An International Comparison of Productivity Change in Agriculture and the Economy as a Whole, Using Production-Frontier Approa

A common and longstanding assumption in the economic growth literature has been that total factor productivity growth is lower in the agriculture sector than in the rest of the economy. Using a stochastic production frontier finite mixture model, labor productivity change is decomposed into catch-up, technological change and factor accumulation effects and stochastic shocks. This decomposition is investigated separately in the agriculture sector and the economy as a whole using a balanced panel data set of 45 countries in different development stages during the time period 1967-1992. The impact of labor productivity change components on the evolution of the cross-country counterfactual distribution of labor productivity is also analyzed. For the overall economy, the empirical results indicate that growth and the twin-peak distribution of labor productivity are driven by capital deepening. However, the results for the agriculture sector suggest that labor productivity distribution is brought by total factor productivity changes rather than factor accumulation. Furthermore, the agriculture sector exhibits reductions in capital per worker as well as stronger catch-up and technological change effects. Thus, growth of the rest of the economy appears to owe more to capital deepening and resource reallocation from agriculture than to faster productivity change.