Efficiency and productivity measurement with persistent benchmarks

In this note we propose a methodology to estimate the efficiency and productivity of decision making units based on persistent benchmarks. Using panel data we identify convex combinations of observations which dominate the evaluated unit over an extended period of time. Thereby, our approach takes into account that the restructuring of production processes or the adjustment of economic policy according to the optimal benchmark may not be possible instantaneously. We apply our model to an analysis of persistent development benchmarks using a sample of Sub-Saharan countries.