| Data Envelopment Analysis (DEA) is a linear programming tool that can handle multiple inputs and outputs in a non-parametric fashion. DEA provides an unbiased analysis of Decision Making Units (DMUs). It can provide a more sophisticated alternative to prequalification systems as well as eliminate any subjectivity from the decision-maker. A three-stage model, employing an output-oriented variable returns-to-scale DEA model, is proposed and compared against an established model. The three stages involve receiving a letter of required bonding, followed by the DEA analysis, and ending by reducing the amount of prequalified contractors to a predetermined amount, referred to as shortlisting. The DEA model produced average efficiencies between 70 and 85 percent for the seven contracts evaluated. Many of the top and bottom contractors received similar rankings in each of the two models. DEA should be a welcomed addition to the construction industry. |
