DP5549 Industrial Productivity in 51 Countries, Rich and Poor

Author(s): Orsetta Causa, Daniel Cohen
Publication Date: March 2006
Keyword(s): industry, Lucas Paradox, productivity
JEL(s): F15, L16
Programme Areas: International Macroeconomics
Link to this Page: cepr.org/active/publications/discussion_papers/dp.php?dpno=5549

This Discussion Paper analyses 23 industrial sector in a sample of 51 developed and developing countries. It distinguishes the contribution of five factors: private capital, infrastructure, education, trade integration, and net efficiency. Several relatively small handicaps, combined multiplicatively, can make a country poor or very poor. In average, the average productivity of the industrial sector is indeed the product of about five times 70%. But 0.70 to the power of five is 17%. The least productive country in the sample, Bangladesh, has a productivity level worth about 2% of that of the richest nations. From this perspective, industry is not much different from aggregate GDP such as analysed in Cohen and Soto (2004) where a similar picture emerged The paper then sheds light on the effect of TFP differential between industry and GDP at large on the relative price of manufactured goods. We show that productivity differentials explain about half of the relative price discrepancy. It then analyses the extent to which this is an explanation of the Lucas Paradox. So far as manufacturing is concerned the paper highlights an "anti-Lucas" paradox whereby the capital output ratio is higher in poor countries than in rich countries. This result tends to deflate the theories according to which fear of expropriation is the critical explanation of the low level of capital in the economy at large.