DMus

DMUs, or decision-making units, are the entities whose efficiency is assessed in data envelopment analysis (DEA). In its simplest form, a DMU uses a set of inputs to produce a set of outputs. Examples include hospitals, schools, bank branches, or manufacturing plants. DEA compares multiple DMUs to determine how efficiently each converts inputs into outputs, relative to an empirical frontier formed by the observed units. All DMUs are assumed to be similar in processes and functions, so that comparisons are meaningful.

Common inputs include labor, capital, and operating costs, while outputs can include services rendered, units produced, or outcome measures such as patient discharges or revenue. The efficiency score is typically between 0 and 1, with 1 indicating a unit on the efficient frontier.

DEA models differ in their assumptions about returns to scale. The CCR model assumes constant returns to scale, providing an overall technical efficiency score. The BCC model assumes variable returns to scale, allowing separation of pure technical efficiency from scale effects. Extensions include cross-efficiency evaluation, where DMUs are judged from the perspectives of other units, and non-radial models that address input and output slacks rather than proportional reductions.

Applications of DMU analysis span public and private sectors, including evaluating hospital performance, school productivity, bank branch efficiency, and manufacturing plants. Limitations include sensitivity to the choice of inputs and outputs, dependence on data quality, and the assumption that DMUs are comparable. DEA provides a relative efficiency frontier rather than an absolute measure, helping identify best practices and areas for improvement among similar units.