Everyone is aware that US economic growth in the last few years has
been robust. Since 1995, real GDP has been growing at close to 5% per
year. By contrast, from 1972 to 1995, it grew at only 2.75% per year.
Just a few years ago, almost no one would have believed that we could
return to pre-1972 growth rates; indeed, many believed that the
"speed limit" for the U.S. economy was probably something
like 2.5% per year, with any higher growth fueling inflation. However,
looking back from the perspective of the first year of the new
millennium, the "doomsayers" appear to have completely
missed the birth of the "New Economy". Not only did economic
growth accelerate, but growth of labor productivity boomed. Many argue
that this productivity increase is more than the "normal"
improvement that can be attributed to a business cycle expansion.
Rather, it is believed that a fundamental change has taken place that
will keep productivity growth high. If true, this means that the slow
growth post-1973 period, as well as the NAIRU inflation-unemployment
trade-off may have been banished.
Chairman Greenspan was an early convert to the view that the rapid
pace of technological innovation since the mid 1990s was fueling
productivity growth that allowed for rapid economic growth without
inflation. Indeed, this recognition is credited for convincing the Fed
to postpone monetary tightening-allowing the economy to continue to
grow more robustly than many had thought possible. Where some had
shown pessimism, enthusiasts for the New Economy view display nearly
unbounded optimism. Nowhere is this more in evidence than in President
Clinton's budget surplus projections-that rely heavily on a sustained
expansion, and that seem to grow by the trillions upon every revision.
By nearly all accounts, the cornucopia is here to stay.
I hate to throw cold water on such Pollyanna thinking. In reality, the
Goldilocks economy is not all that unusual. Rather, we seem to have
simply returned to a more normal growth rate after a quarter century
of below normal growth. The constraints that kept growth low during
most of the final quarter of the 20th century were the same
constraints that have periodically limited growth in the US, and that
have limited growth in other OECD countries-namely, demand
constraints. That is, during the Clinton expansion US aggregate
demand has been remarkably high, allowing firms to utilize their labor
forces and equipment at a high rate, raising measured productivity. In
contrast, demand was generally so low between 1975 and 1995 that
resources could not be fully utilized; hence, measured productivity
growth was low. On this view, the features of the New Economy, as well
as other supply-side factors, have next to nothing to do with rapid
productivity growth in the past five years.
Let me offer two different kinds of evidence in support of this
hypothesis. The first is supplied by economist Robert Gordon, who uses
a typical production function approach to estimate multi-factor
productivity growth for the period 1995-99. (See "Does the ‘New
Economy' measure up to the great inventions of the past?", Robert
Gordon, NBER Working Paper 7833.) He finds that there has been an
acceleration of productivity growth in the durable manufacturing
sector, however, this is limited to the production of computer
hardware and other closely related durables. Surprisingly, there has
been no increase in multi-factor productivity in the 88 percent of the
rest of the private economy. That is, all the productivity growth he
measures is attributed only to production of computers and none
to the computer-using part of the economy. This finding should
be quite disconcerting to those who believe that the tremendous
diffusion of the computer throughout our economy has generated "New
Economy" productivity-enhancing effects that are largely responsible
for the Clinton expansion.
There is another, less direct, method of looking at this issue. One
can apportion per capita GDP growth between growth of the employment
rate (number employed divided by population) and growth of output per
worker. In a sense, a nation can "choose" between employing a greater
percent of its population (a "high employment" path), or getting more
output per worker (a "high labor productivity" path). If we
compare US experience with that of Western Europe or Japan, one finds
a quite remarkable difference: while other nations took a high
productivity path, the US chose a "middle road". Since 1970,
the US has increased its per capita GDP by about 50%; Western Europe
has increased its per capita output by about 55%, and Japan has
increased its by over 100%. However, while the source of US
growth of per capita output is just about evenly apportioned between
increases to its employment rate (which grew by about 25%) and
improvements to labor productivity (which grew by a bit more than
25%), virtually all the per capita growth in Western Europe and Japan
has been due to growth of labor productivity (Japan's employment rate
grew by about 5%, while Western Europe's was essentially flat).
Economists have long argued that the US has much "freer" labor
markets, which encourage employment and allow us to enjoy lower
unemployment rates. This appears to be true. But these freer labor
markets do not give us any advantages with respect to economic growth!
Rather, there appears to be a trade-off, and maybe a choice: we can
choose higher employment rates or greater productivity
growth. This choice is taken within a context of growth of aggregate
demand. Given our "freer" labor markets and high growth of employment
rates, the US must grow much faster than Western Europe or Japan does
in order to raise our labor productivity. The Clinton expansion has
merely allowed aggregate demand to grow sufficiently fast to lower
unemployment and increase labor productivity. Thus, this expansion is
unusual only when compared with the 1973-95 period, when demand
constraints were tighter (so that our output growth barely kept pace
with employment growth-resulting in low measured productivity growth).
When compared with the "Keynesian" 1960s, the years surrounding WWII,
or the "frontier" years after the Civil War-all periods during
which aggregate demand remained high-there is nothing so unusual about
the current expansion. In the past few years, our growth was finally
high enough to raise productivity growth.
Thus, we can reject the supply-side arguments that really underlie the
New Economy view, as well as other arguments that focus on supposed US
advantages deriving from greater reliance on free markets. Neither
slow growth between 1973-93, nor rapid growth post-1995 should be
attributed to slow or fast productivity growth. Indeed, productivity
growth should be seen as an uninteresting residual that results from
the interplay between employment growth and growth of per capita
output (when output grows faster than employment rates, measured
productivity increases)-both of which are primarily demand-driven. The
real constraints in the coming years in the US will be demand
constraints-as they usually are. Thus, while I would argue that the US
Goldilocks expansion is not sustainable, this has nothing to do with
the tenuous nature of the New Economy, but rather with the inevitable
downturn of spending that will come when households slow their record
pace of borrowing. 
|
