M-RCBG Faculty Working Paper No. 2014-04
Asiaphoria Meets Regression to the Mean
Lawrence H. Summers and Lant Pritchett
October 2014
Abstract
Consensus forecasts for the global economy over the medium and long term predict the world’s economic gravity will substantially shift towards Asia and especially towards the Asian Giants, China and India. While such forecasts may pan out, there are substantial reasons that China and India may grow much less rapidly than is currently anticipated. Most importantly, history teaches that abnormally rapid growth is rarely persistent, even though economic forecasts invariably extrapolate recent growth. Indeed, regression to the mean is the empirically most salient feature of economic growth. It is far more robust in the data than, say, the much-discussed middle-income trap. Furthermore, statistical analysis of growth reveals that in developing countries, episodes of rapid growth are frequently punctuated by discontinuous drop-offs in growth. Such discontinuities account for a large fraction of the variation in growth rates. We suggest that salient characteristics of China—high levels of state control and corruption along with high measures of authoritarian rule—make a discontinuous decline in growth even more likely than general experience would suggest. China’s growth record in the past 35 years has been remarkable, and nothing in our analysis suggests that a sharp slowdown is inevitable. Still, our analysis suggests that forecasters and planners looking at China would do well to contemplate a much wider range of outcomes than are typically considered.