Reducing Weight Flexibility in Data Envelopment Analysis
Robert G. Dyson, Emmanuel Thanassoulis
Abstract
Robert G. Dyson, Emmanuel Thanassoulis
Abstract
This paper explores the consequences of total weight flexibility in data envelopment analysis (DEA) assessments of the efficiency of decision-making units, and it suggests one possible way of limiting such flexibility. It is suggested that total weight flexibility can lead to some decision-making units being assessed, in effect, on only a small subset of their inputs and outputs, while the rest of their inputs and outputs are all but ignored. Constraining the weights in DEA assessments overcomes this problem. The paper suggests one way in which constraints can be placed for the case where the decision-making units to be assessed use only a single input. The method is illustrated using data on local-authority rates departments. Finally, the paper discusses the interpretation and usefulness of the information obtained from DEA assessments involving weights constraints.
OpenAlex reports 21 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper explores the consequences of total weight flexibility in data envelopment analysis (DEA) assessments of the efficiency of decision-making units, and it suggests one possible way of limiting such flexibility. It is suggested that total weight flexibility can lead to some decision-making units being assessed, in effect, on only a small subset of their inputs and outputs, while the rest of their inputs and outputs are all but ignored. Constraining the weights in DEA assessments overcomes this problem. The paper suggests one way in which constraints can be placed for the case where the decision-making units to be assessed use only a single input. The method is illustrated using data on local-authority rates departments. Finally, the paper discusses the interpretation and usefulness of the information obtained from DEA assessments involving weights constraints.
Key concepts: Data envelopment analysis, Flexibility (engineering), Computer science, Limiting, Operations research, Purchasing, Project management, Mathematical optimization