The Dynamics of Agricultural Productivity Gaps: An Open-Economy Perspective

Xiao Ma is an assistant professor in the HSBC Business School of Peking University. He holds a Ph.D. in Economics from the University of California, San Diego. His primary research focuses on international trade, macroeconomics, and economic growth. He has published multiple papers in journals such as American Economic Journal: Macroeconomics, International Economic Review, Journal of International Economics, Journal of Political Economy Macroeconomics, Review of Economics and Statistics, Economic Research Journal (in Chinese), Journal of World Economy (in Chinese), and China Economic Quarterly (in Chinese). He has been awarded the National Natural Science Foundation of China's Excellent Young Scientists Fund and the WTO Young Economists Prize.

Agricultural productivity gaps have persisted over the last three decades and have even widened in major economies like China and India, despite global decreases in agricultural employment. This paper takes an open-economy perspective on these patterns, arguing that increases in comparative advantage in non-agricultural activities can lead to growing gaps in sectoral value added per worker in general equilibrium in spite of worker reallocation out of agriculture. We formalize this idea in a quantitative multi-sector model with imperfect mobility of labor across sectors, frictional international trade, and uneven growth of relative prices and real productivity levels across countries and sectors. We estimate the model to match time series patterns of exports by sector and country, among other statistics, which help discipline the model’s real sectoral productivity growth rates. Quantitatively, the model does well in accounting for the dynamics of productivity gaps across countries. Consistent with the model’s predictions, previous exports by sector are strong predictors of agricultural productivity gaps in the current cross-section of countries, in line with the view that the gaps reflect real economic factors rather than mis-measurement.