Efficiency Measurement of Bridge Management with Data Envelopment Analysis

This study explored the feasibility of data envelopment analysis (DEA) to measure objectively the efficiency of Interstate bridge replacement and rehabilitation programs of the state highway agencies in the United States. Past applications of DEA indicated that this method showed promise for such assessments because of the ability to take into account the effect of uncontrollable factors and scale efficiency. The current study developed a modified input-oriented variable returns to scale model that treated the uncontrollable variables as uncontrollable. The two variables selected as outputs were (a) the change in area of structurally deficient bridges and (b) the change in area of functionally obsolete bridges between two consecutive years. The six variables selected as inputs were (a) expenditures on Interstate bridge replacement and rehabilitation, (b) average daily traffic on Interstate bridges per deck area, (c) proportion of Interstate truck vehicle miles traveled (VMT) to Interstate total VMT, (d) average age of Interstate bridges, (e) annual precipitation, and (f) annual freeze–thaw cycles. For each state, the DEA efficiency score and the corresponding target states were identified. The performances of the states and the possible reasons for inefficiencies were assessed. The Malmquist productivity index model was applied to calculate the technical efficiency change (catch-up effect) and the technological change (frontier shift). Over the analysis period, an overall improvement was observed in the frontier of the studied industry; this improvement suggested within this period a general improvement in the efficiencies of the Interstate highway bridge replacement and rehabilitation programs at highway agencies in the United States.