A Comparison of Different Window Models for Measuring Technical Efficiency

This study analyzes the technical efficiency of Iran's 13 major cotton producing provinces over the period 2000-2012. To assess technical efficiency over time we use several non-parametric versions of the Window methodology. These include Window-DEA, Window-FDH, Window order- and Window order-m models. In a Window’s analysis one uses moving average patterns as a provincial observation in each period treated as if it is a different observation leading to a higher number of observations. The aim is to shed light on temporal patterns and heterogeneity in efficiency, time dependency and sensitivity of the results attributed to different efficiency measurement methodologies that are applied using the same dataset. It then invokes four consistency conditions to examine these scores and rankings. According to our results the models Window-FDH, Window order- and Window order-m are consistent. We were able to identify the most (least) efficient provinces and follow temporal patterns of their performance in the production of cotton. The findings reveal that provinces with larger scales of production are not necessarily more efficient. The results also indicate that although there are differences in the levels of efficiency across the models used, efficiency rankings from various models tend to support similar conclusions about the relative performance of the sample provinces.