Scale efficiency from constant and variable returns frontiers

Divides technical efficiency measured against a constant returns to scale frontier by technical efficiency against a variable returns to scale frontier for the same unit and orientation. The variable returns score removes the effect of size, so the ratio isolates the part of inefficiency due to operating at a size other than the most productive scale.

Signature

SE = TE_CRS / TE_VRS
Inputs
InputsDefinitionUnit
TE_CRSScore against a frontier that allows no economies or diseconomies of scalescore from 0 to 1
TE_VRSPure technical efficiency score against a frontier that allows returns to scale to vary, greater than zeroscore from 0 to 1
Output
SEShare of technical efficiency under constant returns not explained by pure technical inefficiencyscore from 0 to 1

Function

Frontier efficiency measurement function

Maps a decision-making unit's observed inputs, outputs and costs, and a production or cost frontier estimated for comparable units, to efficiency scores between 0 and 1. A score of 1 places the unit on the estimated frontier. The frontier itself is usually estimated by data envelopment analysis, a linear programming method, or by stochastic frontier analysis, a parametric method that separates inefficiency from random noise.

Try this function

Implementations

  • Excel

    Scale efficiency from two DEA scores

    With the constant returns score in ScoreCRS and the variable returns score in ScoreVRS, Excel returns scale efficiency.

    =ScoreCRS/ScoreVRS

Assumptions

  • Same units, variables and orientation for both frontiers

    Both scores come from the same sample, inputs, outputs and orientation, differing only in the returns to scale assumption, so TE_CRS is no greater than TE_VRS.

  • Direction of scale inefficiency found separately

    A scale efficiency below 1 does not show whether the unit is too small or too large. A further model under non-increasing returns to scale identifies which applies.

Worked examples

  • Small community hospital

    A community hospital scores 0.72 against a constant returns frontier and 0.80 against a variable returns frontier. Scale efficiency is 0.9, so most of its measured inefficiency is pure technical inefficiency. The figures are illustrative.

    TE_CRS = 0.72; TE_VRS = 0.80; SE = 0.9

Common errors

  • Recommending mergers from scale efficiency alone

    A low scale efficiency score measures production cost only. Centralising services might reduce measured cost while increasing travel, weakening access or reducing resilience, which the score does not capture.

Sources

  • Textbook on scale efficiency in DEA

    Coelli TJ, Rao DSP, O'Donnell CJ, Battese GE. An Introduction to Efficiency and Productivity Analysis. 2nd ed. New York: Springer; 2005. Decomposition of constant returns technical efficiency into pure technical efficiency and scale efficiency, with a non-increasing returns model to identify the nature of scale inefficiency.

    View source →

Canonical Identity

Stable URI · Machine-readable · Resolvable · CC BY 4.0