market believes the policy adjustment is
temporary, permanent, or the first of a series of
policy adjustments in the same direction. I want
to separate this discussion from the immediate
policy situation, so let’s think back to a year ago
(before I arrived at the St. Louis Fed). At that time,
the economy was growing robustly, and there
were possible signs of overheating. Suppose that
last fall the Fed had raised the federal funds rate
by 25 basis points. The economy’s response would
presumably have been minimal if the market
anticipated that the Fed was going to bring the
funds rate back to 51/2 percent in a matter of a
few months. Or, the economy’s response would
be somewhat greater if the market believed that
the 25-basis-point increase was likely to remain
in effect for, say, a year. The effect on the economy would have been even larger if the market
viewed the funds rate as the first of a series of
Everyone, both inside the Fed and out, is
well aware of this argument. But at the time the
Fed is considering a policy change, I have little
insight into how the market will actually react to
a policy change of 25 basis points in the funds rate.
To some extent, the Fed can shape this reaction
by what it says, but I am not myself confident
that I can predict the reaction.
Moreover, uncertainty about how the market
will react is by no means the end of the matter. If
a Fed policy adjustment takes the market largely
by surprise, as was the case with the increase in
the funds rate in February 1994, the market reaction may be one of wholesale revision of expectations about the state of the economy. If the market
believes the inflation environment is benign, for
example, and the Fed tightens policy, the market
may then suddenly develop a case of inflation
jitters simply because it believes that the Fed has
a case of inflation jitters. (Of course, the Fed never
has “jitters,” only “concerns.”) Still, this problem
of making the market’s judgments coincide with
the Fed’s judgments is an extremely difficult part

A Policymaker Confronts Uncertainty

of the policy problem the Fed faces. One of the
biggest risks we face is a gap between the market’s
expectation of Fed policy and what the Fed actually does.

Market Anticipations of Fed Policy
Closely related to the problem of uncertainty
over market reactions to Fed policy is the uncertainty of market anticipations of Fed policy. The
difficulty here is that the Fed should not adjust
policy simply because the market expects it. That
said, mistaken decisions in private markets may
arise if Federal Reserve policy does not match
the anticipated policy.

So, what do we do? At the outset, we need to
be clear about several things. First of all, at any
given time the Fed receives an ample amount of
advice from every direction and, therefore, after
the fact someone or other will always have had
things pegged better than the Fed did. But that is
not the right criterion for judging the wisdom of
policy decisions. We must make the judgments
on the basis of ex ante considerations— you pay
your money and make your bets before the game
is played. The issue is always whether the Fed is
making a mistake that is predictable at the time
the decision is being made. With all of the uncertainties in the environment, it is rarely the case
that the Fed make obvious ex ante mistakes. Or
at least so it seems to me in recent years.
Second, we should always keep in mind that
the Federal Reserve has essentially only one policy
instrument. I like to think of that instrument as
the amount of money the Fed creates, but for those
who prefer to concentrate on the federal funds rate,
the issue is exactly the same. However you put
it, the Fed controls just one policy instrument—
money growth or the funds rate.
With one policy instrument, the Fed can
achieve at best one policy goal. My conviction,
which I believe is the overwhelming view in the
economics profession, is that the goal must be
the rate of inflation. If the Federal Reserve is suc-

cessful in keeping the rate of inflation low and
stable if, roughly speaking, the economy enjoys
sustained price stability, then the Fed will have
done its job and done it well. Price stability aids
the efficiency of the economy and from recent
evidence it also appears to be helpful in holding
down the average level of unemployment. But
with one policy instrument, the Fed cannot aim
directly at economic growth, the unemployment
rate, the level of the stock and bond markets, the
exchange rate, the growth of the Japanese economy, and so forth and so on. To use a simple
analogy, when driving a car we need a steering
wheel to control direction, an accelerator and
brake to control speed, and a transmission to
control forward or reverse. If the only control
instrument available were the steering wheel, at
best we could control direction. There would be
no use in also trying to control speed.
This point, which is so obvious to professional
economists, is extremely important in the current
environment. Many are calling on the Fed to
reduce interest rates to calm the market. Some
argue that cutting rates in the United States will
promote a more healthy financial and growth
environment abroad, will ease the pressing financial problems in Russia, and so forth and so on.
The fact of the matter is that with one policy
instrument the Federal Reserve cannot do all of
these things. An attempt to pursue other objectives
that in fact damaged the pursuit of price stability
in the United States would only add to world
financial problems. The most important contribution the Federal Reserve can make is to maintain
price stability at home, and the economic stability
that accompanies achievement of that goal. Do
not misinterpret me; improving growth prospects
in Japan and elsewhere around the world is an
important objective. It is just that, in my judgment,
the Federal Reserve does not have policy instruments to further this objective. Of course, if financial turmoil in the world threatens to depress the
U.S. economy, then adjustment of U.S. monetary
policy is appropriate, and I have no doubt that
the Fed will make that adjustment when the case
for it is clear.


A second important policy principle for dealing with uncertainty is that we need to minimize
errors in the market’s expectation of what the
Federal Reserve is going to do. These errors can
never altogether be eliminated, because what the
Federal Reserve will do depends in good measure
on events that no one can forecast. But what the
Fed can do is to reiterate its conviction that maintaining price stability is the primary, indeed
almost exclusive in my view, goal of monetary
policy. I believe that the market in recent years
has in fact come to interpret the Fed’s actions and
analyses within that context. The Federal Reserve
should also share its views with the markets and
explain the reasons for its policy decisions. This is
not by any means an easy job, because individual
FOMC members may differ on the rationale for a
given policy change. That is, it is typically the
case that FOMC members can agree on what to do
without being able to agree with great precision
on exactly why the action is appropriate. Different
economists, quite naturally, have different views
on what the most compelling circumstances are
in any given situation.
Another important principle is that we should
rely on built-in market stabilizers to the maximum
possible extent. Let me provide a very brief background on this important topic, which is by no
means fully explored or understood at the level
of the professional journal literature. Chicagoschool economists have long argued that a steady
rate of money growth would set a firm basis for
the market to form expectations about the longrun course of monetary policy. Given confidence
in price stability, markets could then respond to
current information and current economic conditions by adjusting interest rates to equilibrate
saving and investment at full employment. This
vision of monetary policy did not ultimately
prevail, most likely because of the disturbances
caused by unpredictable changes in money
demand in the short run. Still, the principle of
relying on market mechanisms to equilibrate markets to the maximum possible extent still holds.
Under the Fed’s operating procedure of fixing
the federal funds rate in a narrow range in the

short run, it appears that a built-in stabilizing
mechanism, a mechanism not anticipated in the
policy debates of the 1960s and 1970s, does operate. The market moves longer-term interest rates
by substantial amounts given the base of the federal funds rate and given market confidence in
continued low inflation. The federal funds rate has
been constant at 51/2 percent from March of last
year to the time of this writing. Over this period,
long-term interest rates fluctuated very substantially. The way the process seems to be working
is that the market has increasingly become convinced of the Federal Reserve’s objective of maintaining price stability and has been adjusting
longer-term interest rates on the basis of those
firm low-inflation expectations. As market rates
fluctuate, the incentives for businesses and households to spend on capital goods and consumer
durables rise and fall. Of course, this process can
work only if the Federal Reserve does in fact adjust
the funds rate in due time, as required by emerging economic conditions. And that, I believe, is
exactly what the Federal Reserve has done to the
best of its ability.
If I am correct with this analysis, then the
market should have confidence that the Federal
Reserve will adjust the federal funds rate up or
down as required by emerging developments to
keep the economy at low inflation. There does
seem to be a substantial amount of room to time
these adjustments so as to further the objective
of stability in growth and employment. But the
key to this process is that the market must believe
that the Fed’s long-term goal is (after allowing for
the biases in broad price indexes) true price stability—neither inflation nor deflation.
In setting interest rates, the market should
always concentrate on the fundamentals of the
situation, rather than on trying to guess what the
Federal Reserve is going to do next. The Fed is
concentrating on the fundamentals; if the market
is correct in judging that the fundamentals will
require an interest rate adjustment, then as the data
arrive and future prospects change the Federal
Reserve will indeed change interest rates in the
anticipated direction. However, under current
circumstances, the Federal Reserve does have the

A Policymaker Confronts Uncertainty

luxury of waiting to be sure that the fundamentals
do indeed point in a particular direction. That
luxury is an extremely important gain from the
Fed’s investment in policy credibility. Because
the market trusts the Fed to keep inflation low and
stable, we do not have large changes in inflation
expectations complicating our policy decisions.

Applying these comments to today’s circumstances, it is obvious to everyone who watches the
markets that the level of interest rates has declined
in anticipation of a Fed easing of the federal funds
rate sometime within the next few months or few
quarters. I can be quite vague about the timing
because the timing doesn’t really matter very
much to a holder of a 30-year bond. If the fundamentals do come in as the market anticipates, and
let me reiterate the importance of that conditional
word “if,” then of course the Fed will act. However, the possible future events are by no means
one-sided. If the situation were perfectly clear, the
Federal Reserve would already have adjusted rates.
I am, I hope, only saying very obvious things,
but it is extremely important that we be clear
about these obvious things to avoid misinterpretation of the Fed’s policy and what I am saying
about it. If it turns out that the economy remains
much more buoyant than the market seems to
anticipate at this time, and especially if it turns
out that we see inflationary pressures rising, then
the next adjustment in interest rates could be up,
rather than down. Most of the adjustment would
be in market rates, as new information changes
market perceptions about the likely course of
events. In time, the Fed might want to move the
funds rate up under these circumstances, but
analysts should not underestimate the importance
of the built-in stabilizing effects from market
adjustments of interest rates.
My discussion so far has been about what I call
the first great principle of monetary policy—the
pursuit day in and day out of price stability in an
environment inherently riddled with uncertainties.

Let me finish with a brief word about the second great principle of monetary policy: the need
for the Federal Reserve to act under extraordinary
circumstances to provide extra liquidity to the
marketplace. Every now and then, a crisis erupts
in which the market mechanism itself breaks
down. These are rare occurrences, as evidenced
by the list of events in recent years that have
required extraordinary liquidity provision. (The
most famous recent case is the stock market crash
of 1987, during which the market mechanism for
trading equities was under extraordinary stress
as evidenced by trading volume outrunning the
capacity of the marketplace, the panic conditions,
and the concerns over the viability of several securities firms.) In this circumstance, the Fed acted
vigorously to provide extra funds to the market
until conditions quieted and the panic subsided.
Another circumstance was the near failure of
Continental Illinois Bank in 1984, during which
the market mechanism for large-bank CDs was
severely strained and spreads widened as investors
developed great uncertainties about the entire CD
market. The Fed intervened to smooth over that
market upset. Another case was the Penn Central
Railroad bankruptcy in 1970. In this case, the
market for commercial paper was disrupted.
These infrequent events require prompt and
vigorous Federal Reserve actions to provide extra
liquidity and reassurance to the market. The
essential feature of these periods is not only that
market prices are moving, but also that the market
mechanism itself is breaking down in some way.
Remember once again that the Federal Reserve
has only one policy instrument. We cannot simultaneously achieve our goal of price stability and
provide support to each individual market in the
economy. The Fed’s day-in and day-out policy
must be concerned with price stability. As a nation
we rely on many different mechanisms for keeping the economy humming along smoothly and
efficiently. We rely on private markets with decisions made by millions of individual participants
to solve problems in the distribution of goods
and in the determination of securities prices. We
rely on the mechanisms of government spending
and taxes and regulation to solve many other


allocation and distribution problems in our economy. The Federal Reserve’s role is a key one, but
it is just part of the society’s mechanism for producing good outcomes for the long run.

I’ll finish with a very brief summary of my
analysis. The Fed faces many uncertainties, and
must adjust its one policy instrument to navigate
as best it can this sea of uncertainty. The fundamental principle is that the Fed should use that
one policy instrument to achieve long-run price
stability. The Fed, by making clear to everyone its
commitment to that goal, is able to rely increasingly on market adjustments of interest rates to
equilibrate the economy. The Fed needs to adjust
the funds rate target from time to time, but not
day in and day out. The second great principle
of monetary policy is that the Fed should stand
ready to provide extra liquidity to stabilize markets
in those rare circumstances that panic is breaking
down the market mechanism itself. Because U.S.
markets work so well, extraordinary policy intervention is fortunately extraordinarily rare. My
bottom line is that market participants should
concentrate on the fundamentals, and so should
the Fed. With success in maintaining price stability, the Fed will need to adjust its policy instrument from time to time, but if the bond traders
can get it right they’ll do much of the stabilization
work. We at the Fed can then sit back and enjoy