Abstract
In this article, we conduct a productive efficiency analysis using the two major methods: stochastic frontier analysis (SFA) and data envelopment analysis (DEA). First, we demonstrate the steps and assumptions involved in SFA and DEA, and highlight their strengths and weaknesses. Then we apply them to the case of bleached softwood kraft pulp production to derive technical and allocative efficiency measures for 102 mills in the world. We find that in general, the SFA cost-efficiency levels are higher than their DEA counterparts, and much of the cost inefficiency is caused by inappropriate input mix, or allocative inefficiency, instead of the way with which inputs are converted into outputs, or technical inefficiency. In addition, SFA and DEA results rank some of the producers in different orders. We also find that while cost-efficiency levels vary in different regions, most producers have substantial opportunities to make improvements. Finally, we discuss implications of our study. FOR. SCI. 46(4):558–569.