Benchmarking a nationwide non-profit service organization using DEA

We used 2009 data from a nationwide church organization to assess operational productivity and efficiency utilizing Data Envelopment Analysis (DEA). This non-parametric methodology has been used to assess not-for-profit organizations before, but this is one of the first papers using this technique to study a religious organization and a network of 65 synods. Such homogenous Decision Making Units (DMU´s) and the uniform and measurable nature of their resources and deliverables (inputs and outputs) made this an ideal case for analysis. By using a Variable Returns to Scale and output-oriented model with as many as three inputs and six outputs, as many as 60% of the synods were found to be efficient. Nevertheless, the model produces enough nuances to allow identification and further analysis of the DMU´s deemed to be inefficient, as well as the selection of “best practice” performers to use in organizational improvement efforts. In addition to considering and discussing the practical use of slack variables and multipliers, the paper also introduces some statistical and graphical tools to further analyze the results.