Recent research on economic growth.

• Author: Barro, Robert J.
• Position: Research Summaries

Economic growth has been a lively area of research since the mid-1980s. Important advances have been made in both theory and empirical analyses, and much of this progress has been reported on at NBER conferences on growth, which first took place in 1989.

My research over the last five years has focused on growth; the main results are included in Economic Growth, coauthored with Xavier Sala-i-Martin and forthcoming this fall from McGraw-Hill.(1) Aimed at the level of first-year graduate students in economics, the book combines new results with expositions of theories from the 1950s through the 1990s.

To appreciate the importance of growth rates, start with the observation that the U.S. real per capita GDP, measured in 1985 dollars, grew from $2244 in 1870 to $18,258 in 1990, or by 1.75 percent per year. This performance gave the United States the highest real per capita GDP in the world in 1990 (with the possible exception of the United Arab Emirates, an oil producer with a small population). If the U.S. growth rate had been just one percentage point per year less--that is, 0.75 percent per year--then the real per capita GDP in 1990 would have been $5519, only 30 percent of the actual value. Then, instead of ranking first in the world in 1990, the United States would have ranked 37th out of 127 countries with data available. To put it another way, the U.S. real per capita GDP in 1990 would have been close to that in Mexico and HUngary, and would have been about $1000 less than that in Portugal and Greece.

Alternatively, if the U.S. growth rate had been one percentage point per year higher--that is, 2.75 percent per year--then the real per capita GDP in 1990 would have been $60,841, or 3.3 times the actual value. A real per capita GDP of $60,841 is well outside the historical experience of any country and may, in fact, be infeasible. But, in any event, a continuation of the long-term U.S. growth rate of 1.75 percent per year implies that the United States would not attain a real per capita GDP of $60,841 until the year 2059.

For 114 countries between 1960 and 1990 the average growth rate of real per capita GDP was 1.8 percent per year--nearly the same as the long-term U.S. rate. The range is -2.1 percent per year for Iraq to 6.7 percent per year for South Korea. Thirty-year differences in growth rates of this magnitude have enormous consequences for standards of living. South Korea raised its real per capita GDP from $883 in 1960 (rank 83 out of 118 countries) to $6578 in 1990 (rank 35 of 129), while Iraq lowered its real per capita GDP from $3320 in 1960 (rank 23 of 118) to $1783 in 1990 (rank 82 of 129).

In 1990, the mean of real per capita GDP for 118 countries was $2737. The highest value--$18,399 for the United States--was 65 times the lowest value--$285 for Ethiopia. To understand why countries differ this much in standards of living requires knowing why they experience correspondingly sharp divergences in long-term growth rates. Even small differences in these growth rates, when cumulated over a generation or more, have much greater consequences for standards of living than the kinds of short-term business fluctuations that typically have occupied most of the attention of macroeconomists. To put it another way, if we can learn about government policy options that have even small effects on the long-term growth rate, then we can contribute much more to improvements in standards of living than has been provided by the entire history of macroeconomic analysis of countercyclical policy and fine-tuning. Economic growth is the part of macroeconomics that really matters.

Modern growth theory begins with the neoclassical model, as developed by Frank Ramsey (1928), Robert M. Solow (1958), Trevor W. Swan (1956), David Cass (1965), and Tjalling C. Koopmans (1965). A key prediction of these models--which has been exploited seriously as an empirical hypothesis only in recent years--is conditional convergence. The lower the starting level of...