Evaluating the efficiency of local municipalities in providing traffic safety using the Data Envelopment Analysis. (May 2015)

• Record Type: Journal Article
• Authors: Alper, Doron; Sinuany-Stern, Zilla; Shinar, David
• Is Part Of: Accident analysis and prevention. Volume 78(2015)
• Page Start: 39
• Page End: 50
• Publication Date: 2015-05
• Subjects: Traffic safety -- Data Envelopment Analysis -- Local municipalities -- Efficiency -- Demographic parameters -- Two-stage DEA -- Undesirable outputs; Accidents -- Prevention -- Periodicals; Accident Prevention -- Periodicals; Accidents -- Prévention -- Périodiques; 363.106
• DOI: 10.1016/j.aap.2015.02.014
• Languages: English
• ISSNs: 0001-4575

Abstract

Highlights: We evaluated the relative efficiency of 197 local municipalities on their traffic safety. The Data Envelope Analysis (DEA) was used to rank-order the municipalities in terms of their input/output ratios. A regression analysis showed that the municipality's population size is a key factor: with the smaller communities being more efficient. Abstract: The purpose of this study was to estimate the relative efficiency of 197 local municipalities in traffic safety in Israel during 2004–2009, using Data Envelopment Analysis (DEA). DEA efficiency is based on multiple inputs and multiple outputs, when their weights are unknown. We used here inputs reflecting the resources allocated to the local municipalities (such as funding), outputs include measures that reflect reductions in accidents (such as accidents per population), and intermediate variables known as safety performance indicators (SPI): measures that are theoretically linked to crash and injury reductions (such as use of safety belts). Some of the outputs are undesirable. Using DEA, the local municipalities were rank-scaled from the most efficient to the least efficient and required improvements for inefficient municipalities were calculated. We found that most of the improvements were required in two intermediate variables related to citations for traffic violations. Several DEA versions were used including a two-stage model where in the first stage the intermediate variables are the outputs, and in the second stage they are the inputs. Further analyses utilizing multiple regressions were performed to verify the effect of various demographic parameters on the efficiency of the municipalities. The demographic parameters tested for each local municipality were related to the size, age, and socio-economic level of the population. The most significant environmental variable affecting the efficiency of local municipalities in preventing road accidents is the population size of the local authority; the size has a negative effect on the efficiency. As far as we could determine, this is the first time that the DEA is used to measure the efficiency of local municipalities in improving traffic safety.