BY DIETRICH VOLRATH 13/11/2015
Most macroeconomic textbooks make a vague distinction between the ‘short run’ and the ‘long run’, typically saying the short run is somewhere between a few quarters and a few years, while the long run is perhaps five years out. When growth economists say ‘long run’, they really mean it. Studying long-run economic growth involves looking over decades, centuries, and occasionally millennia.
Figure 1 plots both global output per worker and global population from the year 0 AD to 2010. As can be seen, for the period from 0 to 1000 both series are close to flat, indicating very slow growth. Between 0 and 1000, GDP per capita did not grow at all. From 1000 to 1870, it grew at just about 0.2 percent per year. But from 1870, just as population begins to explode, so does output per capita, which grew at about 1 percent per year until 1950, and then by 1.6 percent per year from 1950 until today.
This data shows that sustained economic growth is a relatively recent historic event. If we extended the figure back further, we would see that from the onset of settled agriculture around 10,000 BCE to the beginning of the 19th century there was not any appreciable growth in world output per capita. There were some periods in which living standards did surge ahead, as under the Roman Empire, or in Western Europe following the Black Death, or in China during the Song dynasty. But these surges never developed into the sustained economic growth that has slowly been spreading around the world since about 1850.
Unified growth theory (Galor, 2011) is a body of research that is trying to explain what happened in the middle of the 19th century to set at least some countries on the path to sustained growth. At its core, the theory tells us that there must have been a fundamental change in the relationship between living standards and population growth to allow for sustained economic growth. Prior to about 1850, increases in living standards tended to lead to increases in population, which then in turn caused living standards to fall. This is the basic logic described by Thomas Malthus in 1798. But starting in the UK around 1850, and then following quickly across much of Western Europe and the US, increased living standards did not generate population growth sufficient to push those standards back down. As the 20th century dawned, rising living standards became associated with lower population growth rates in what is called the Demographic Transition. Sustained economic growth is partly the result of this breakdown of the old Malthusian population response.
Whether this breakdown occurred because technology was changing so rapidly that population growth could not keep up, or because technology directly changed the incentives of families towards having fewer children, is not obvious. And this is where the study of economic growth bleeds into the study of economic history. By digging into history, we are hoping to identify the origins of the technological and demographic changes. The issue is that we have so few examples of the transition to sustained growth, and in each of them a number of factors may have been involved. For the UK alone, people have proposed the Glorious Revolution, common law, the Enlightenment, canals, colonies, finance, coal, steam engines, spinning jennies, and pure dumb luck as reasons for its take-off to sustained growth. It could be any, all, or none of these.
Whatever the ultimate reason, sustained growth did begin, and we are living in a unique historical era. And we have been living in that era for long enough to be able to make some generalizations about sustained economic growth. To do so, we can examine the relatively short period of time from 1870 to 2010 in which we have more reliable data. Figure 2 plots output per capita for six countries over those years. The output per capita numbers are plotted on a log scale, so that movements up the vertical axis capture percent changes.
The first thing to note is that the growth rate (i.e. the slope) of the lines for the UK, US, and Germany are similar, and persistent over time, implying roughly 1.8 percent growth in output per capita per year. When sustained growth hit, as it did in these countries in the mid-19th century, it had remarkably similar effects on all of them.
As time rolled forward, sustained growth began to diffuse to other countries. In Figure 2 you can see both South Korea and China experience incredibly rapid growth starting in 1950 and 1980, respectively. In both cases growth remained higher than 1.8 percent per year for decades, and this allowed them to converge to the rich country levels of output per capita. But these rapid growth rates don’t persist. South Korea, for example, reached parity with Germany in output per capita and it now grows at roughly 1.8 percent per year as well. China is on a similar trajectory, and its growth rate is falling as it continues to get richer over time.
This diffusion of sustained growth is incomplete, however, as one can see with respect to Kenya. It, and numerous other countries in Sub-Saharan Africa, South Asia, and Central America remain stubbornly stuck at low levels of output per capita, and have not experienced any sustained economic growth. The current dispersion of output per capita across countries is due to the incomplete diffusion of sustained economic growth. This lack of diffusion could be due to institutional differences, geographical constraints, or cultural factors, but at this point we do not have enough evidence to definitively pin down the reasons.
Robert Solow (1956) formed the baseline for our understanding of sustained economic growth. He explained why economies tend to gravitate towards what we call a balanced growth path with a constant growth rate, as in the UK, US, or Germany. His explanation was that the marginal return on capital rises as the economy gets farther from the balanced growth path. Hence output goes up rapidly when an economy is relatively poor compared to its balanced growth path, and it converges back towards its balanced growth path over several decades. The classic example of this is Germany after World War II, which you can see in Figure 2.
Solow also explained why countries do not grow rapidly forever, as shown by the experiences of both Germany and South Korea. As they accumulated more capital, the marginal return fell, and output grew only slightly as they invested. Hence the growth rate slowed down and they ended up along a balanced growth path. This same reasoning is why growth economists fully expect that China’s growth rate will slow down, and it will find itself with a level of output per capita similar to currently rich countries, but growing at the same steady 1.8 percent per year.
Why 1.8 percent, though? The last thing the Solow models tells us is that this growth rate is determined entirely by the growth rate of technology. In other words, capital accumulation does not by itself generate sustained growth. Preferences and policies influencing savings rates or education levels can explain the small differences between the US and UK, for example, but they do not explain the long-run growth rate itself.
Given the importance of technology for explaining sustained growth, endogenous growth theory took on the task of studying the incentives involved in innovation, why they lead to that specific rate of 1.8 percent growth, and if there is anything we can do to raise (or lower) that rate. These theories show us that the rate of growth is determined in part by the ability of firms to earn rents from their innovations. To the extent that policy can influence the rents that firms earn on innovations, such as with patents, R&D subsidies, trade policy, or taxes, it can influence the rate of growth.1
But that influence is probably only temporary. Charles Jones (1995a, 1995b) argued that the 1.8 percent growth rate ultimately doesn’t depend on policies at all. The support for this comes from the observation that while the share of GDP spent on R&D and the share of workers doing R&D have been rising steadily over the 20th century, the long-run growth rate has not changed. Instead, the rate of growth is pinned down by the inherent speed of technological progress, which is tied positively to the growth rate of the population. In this case, policies regarding innovation can temporarily change the growth rate, but not permanently.
What history has shown is that sustained economic growth is incredibly resilient. Changes in tax policies, wars, trade policies, political realignments, and the like have not had much, if any, effect on the sustained long-run growth rate in currently developed countries. If long-run growth rates are going to significantly change in the future (for better or worse), then this will be an unprecedented event.
At the same time, sustained economic growth was incredibly difficult to achieve in the first place. An answer to what ignites sustained growth is important not only because it could provide a way out of poverty for millions, but because it could alert us to crucial institutions or policies that support sustained growth in currently rich countries. Economic growth has been very resilient over the last century and a half, but that does not mean it will be in the future. The origins of sustained growth are relevant not only for explaining the past, but for forecasting the future.
Aghion, P. and Howitt, P. (2009), “The Economics of Growth”, MIT Press.
Galor, O. (2011), “Unified Growth Theory”, Princeton University Press, Princeton, NJ.
Jones, C. (1995), “R&D-Based Models of Economic Growth”, Journal of Political Economy 103 (1995), 759–784.
Jones, C (1995), “Time series test of endogenous growth models”, Quarterly Journal of Economics 110, 495–525.
Maddison, A (2008), “Historical Statistics of the World Economy”, available at http://www.ggdc.net/, 2010.
Romer, P. (1986), “Increasing returns and long-run growth”, Journal of Political Economy 94, no. 5, pp.1002–1037.
Romer, P. (1990), “Endogenous Technological Change”, Journal of Political Economy 98, no. 5, pp.S71–S102.
Solow, R. (1956), “A Contribution to the Theory of Economic Growth”, The Quarterly Journal of Economics 70, no. 1, pp. 65–94.