 n
this essay, Dr. Robert J. Barro, the Robert C. Waggoner
Professor of Economics at Harvard University, discusses his
highly cited work in the framework of the evolution of
economic growth models. In our analysis of high-impact papers
in economics, five of Dr. Barro’s papers were cited a total
of 852 times, making him the most-cited economics author of
the past decade. His most-cited paper is "Economic growth
in a cross section of countries" (Quarterly
Journal of Economics 106[2]:407-43, May 1991) which has
been cited 577
times to date. ESI data indicate that Dr. Barro’s citation
statistics currently include 18 papers with a total of 1,202
citations.
Dr. Barro is also
a Senior Fellow of the Hoover Institution, a contributing
editor of the Wall
Street Journal,
a contributing author for the bipartisan, weekly e-zine
Intellectualcapital.com, and has authored several books.
|
Research on economic gr owth
was very active from the late 1950s through the 1960s. (A brilliant
paper by Frank Ramsey in the 1920s anticipated much of the later
research.) The work through the 1960s produced the "neoclassical
growth model," which is a core conceptual framework used now by
most economists. The model stresses the accumulation of capital,
broadly defined, as a source of growth. An important prediction of
this model is convergence, that is, a tendency for poor economies to
catch up to rich ones. However, the model is  less interesting in its
predictions about long-run economic growth, which depends entirely on
unexplained factors, especially the exogenous rate of technological
progress.
The research program that led to the neoclassical growth model
became less active by the end of the 1960s. One reason was that
researchers were no longer making important theoretical advances.
Another was that the model was not being used effectively for
empirical applications.
Economic growth became a vigorous area of research again in the
late 1980s, propelled by a new line of theory that focused on the
origins of technological change. This work stressed that technological
advance amounted to the creation of new ideas, which differed from
standard inputs to production because ideas could be used freely by
any number of producers. This characteristic of ideas meant that some
kind of monopoly power over new products or processes—that is, a
type of imperfect competition—was necessary for motivating the
discovery of better technologies. Numerous models with these features
were developed during the 1990s. These theories are described as
"endogenous growth models," because they determine within
the model the rate of technological change and, hence, the economy's
long-term growth rate.
In the early 1990s, empirical work on
the determinants of economic growth also became important. However,
most of this work related to extended versions of the neoclassical
growth model, which was developed decades earlier, rather than to the
recent endogenous growth theories. My papers that Essential
Science Indicators selected as heavily cited during the
1990s are contributions to this empirical literature. The articles
include my 1991 paper from the Quarterly Journal of Economics
("Economic growth in a cross section of countries" and my
1992 paper from the Journal of Political Economy
("Convergence," 100[2]:223-51, April 1992; co-authored with
Xavier Sala-i-Martin, with whom I also wrote the 1995 book Economic
Growth).
The recent empirical work—partly
across countries, partly across regions of countries, and partly
across time periods—suggests that the convergence implications of
the neoclassical growth model have a lot of empirical content.
However, for heterogeneous economies, such as a broad group of
countries, convergence operates only in a conditional sense.
Specifically, countries grow faster if they are poorer but only after
holding constant an array of explanatory variables. These variables,
which enter into extended forms of the neoclassical model, include
government policies and institutions, the nature of educational
systems, people's propensities to save and to have children, possibly
a nation's colonial history and religious traditions, and so on.
Because poor countries typically have less favorable values overall
for these explanatory variables (which explains why they are poor),
they tend not to grow faster than average. An important part of the
research program has been to isolate particular explanatory variables,
especially indicators of policies and institutions, that affect growth
in a significant way. Results that look important include a positive
effect from successful maintenance of the rule of law, a positive
effect from educational attainment, a negative effect from high
fertility rates, and a negative effect from high inflation.
This type of cross-country empirical
work has exerted a good deal of influence, as indicated by the
citation counts. However, the methodology is also controversial,
because there are numerous problems in making causal and robust
inferences from the available data. I agree that these problems are
serious, but I also feel that, if policy inferences relevant for
long-term growth cannot be drawn from the cross-country data, then
they probably cannot be garnered from any data. I also think that many
of the results already obtained provide clear insights into the
determination of growth rates. In any event, this area of research
remains active, and we are still learning more about appropriate
methodologies, as well as the details of empirical relationships.
In contrast, empirical analysis has
thus far been less successful in verifying the importance of the
recent endogenous growth theories, notably their implications for
technological progress. There has been some interesting work involving
the role of R&D outlays and the nature of the diffusion of
technologies from leading countries to followers. I anticipate that
empirical work in these areas will be important in the next decade.
Most likely, we will ultimately find that some combination of the
recent endogenous growth theories with the neoclassical growth model
will provide the best framework for understanding the determinants of
economic growth.
Dr. Robert J. Barro
Harvard University
Department of Economics
Cambridge, MA, USA
|
cites
:
