Citation

Stevens RS, Dean MD, Miller JS, Dougald LE. Case Stud. Transp. Policy 2020; 8(2): 648-657.

Copyright

(Copyright © 2020, World Conference on Transport Research Society, Publisher Elsevier Publishing)

DOI

10.1016/j.cstp.2020.02.006

PMID

unavailable

Abstract

Construction of an arterial roadway with a limited number of commercial access points may be viewed as the final outcome of a one-time planning process; however, the preservation of this corridor is a series of continuous management decisions made over its functional life. For this reason, at least 25 U.S. states maintain access management spacing standards, which set minimum spacing between access points in the corridor. Landowners may successfully petition for an exception to these standards, leading policymakers to ask whether such requests are associated with an increase in crashes. Because about 15% of all commercial entrance permits issued in Virginia from July 2011 to July 2016 entailed an access exception request, this state served as a useful case study for developing a program to monitor the crash impacts of these exception requests. A before-after analysis of crash frequencies at 64 exception sites, mostly built during 2011-2015, showed no significant difference (p = 0.63). A matched-pairs analysis, where the ratio of crashes at each exception site (after period/before period) is compared to the corresponding ratio at a comparison site (for the same time interval) showed no significant difference (p = 0.24). Execution of a negative binomial regression model with these 64 exception and 64 comparison sites also indicated no change in crash frequency associated with exception requests: a variable indicating presence of an exception site was significant both before (p = 0.04) and after (p = 0.02) the site was constructed. These results suggest Virginia's exception request program is not having an adverse safety impact. To be clear, these findings only apply to Virginia's particular exception program for the period of operation and do not indicate that elimination of access standards does not have a safety impact. (An area of further research is to determine whether the use of mitigation strategies is a contributing factor to this finding.) A surprising challenge was the difficulty in identifying ideal comparison sites, exacerbated by the diversity of exception requests that include sites with local roads, principal arterials, speed limits from 25 to 55 mph, small housing developments, and large industrial parks. The findings herein suggest that the matched-pairs technique can use appropriate but imperfect comparison sites fairly efficiently, allowing other states to establish monitoring programs in a cost-effective manner. For other states that wish to monitor the crash impacts of exception requests, the results of this analysis in Virginia suggest two lessons applicable elsewhere--each of which applies to the narrow problem of evaluating access exception requests but also to the larger challenge of monitoring the transportation system in an effective manner with imperfect data sets. First, multiple performance measures are beneficial given the specific challenge of finding suitable comparison sites. In this case study, one measure (differences in crash frequency) supports fast detection and the other measure (whether an exception site has a statistically significant impact on the natural logarithm of crash frequency) supports an assessment of impacts. Second, the use of the matched-pairs technique can mitigate challenges in developing robust comparison sites and thus encourage system monitoring. Execution of the tiered criteria for selecting these sites led to the identification of comparison sites where there was a strong degree of correlation with the exception sites, yet there were some numerical differences. These differences can be problematic with more detailed statistical modeling. For this reason, the latter technique can, support a two-stage monitoring effort: first, systematically identify areas of concern (using matched-pairs, which only requires that comparison and exception sites be correlated) and then, if areas of concern are identified, one may conduct a more detailed site-specific analysis.

Language: en

Keywords

Access control; Access management; Administrative procedures; Highway safety; System monitoring; Transportation planning