Bridging the gap between partial and total factor productivity measures using directional distance functions

Technical dependencies as well as data constraints limit our ability to allocate inputs across sectors and hence our ability to measure sectoral productivity. We adapt a directional measure of efficiency to the measurement of sector-specific productivity that does not require allocating all inputs across sectors. Applied to the agricultural sector of a group of countries, the results show important differences in livestock and crops productivity growth. Commonly used partial factor productivity measures for livestock and crops tend to overestimate productivity growth in most developing countries while underestimating it in European countries.