Using Panel Data Analysis to Estimate DEA Confidence Intervals Adjusted for the Environment

Abstract

This paper illustrates three concepts new to the data envelopment analysis (DEA) literature, and applies them to data from Canadian paratransit agencies. First, it predicts valid confidence intervals and trends for each agency’s true efficiency. Second, it uses panel data analysis methodology, a set of statistical procedures that are more likely to produce valid estimates than those commonly used in DEA studies. Third, it uses a new method of identifying and adjusting for environmental effects that has more power than conventional procedures.