Go to Brad De Long's Home
What Does the Economics of Growth Have to Say About Public Policy?
J. Bradford De Long
University of California at Berkeley
"Growth economics" began in the 1950s--created in large part by
M.I.T.'s Robert Solow. It was reborn, with somewhat different aims and emphases,
in the 1980s.
The 1950s growth economics tradition led economists to take up and hold
on to two sets of positions. The first set was empirical: the bulk
of economic growth in living standards and material prosperity appeared
to be due not to investment--not to capital formation that "deepened"
the quantity and value of the capital available to amplify the productivity
of each worker--but instead to disembodied "total factor productivity
growth". This "total factor productivity growth" was seen
in large part as due to the background progress of science and technology,
and as determined by factors outside of--hence hardly influenced by--the
The second set of positions led to by the 1950s growth economics tradition
was theoretical: that differences in national economies' rates of
investment and savings would, in the long run, lead to some differences
in levels of output per worker but not to permanent differences in growth
rates: countries would array themselves in order, with the highest investment
countries having the highest productivity levels and living standards at
any single point in time. But all would see roughly the same rate of output-per-worker
growth: all would grow together.
The 1980s rebirth of growth economics in large part sprang from dissatisfaction
with these two sets of positions that had been the conclusions of the 1950s
growth economics tradition.
First, the theoretical conclusion that all national economies should, in
relatively short order, find themselves with approximately the same rate
of output-per-worker growth appeared to be false: there was little sign
that in the post-World War II period the different national economies of
the world had been "growing together". Second, the empirical conclusion
that the bulk of output-per-worker growth came from disembodied "total
factor productivity growth"--and was, hence, largely determined by
factors outside the scope of political economy--appeared unattractive as
well: good government, free trade, and high investment appeared to have
much stronger links to rapid economic growth than the first round of growth
economics had suggested.
The new growth economics tradition has now been at work for more than a
decade. For more than a decade the--largely technical, frequently over-mathematical,
frequently obscure--written literature in the "new growth economics"
tradition has been developing. Yet the overall lessons to draw from this
tradition have not been widely disseminated: "what it all means"
is obscure to those outside the relatively small circle of largely-academic
initiates--and is, if truth be told, obscure even to many inside
the relatively small circle of largely-academic initiates.
This essay takes a look at the "new growth economics" tradition,
and tries to outline what those making economic policy should take from
the tradition. There are four lessons:
There is still an enormous amount that remains unsettled in the literature
of the "new growth economics." But these four lessons appear clear--and
clear enough to begin to guide applications to economic policy.
- First, that it is very hard to continue to believe in the conclusions
of the 1950s literature--that the bulk of growth over the timespan of even
a generation is due to total factor productivity growth that springs from
pure research and development outside the economic system, and that national
economies' savings and investment efforts have some impact on relative levels
but little impact on relative long-run growth rates. This does not mean
that some do not try to reestablish the major positions of the 1950s tradition:
Greg Mankiw, for example, tries to interpret the cross-country pattern of
growth as supporting a dominant role for disembodied total factor productivity
growth. Robert Barro, for example, tries to find evidence for "conditional
convergence"--that even though countries have not been growing
at roughly the same rate, and even though differences in investment have
not been primarily associated only with differences in relative levels
of prosperity, that these conclusions of the 1950s will soon come to pass
as the world moves out of the current "transitional" stage to
achieve its "steady-state" relative income and productivity distribution.
The problem is that in order to rescue any one particular piece of the 1950s
set of conclusions, practically all of the other assertions have to be thrown
- The second lesson is that rapid total factor productivity growth is
not unconnected with but closely connected with economic factors: without
high levels of investment and without good government, increases in productivity
associated with the progress of science and technology and the diffusion
of better organizations and techniques simply do not happen.
- The third lesson is that untangling exactly which particular components
of investment promise the greatest benefits in terms of boosting productivity
growth is very hard. Practically everyone agrees that high rates of investment
in machinery and equipment are a crucial channel through which developing
economies assimilate the high-productivity modern technologies of the industrial
revolution--but some doubt that already-industrialized economies would see
similar boosts in productivity growth from higher machinery investment,
and claim that the growth-machinery correlation in already-rich economies
arises primarily because fast growth creates many investment opportunities,
and triggers high investment. Practically everyone agrees that a public
sector that fails to provide and maintain infrastructure can cripple private-sector
productivity--but some doubt that modern governments have the competence
to direct additional infrastructure dollars to high-value uses. Practically
everyone agrees that research and development expenditures have had very,
very high social rates of return. But research and development expenditures
are and are almost surely destined to be only small components of investment.
The natural conclusion is that--when we do not truly know which components
of investment are the most crucial ones, but have strong reason to
believe that many types of investment generate healthy rates of return for
society, and that some types of investment trigger enormous benefits in
faster productivity growth--good government policy involves placing eggs
in many baskets: boosting the different components of investment along a
very broad range.
- The fourth lesson is that the factors that have in the past led to
the conclusion that a good government takes steps to boost investment are
growing stronger, not weaker, as time passes. Improvements in communication
and in the speed of technological diffusion imply that there are growing
"wedges" between the private and the social returns to investments
that create new knowledge and new organizations that can effectively utilize
Long-Run Economic Growth
Compare America's standard of living today to 1870. The inflation-adjusted
productivity of American workers, real wages, and real GDP per head are
all now some ten times what they were then according to official statistics--and
there is good reason to think that official statistics substantially understate
A small part of this enormous amplification of material prosperity comes
from committing a somewhat greater share of production to investment: perhaps
20%. Another part of the past century's growth comes from committing a larger
share of production to education, boosting the skills and competencies of
American workers: perhaps 30%.
But the main engine of economic growth has been the advance of economically-useful
knowledge: better ways of making machines, better ways of using machines,
better ways of organizing production and communication, and better ways
of using limited natural resources. The advance of knowledge woul dhave
generated a three-fold multiplication in productivity, even if the shares
of national product invested in human and physical capital had not risen.
No one sane thinks that the economically-useful knowledge in the brains
of workers and the standard operating procedures of organizations advances
at a steady pace, unrelated to the rest of the economy and the polity. Even
the purest and most abstract of pure sciences depend on the number of and
the resources devoted to cosmologists, elementary partical physicists, and
paleontologists. The applied science and organizational practices that boost
the productivity of workers and businesses are closely tied to the rest
of economic life.
So there is no reason to think that the trend rate of productivity growth--a
function of the rates of investment and of the advance of economically-useful
knowledge--is a fixed constant. After World War II continental Europe grew
rapidly as it built its capital stock and worker skills back to pre-WWII
levels. But continental Europe did much better than simply return to its
pre-WWII growth trend: today it has output per capita levels more than forty
percent above what you would have expected from simple extrapolations of
pre-WWII trend growth.
On the eve of WWII Argentina was a wealthy member of the first world, perhaps
fourth in the world among nations in automobile ownership per capita and
higher in estimated GDP per capita than Germany, Italy, or France. Yet Argentina
has gone from 50% of America's--150% of Italy's--GDP per capita in 1950
to less than 30% today. There is no basis for the often-heard claim that
countries must learn to live with rather than try to change their long-run
growth trend, and every reason to think that pro-growth policies can nurture--and
anti-growth policies destroy--long-term economic growth.
Knowledge and Growth
But how about the argument that market forces will generate the "right"
amount of economic growth? We let market forces decide the number of books
relative to CD's to produce, and even to switch from analog LP to digital
CD systems. Why not let market forces decide how much of society's collective
work time and effort to devote to pursuing advances in economically-useful
The answer is that there is good reason to think that the Invisible Hand
will do a bad job. The Invisible Hand is very good at directing economic
activity when resources are scarce--either I have it or you have it--and
property rights are straightforward. But economically-useful knowledge is
not scarce in this sense: just because I am making use of it doesn't mean
that you cannot use the same piece of knowledge. And information wants to
be free: it is very, very hard to keep people from making full use of whatever
they know no matter who holds formal title to the "intellectual property."
The libertarian science fiction writer Robert A. Heinlein once set out the
principles of a [fictional] Fifth Socialist International: "Private
where private belongs, public where it's needed, with an admission that
circumstances alter cases." The nature of knowledge-as-commodity guarantees
that it is a broad and important area of the economy where public involvement
is needed: reliance on pure laissez-faire will not produce good outcomes.
To date the "endogenous growth theories" of economists are signposts
that point to problems and unresolved issues. They may provide a few principles
for how to think about the relationship between public policy, the advance
of knowledge, and economic growth. They do not provide settled directives,
or comfort for the dogmas of old ideologies. But political movements that
refuse to think about such issues are certain to become more and more irrelevant.
For the advance of knowledge continues to become--as it has been doing since
the eighteenth-century industrial revolution--a more and more dominant component
of modern economies.
What about economic policy? To some extent the principles derived from economists'
theories of knowledge and growth can guide thinking about the design
of a good economic growth agenda. In my view the policies to be extracted
from the principles were:
This growth agenda is bipartisan. For Democrats, it urgency is underlined
by the recognition that the United States' social insurance system was designed
for the pre-1973 rapid rather than the post-1973 slow pace of growth. Democrats
who value that social insurance system--who think that equality of opportunity
and social insurance are better principles than the perpetuation of privilege
and the multiplication of poverty--recognize that without faster long-term
growth America's social insurance system will be dead in two decades. Republicans
who fear that the New Deal puts too great a burden on the private economy
similarly recognize that only faster growth will diminish the relative size
of the social insurance state, and reduce the potential economic drag.
- Support pure research. Repair the damage done to public support
of civilian science and technology during the 1980s budget stringency, and
expand the federal government's commitment to basic research and technological
- Provide true public goods. Restore government investment in
- Boost private investment. Sharply reduce the federal deficit
to lower interest rates, boost investor confidence, and thus stimulate a
high rate of investment in machinery and equipment that is such an important
embodiment of technological knowledge.
- Make it easier for people to invest in themselves--through
"making work pay" so that joining the labor force will always
seem a more attractive option, through diminishing the cost of student loans,
through making it easy for individuals to invest in themselves through more
education and training.
- Expand markets by expanding world trade, because private investments
in knowledge are more valuable the greater is the size of the marketplace
over which they can be leveraged.
Over the past four years a proportion of this economic growth agenda has
been put into operation. Trade expansion--yes. Deficit reduction--yes (and
deficit reduction has paid dividends in higher investment and lower interest
rates perhaps twice as great as even optimistic projections had forecast).
But support for research and development--no. Increases in public investment
as opposed to redistribution--no. Making it easier for America's workers
to invest in themselves--largely no.
Because perhaps half of a pro-growth agenda has been enacted, America's
trend productivity growth rate is perhaps higher than it would otherwise
have been by some 0.2 percentage points of growth per year or so. While
masked in the short term by the business cycle, in the long term such a
boost to trend growth mounts up: by 2010 there will be perhaps an extra
$400 in annual U.S. GDP.
But we can do better.
of Economics Brad De Long, 601 Evans
University of California at Berkeley; Berkeley, CA 94720-3880
(510) 643-4027 phone (510) 642-6615 fax