This semester I have been teaching a class at BYU on
economic growth. One of the
anticlimactic moments in the class comes near the end when I try to summarize
the various theories and models of growth we have covered. We try to determine by an appeal to the
evidence which of these is the best explanation of observed growth
patterns. The bottom line is that we
don't really know.
Growth is one of the most important issues economist study. As Bob Lucas, a Nobel Prize winner from the University of Chicago has said, the consequences for human welfare from correctly understand the mechanisms that drive economic growth are simply profound. So profound, in fact, that when one begins to think seriously about growth, it is difficult to concentrate on other topics.
The modern experience with economic growth is unique in history. Prior to the 17th century standards of living changed very slowly over time, if at all. Something happened in the 18th century that changed that dynamic dramatically. Standards of living now rise rapidly from one generation to the next. And this is true even for some of the poorest places on the earth. That is, even the typical citizens of poor countries like Mali or Bangladesh are better off today than their parents were a generation ago.
As economists, we now understand growth a great deal better than we did even twenty years ago, but the root causes are still elusive. Back in the 1950's Nobel Laureate Robert Solow of M.I.T. showed that growth in per capita income cannot be driven in the long-run by simple mechanisms such as the accumulation of physical capital. Such mechanisms can cause growth spurts as economies build up their capital to some long-run optimal level, but they cannot explain sustained episodes of rising per capita income over very long periods of time. The Japanese post-war growth spurt is a prime example. Growth was very rapid for several decades, but once the Japanese economy reached a level of capital and output per capita similar to the rest of developed world, growth rates fell to more common rates. Today the Chinese economy is booming in a similar way and it will likely continue to do so for quite a while. But eventually growth rates in China will also fall to rates closer to the rest of the world. By then, of course, incomes in China will be very similar to levels in other developed economies.
Later work by Solow showed that much of the growth we observe is attributable to increases in technology. Discovery of new knowledge and application of that knowledge to the production of goods and services allows us to produce more goods and services with the same amount of labor and capital. This discovery only pushes the question back a stage. If growth in standards of living is driven by growth in technology, what causes the rise in technology?
Modern growth theory has no shortage of plausible explanations. For example, some economists argue that economic growth is driven by increasing returns-to-scale. In other words, doubling the capital and labor and all other factors in an economy, more than doubles the number of goods and services that are produced. Others point to learning-by-doing arguments. As firms produce more they discover, as a byproduct of production, more efficient ways to produce and these production methods can be adopted by other firms, expanding overall technology. Still others believe that growth is driven primarily by R&D firms that are attempting to corral monopolies on new goods. The lure of these monopoly profits induces the firms to hire researchers to work on developing a new and improved product that will allow them to drive existing producers out of the market. Many researchers focus on the role of government and public policy in generating improvements in technology and human capital. Each of these mechanisms implies a mostly unique set of economic indicators should be highly correlated with economic growth.
Empirical evidence able to distinguish these various hypotheses is difficult to come by. The main reason is that while any of these theories in isolation implies a particular set of data should explain growth, when taken together all the data tell much the same thing. For example, some theories suggest that higher rates of primary or secondary education should cause faster growth. Other theories suggest that an increase in legal enforcement of property rights should cause more rapid growth. In reality we observe all three of these rising simultaneously. Is rising education causing more rapid growth and better enforcement of property rights? Or are property rights causing rising growth and, hence, a better educations system. Or are all three being driven by some other factor.
An appeal to the data does not allow us to resolve the issue. In one sense this is good as it may imply that the sources of growth are various and if one source disappears growth can be sustained by the others. This is not very satisfying to students or professors hoping to learn the ultimate cause of economic growth, however.