Highlights This study estimates the effect of red light cameras on collisions in Los Angeles. Intersections with cameras are compared to nearby intersections without cameras. A correlated panel count data model is used to account for spatial correlation. Cameras cause red light running related collisions to decrease. Right angle, rear end, injury related crashes cause collisions to increase overall.
Abstract This study estimates the effect of red light cameras (henceforth cameras) on collisions under the Los Angeles Automated Photo Enforcement Program that ran from 2006 to 2011. To control for selection bias and unobservables, a data set is constructed such that intersections with cameras are compared to control groups of nearby intersections without cameras, matched on observable characteristics. To capture potential spillover effects of cameras, control groups at various distances from the intersections with cameras are considered. A Poisson panel data model with random coefficients is applied to these data and estimated using Bayesian methods. The program suffered from weaknesses in enforcement. The city’s courts did not uphold citations and this dampened the effect cameras had on drivers. These problems are accounted for in modeling. Controlling for these concerns, results indicate that the cameras decreased red light running related collisions, but increased right-angle and injury collisions, as well as collisions overall.
Lights, camera, legal action! The effectiveness of red light cameras on collisions in Los Angeles
Transportation Research Part A: Policy and Practice ; 69 ; 165-182
2014-08-22
18 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Lights, camera, legal action! The effectiveness of red light cameras on collisions in Los Angeles
Online Contents | 2014
|Lights, cameras — Machine vision!
Online Contents | 2014
The Big Picture - Lights! Camera! Action!
Online Contents | 2012
|Effectiveness of Red-Light Cameras
Online Contents | 2010
|Online Contents | 2001