FRBSF Economic Letter
2000-13; April 28, 2000
Structural Change and Monetary Policy
This Economic Letter summarizes the papers presented at
the conference "Structural Change and Monetary Policy" held in San Francisco
on March 3-4, 2000, under the joint sponsorship of the Federal Reserve
Bank of San Francisco and Stanford University's Stanford Institute for
Economic Policy Research.
Pronouncements about the "new economy" in the U.S. are made with such
frequency that they may soon become tiresome and trite. From an economist's
perspective, however, the discussion about recent changes in the structure
of the economy is just starting to get interesting, as enough data finally
are becoming available to begin a reasoned debate about what is happening
and why. Indeed, from an economist's point of view, almost all of the
heavy lifting in terms of analysis and explanation regarding the new economy
remains to be done.
The six papers presented at this conference provide some first steps in
defining the recent changes in the U.S. economy and in describing the appropriate
behavior of monetary policy in the face of such changes. The papers are
listed at the end and are available--along with comments by discussants
and the keynote speech by Federal Reserve Board Governor Laurence Meyer--at
Two papers focus on documenting recent changes in the structure of the U.S.
economy. One paper examines the apparent moderation in business cycles since
the early 1980s. The authors find a distinct decline in the volatility of
real output growth and provide some evidence to suggest that this change
reflects a behavioral adjustment on the part of durable goods producers
to keep better control of their inventories. The other paper focuses on
what is the most acclaimed attribute of the new economy: the remarkable
rise in productivity growth since 1995. The authors show that both the growing
use of information technology in businesses and the gains in the efficiency
of producing computers and semiconductors have made substantial contributions
to the recent surge in productivity growth. A separate panel discussion
by Chad Jones and John Taylor, both of Stanford University, and Mark Watson,
of Princeton University, also considers some of the recent changes in trend
Three of the conference papers explore how monetary policy should operate
during periods of structural change--particularly when the degree of this
change is unknown. Thus, the key question investigated is how monetary policymakers
should take into account uncertainty about potential output or the level
of the natural rate of unemployment. One of these papers provides an interesting
episodic analysis that is calibrated to the mid-1970s productivity slowdown
and to the mid-1990s productivity speedup. The other papers provide a general
theoretical analysis of optimal policy under data uncertainty. Governor
Meyer's keynote speech focuses on how such research can be applied to the
conduct of monetary policy on a practical level.
Finally, one paper examines the implications of structural change for
the behavior of agents in the economy. It tries to elucidate how businesses
and consumers may change their behavior in response to shifts in the policy
Output fluctuations in
the United States: what has changed since the early 1980s?
The McConnell and Perez Quiros paper analyzes quarterly movements in
real output and its broad components since the early 1950s. The paper
identifies a large and statistically significant decline in the volatility
of U.S. real GDP growth that took place in the early 1980s. Indeed, the
standard deviation of output fluctuations during the earlier period (1953-1983)
is about twice as large as during the more recent one (1984-1999).
Of particular interest is the source of this decline in volatility. It
may reflect good luck in the latter period (for example, fewer oil price
shocks and other disturbances), or improved monetary policymaking (as
suggested in Judd and Rudebusch 1998 and the conference panel discussion
by John Taylor), or a structural change in the economy (say, a shift to
a more stable service-oriented economy). Of course, a combination of these
factors also may be at work. To shed some light on this issue, the authors
disaggregate output into nondurable goods, durable goods, services, and
structures, and find that shifts in the shares of these components--and
particularly the growth in the importance of the service sector--do not
appear able to explain the decline in volatility. Instead, the authors
note that much of the decline in overall volatility can be attributed
to smoother durable goods production. Furthermore, there is evidence of
a change in the behavior of durable goods inventories but not durable
goods sales. Thus, the authors suggest that a change in the management
of durable goods inventories, perhaps including the just-in-time techniques
and tight control made possible by computers, may have played an important
role in the reduction in overall output volatility.
The resurgence of growth in the late 1990s:
is information technology the story?
Since 1995, rapid growth in real output has been accompanied by an average
annual increase in nonfarm business productivity of about 2-3/4 percent,
which is nearly double the average pace over the preceding 25 years. The
Oliner and Sichel paper adopts the standard neoclassical growth accounting
framework to determine the source of this pickup in growth. In their version
of this framework, the annual growth in output is attributed to increases
in labor, information technology capital (including computer hardware,
software, and communication equipment), other capital, and a residual
component that measures general technological change.
Their results indicate that the contribution to productivity growth from
the use of information technology capital jumped in the second
half of the 1990s, as U.S. firms invested heavily in the "high-tech" revolution.
In addition, technological advances in the production of computers
and semiconductors also appear to have made an important contribution.
Overall, the authors estimate that these two factors accounted for about
two-thirds of the recent jump in productivity growth.
Learning about a shift in trend output:
implications for monetary policy and inflation
The Lansing paper considers the consequences of a shift in trend output
for a monetary policy that is based at least in part on the difference
between actual and trend output--the "output gap" which is used in the
popular Taylor rule (Judd and Rudebusch 1998). Under such a policy, the
productivity slowdown of the early 1970s may have contributed to the substantial
rise in inflation in the latter part of that decade. This may have happened
if monetary policymakers only gradually perceived the slowdown in productivity
and trend output; thus, actual output appeared lower relative to trend
than it actually was. Consequently, monetary policy might have been inappropriately
loose, which would foster inflation.
The Lansing paper formalizes this intuition in a small forward-looking
macroeconomic model where the Federal Reserve's regression-based perceived
gap between actual and trend output is used as an input to the monetary
policy rule in real time, while the true gap influences aggregate demand
and inflation. The author calibrates two experiments to match the structural
breaks in trend output in the 1970s and the 1990s. He finds that in this
framework, errors in estimating potential output can account for some,
but by no means all, of the historical long-term movements in U.S. inflation.
Indicator variables for monetary policy
A general guideline that economic analysis gives to policymakers is the
principle of certainty equivalence, which states that optimal policy requires
the same response when there is only partial information about the state
of the economy as when there is full information (Walsh 2000). However,
under partial information, the policymaker doesn't react in the same given
fashion to the known value of, say, the output gap, but to the best
estimate of the unknown output gap. Consequently, there is a separation
between the selection of the optimal policy (the optimization problem)
and the estimation of the current state of the economy (the signal extraction
problem). Once the policymaker has obtained the best guess of the state
of the economy, he or she can then set policy as if there were no uncertainty.
In this case, more uncertainty does not lead to more cautious policy actions.
The Svensson-Woodford paper extends this result to the much more complicated
case when some of the variables that the central bank reacts to depend
on private-sector expectations of future developments in the economy.
Examples of such forward-looking variables include exchange rates, bond
rates, and inflation expectations. However, these forward-looking variables
depend on an estimate of the current state of the economy, which in turn
depends on an observation of the forward-looking variables. This circularity
in the presence of forward-looking behavior greatly complicates both the
optimization and signal extraction problems. However, this paper overcomes
these problems and shows that the certainty-equivalence principle continues
to hold in the case of a linear forward-looking model (with a standard
loss function). Thus, even in a forward-looking setting, the authors note
that the proper weight to be placed on an efficient estimate of the output
gap is unaffected by the degree of uncertainty in that measure.
On signal extraction and non-certainty
equivalence in optimal monetary policy rules
The Swanson paper explores some exceptions to the principle of certainty
equivalence. One exception noted by Smets (1998) is that certainty equivalence
fails to hold when policymakers can respond only to some of the important
determining variables in the system. For example, the coefficients of
the optimal Taylor rule--which responds only to the output gap and inflation--would
depend on the amount of uncertainty about these variables. Another exception,
explored by Rudebusch (1999, 2000) and the Lansing paper cited above,
relies on the fact that the real-time estimate of the output gap may not
be a completely efficient estimate of the actual output gap. In this case,
the optimal coefficient, say, on the output gap, also would depend on
the amount of noise in the real-time output gap estimate. The Swanson
paper generalizes this result and places it in an arguably more realistic
setting. Namely, if the output gap is taken to be one of many indicators
of a more general state of "inflationary pressures," then the weight to
be placed on the output gap is also dependent on the accuracy of its measurement.
In this case, more uncertainty calls for more timid policy actions.
Near-rationality and inflation in two
The Ball paper focuses on the very different behavior of U.S. inflation
during two periods: 1879-1914 and 1960-1997. During the early period,
when the U.S. had a gold standard before the founding of the Fed, inflation
fluctuated around a constant level throughout the sample. In contrast,
during the postwar period with discretionary monetary policy, the rate
of inflation has shown large and persistent deviations from its average.
This evidence suggests that the stochastic process generating inflation
cannot always be viewed as independent of monetary policy, as is often
assumed in economic modeling. The Ball paper proposes a near-rational
model of expectations, in which agents make optimal univariate
forecasts, in order to explain both episodes. In this model, the structural
change embodied in the evolution of the monetary system is reflected in
the behavior of the agents.
Glenn D. Rudebusch
Senior Research Officer
Ball, Laurence. "Near-Rationality
and Inflation in Two Monetary Regimes." Johns Hopkins University.
Lansing, Kevin. "Learning
about a Shift in Trend Output: Implications for Monetary Policy and Inflation."
Federal Reserve Bank of San Francisco.
McConnell, Margaret, and Gabriel Perez Quiros. "Output
Fluctuations in the United States: What Has Changed since the Early 1980s?"
Federal Reserve Bank of New York.
Oliner, Stephen, and Daniel Sichel. "The
Resurgence of Growth in the Late 1990s: Is Information Technology the
Story?" Board of Governors of the Federal Reserve.
Svensson, Lars, and Michael Woodford. "Indicator
Variables for Monetary Policy." Princeton University.
Swanson, Eric. "On Signal Extraction
and Non-Certainty Equivalence in Optimal Monetary Policy Rules." Board
of Governors of the Federal Reserve.
Judd, John P., and Glenn D. Rudebusch. 1998. "Taylor's Rule and the Fed:
1970-1997." Federal Reserve Bank of San Francisco Economic Review
3, pp. 3-16. http://www.sf.frb.org/econrsrch/econrev/98-3/3-16.pdf
Rudebusch, Glenn D. 1999. "Is the Fed Too Timid? Monetary Policy in an
Uncertain World." Federal Reserve Bank of San Francisco Working Paper
Rudebusch, Glenn D. 2000. "Assessing Nominal Income Rules for Monetary
Policy with Model and Data Uncertainty." Federal Reserve Bank of San Francisco
Working Paper 2000-03. http://www.sf.frb.org/econrsrch/workingp/2000/wpgr00-02.pdf
Smets, Frank. 1998. "Output Gap Uncertainty: Does It Matter for the Taylor
Rule?" In Monetary Policy under Uncertainty, eds. B. Hunt and
A. Orr, pp. 10-29. Wellington: Reserve Bank of New Zealand. http://www.rbnz.govt.nz/econ/monpol.htm.
Walsh, Carl. 2000. "Uncertainty and Monetary Policy." FRBSF Economic
Letter 2000-08 (March 17). http://www.sf.frb.org/econrsrch/wklyltr/2000/el2000-08.html.
Opinions expressed in this newsletter do not necessarily reflect
the views of the management of the Federal Reserve Bank of San Francisco
or of the Board of Governors of the Federal Reserve System. Editorial
comments may be addressed to the editor or to the author. Mail comments
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120