focus on technology--particularly the Internet--and its implications is
most timely, because as this century draws to a close, the defining
characteristic of the wave of technological innovation sweeping over the
U.S. economy is the role of information.
The veritable avalanche of real-time
data has facilitated a marked reduction in the hours of work required per
unit of output and a broad expansion of newer products whose output has
absorbed the workforce no longer needed to sustain the previous level and
composition of production. The result during the last five years has been
a major acceleration in productivity and, as a consequence, a marked
increase in standards of living for the average American household.
Prior to this revolution in technology,
most twentieth-century business decision making had been hampered by less
abundant information. Owing to the paucity of timely knowledge of
customers' needs and of the location of inventories and materials flows
throughout complex production systems, businesses, as many of you well
remember, required substantial programmed redundancies to function
Doubling up on materials and people was
essential as backup to the inevitable misjudgments of the real-time state
of play in a company. Decisions were made from information that was hours,
days, or even weeks old. Accordingly, production planning required costly
inventory safety stocks and backup teams of people to respond to the
unanticipated and the misjudged.
Large remnants of information void, of
course, still persist, and forecasts of future events on which all
business decisions ultimately depend are still unavoidably uncertain. But
the remarkable surge in the availability of timely information in recent
years has enabled business management to remove large swaths of inventory
safety stocks and worker redundancies.
Businesses not only respond more
accurately to changes in demand, they can respond more quickly and
efficiently as well. Information access in real time--resulting, for
example, from such processes as electronic data interface between the
retail checkout counter and the factory floor, or the satellite location
of trucks--has fostered marked reductions in delivery lead times and the
related workhours required for the production of all sorts of goods, from
books to capital equipment. This, in turn, has reduced the relative size
of the overall capital structure necessary to turn out our goods and
Intermediate production and distribution
processes, so essential when information and quality control were poor,
are being reduced in scale and, in some cases, eliminated. The increasing
ubiquitousness of Internet sites is promising to significantly alter the
way large parts of our distribution system are managed.
The process of innovation goes beyond
the factory floor or distribution channels. Design times have fallen
dramatically as computer modeling has eliminated the need, for example, of
the large staff of architectural specification-drafters previously
required for building projects. Medical diagnoses are more thorough,
accurate, and far faster, with access to heretofore unavailable
information. Treatment is accordingly hastened, and hours of procedures
eliminated. In addition, the dramatic advances in biotechnology are
significantly increasing a broad range of productivity-expanding efforts
in areas from agriculture to medicine.
One result of the more-rapid pace of
innovation has been an evident acceleration of the process of
"creative destruction," which has been reflected in the shifting
of capital from failing technologies into those technologies at the
cutting edge. The process of capital reallocation across the economy has
been assisted by a significant unbundling of risks in capital markets made
possible by the development of innovative financial products.
Every innovation has suggested further
possibilities to profitably meet increasingly sophisticated consumer
demands. A significant percentage of new ventures fail. But among those
that genuinely reduce costs or enhance consumer choice, many will prosper.
The newer technologies, as I indicated
earlier, have facilitated a dramatic foreshortening of the lead times on
the delivery of capital equipment over the past decade. When lead times
for equipment are long, the equipment must have multiple capabilities to
deal with the plausible range of business needs likely to occur after
these capital goods are delivered and installed. In essence, those capital
investments must be structured in a manner sufficient to provide insurance
against uncertain future demands. As lead times have declined, a
consequence of newer technologies, less judgment about the potential
alternative economic environments in which the newly ordered equipment
will be functioning is needed. Accordingly, foreshortened future
requirements have become somewhat less clouded, and the desired amount of
lead-time insurance, in the form of what after the fact would turn out to
have been a partially unproductive addition to the capital stock, has
Indeed, these processes emphasize the
essence of information technology--the expansion of knowledge, and its
obverse, the reduction in uncertainty. The use of information in business
decisionmaking can be best described as an effort to reduce the fog
surrounding the future outcomes of current decisions.
Because the future is never entirely
predictable, risk in any business action committed to the future--that is,
virtually all business actions--can be reduced but never eliminated.
Information technologies have improved our real-time understanding of
production processes, reducing the degree of uncertainty and, hence, risk.
This, in turn, has lessened the need for a whole series of programmed
redundancies from which, in the end, little to no productive capability is
In short, information technology raises
output per hour in the total economy by reducing hours worked on
activities needed to guard productive processes against the unknown and
the unanticipated. Narrowing the uncertainties reduces the number of hours
required to maintain any given level of readiness.
But, obviously, not all technologies,
information or otherwise, affect productivity by reducing the inputs
necessary to produce the current level of existing products. Some
information made possible by technological advance more readily
contributes to developing new products that consumers value rather than to
reducing the required inputs for existing products. Indeed, in our dynamic
labor markets, the resources made redundant by better information are
drawn to newer activities and newer products, many never before
contemplated or available.
The personal computer, with its
ever-widening applications in homes and businesses, is one. So are the fax
and the ubiquitous cell phone. The newer biotech innovations are most
especially of this type, particularly the remarkable breadth of medical
and pharmacological product development. Information has armed many firms
with detailed data to fashion product specifications to most individual
customer needs. Owing to advancing information capabilities and the
resulting emergence of more accurate price signals and less costly price
discovery, many market participants are better able to detect and to
respond to finely calibrated nuances in customer demand. Value added,
accordingly, is enhanced per workhour.
The Internet offers an admixture of
potential new goods and services and potential lower costs of production.
A major part of our current GDP reflects distribution cost, and it is
evident that much of that is subject to potential competitive reduction
through Internet marketing. I do not perceive the end of the shopping
mall, if for no other reason than I have been strongly advised that
shopping is not solely an economic phenomenon. But the relationship
between businesses and consumers already is being changed by the expanding
opportunities for e-commerce. The forces unleashed by the Internet may be
even more potent within and among businesses, where uncertainties are
being reduced by improving the quantity, the reliability, and the
timeliness of information, as I am sure your sessions today and tomorrow
will have made clear.
The newer technologies obviously can
increase outputs or reduce inputs only if they are embodied in capital
investment. Capital investment here is defined in the broadest sense as
any outlay that enhances capital asset values or, for that matter, even
enhances the value of an idea.
But for capital investments to be made,
the prospective rate of return on their implementation must exceed the
cost of capital. That has clearly happened in the last five years.
In particular, technological synergies
appear to be currently engendering an ever-widening array of prospective
new capital investments that offer profitable cost displacement. In a
consolidated sense, reduced cost is reflected mainly in reduced labor cost
or, in productivity terms, fewer hours worked per unit of output.
It would be an exaggeration to imply
that whenever a cost increase emerges on the horizon, there is a capital
investment that is available to quell it. Yet the veritable explosion of
equipment and software spending that has raised the growth of the capital
stock dramatically over the past five years could hardly have occurred
without a large increase in the pool of profitable projects becoming
available to business planners. Had high prospective returns on these
projects not materialized, the current capital equipment investment
boom--there is no better word--would have petered out long ago. Indeed,
equipment and capitalized software outlays as a percentage of GDP in
current dollars are at their highest level in post-World War II history.
To be sure, there is also a virtuous
cycle at play here. A whole new set of profitable investments raises
productivity, which for a time raises profits--spurring further investment
and consumption. At the same time, faster productivity growth keeps a lid
on unit costs and prices. Firms hesitate to raise prices for fear that
their competitors will be able, with lower costs from new investments, to
wrest market share from them. Such circumstances lead to a very favorable
period of strong growth of real output and low inflation.
But the degree to which the growth rate
of productivity has been rising--indeed, whether in a long-term sense it
is rising at all--is subject to considerable debate among economists. This
results, in part, from major disputes about our national data system.
Gross product per workhour measured for
the nonfarm business sector, employing the newly revised data made
available this morning, rose an average 2-1/4 percent per year over the
past five years, and nearly 2-3/4 percent over the past two, after
averaging 1-3/4 percent over the previous two decades. Because in the past
we have had episodes of similar improvements in productivity performance
that failed to persist, these data, on their own, cannot be relied upon to
draw broad conclusions about whether an acceleration in trend productivity
is under way.
But other data are more compelling.
Growth in gross domestic income has outstripped the growth of the
conceptually equivalent gross domestic product in recent years, producing
a dramatic widening of the statistical discrepancy. Productivity growth in
the nonfarm business sector, estimated as real gross income per
hour rather than real gross product per hour, over the past two
years is, thus, a more noticeable 3-3/4 percent at an annual rate, 1
percentage point faster than measured from the product side.
Finally, because the measured level of
productivity in the noncorporate business sector exhibits noncredible
weakness for substantial spans of time, I believe data for the
nonfinancial corporate sector afford a more accurate, though
admittedly more narrow, measure of productivity performance. And here the
numbers are still more impressive, nearly 3 percent on average over the
past five years, and more than 4 percent over the past two. By this
measure, productivity growth in the 1970s and 1980s also averaged about
1-3/4 percent per year. Moreover, the acceleration in productivity appears
reasonably widespread among nonfinancial corporate firms beyond the
high-tech industries themselves, even though gains in output per hour in
the advanced technology companies have verged on the awesome.
Although it still is possible to argue
that the evident increase in productivity growth is ephemeral, I find such
arguments hard to believe, and I suspect that most in this audience would
But how long can we expect this
remarkable period of innovation to continue? Many, if not most, of you
will argue it is still in its early stages. Lou Gerstner (IBM) testified
before Congress a few months ago that we are only five years into a
thirty-year cycle of technological change. I have no reason to dispute
that, although forecasting the evolution of technology is a particularly
precarious activity. It nonetheless seems likely that we will continue to
experience vast advances in the application of the newer technologies and
their associated increases in output per workhour.
But in gauging pressures on cost growth
and prices, the critical issue is not how much of the current wave of
innovation lies ahead of us, but how rapidly the exploitation of the newer
technological synergies proceeds.
If, using Gerstner's figure, the
remaining twenty-five years of the thirty-year cycle of technological
change is exploited at a much more leisurely pace than the first five
years, the rate of productivity growth will fall. To be sure, the level
of productivity will continue to rise but at a slower pace.
A leveling out or decline in the growth
of productivity would have a profound effect on the intermediate outlook
should it occur. I say, should it occur, because evidence of a downward
bend point in productivity growth is not yet evident in our most recent
data. All the same, the rate of growth of productivity cannot continue to
increase indefinitely. At some point it must, at least, plateau. Should,
at that point, labor market tightness result in faster growth of nominal
wage rates, there would be no offset from accelerating productivity. As a
consequence, unit costs would likely rise, pressuring profit margins and
That scenario of rising cost and price
pressure is one policymakers have dealt with before, and the actions
called for, while by no means easy, are readily discernible. What modern
monetary policymaking has not faced for quite some time, if ever, has been
a major surge in innovation--matching, if not exceeding, the other great
waves this century--followed by an apparent elevation of productivity
growth. Yet even these welcomed circumstances create challenges for
Accelerating productivity poses a
significant complication for economic forecasting. For many years,
forecasters could assume a modest, but stable, trend productivity growth
rate and fairly predictable growth in the labor force. Given the resulting
growth of potential GDP, forecasting largely involved evaluating demand
growth. If it appeared to be running in excess of trends in potential, the
economy could be expected to eventually overheat, with inflation and
interest rates moving up. In the end, the economy would, at some point,
fall into recession.
With trend growth in productivity now
clearly in play, the weakness of a simple demand-side evaluation of
economic forces has been brought into sharp focus. It may no longer be the
case that an acceleration in demand presages an overheated and unstable
economy, if the demand growth is caused by growth in trend productivity.
Higher productivity growth must eventually show up as increases in
employee real incomes, in profit, or more generally both. Unless the
propensity to spend out of real incomes falls, consumption and investment
growth will rise, as indeed they must over time if demand is to keep pace
with faster supply.
But consumer demand can accelerate so
much that total demand could rise above even the productivity-augmented
overall growth of potential. This seems to have been happening in recent
years, owing to an expanding net worth of households relative to income
and perhaps a perception that the recent acceleration in real incomes will
This extra demand can be met only with
increased imports or with new domestic output produced by employing
additional workers either from drawing down the pool of those seeking
work, or from increasing net immigration.
Imports presumably can continue to
expand for awhile, since the rising rate of return on U.S. assets has
attracted private capital inflows, particularly a major acceleration of
direct foreign investment, into the United States. For the recent past,
direct foreign investment inflows have almost matched the total current
account deficit. But a continued widening of that deficit could eventually
raise financing difficulties, ultimately limiting import growth.
In addition, over the past two years,
the pool of people seeking jobs--the sum of the officially unemployed plus
those not in the labor force but wanting to work--has declined from 11.2
million to 9.6 million. The number of workers drawn into employment in
excess of the normal growth in the workforce has been running at the
equivalent of roughly a half of a percentage point of annual GDP growth.
This gap must also eventually be closed if inflationary imbalances are to
continue to be contained.
Clearly, the growth in gross domestic
product cannot exceed the sum of growth in structural productivity and in
the working-age population indefinitely. Market pressures must eventually
emerge that work to contain such unsustainable growth.
The process of containment may already
be significantly advanced. Increasing demand for financing capital goods
relative to domestic savings, a reflection of the previously cited
imbalances, has apparently been exerting marked upward pressure on real
long-term market interest rates, especially as economies abroad
The measurement of real yields, that is,
nominal interest rates less expectations of inflation over the maturity of
a debt instrument, is inevitably imprecise. It depends, of course, on
estimates of inflation expectations, which are very difficult to
accurately pin down. But judging by yields on U.S. Treasury
inflation-indexed securities, the real riskless interest rate has risen
about half a percentage point for ten-year maturities since late 1997.
Private long-term real rates have apparently risen even more. The spreads
of corporates against Treasuries have widened significantly for
investment-grade and, especially, high-yield debt over this period. As a
consequence of these higher real interest rates, the ratio of net worth to
income for the average household is already lower than it was earlier this
We do not have enough experience with
technology-driven gains in productivity growth to have a useful sense of
the time frame in which market pressures contain demand. Moreover, it is
not clear as yet how much cumulative impact the rise in real long-term
interest rates over the past two years will have on future demand.
Going forward, the Federal Reserve must
monitor not only this response, but also the evolving capacity of our
economy to meet higher levels of demand. Maintaining balance between these
forces will be essential to preserving the stable price environment that
has provided a firm foundation for this period of extraordinary innovation
and progress in the U.S. economy.