Installation of Pedestrian Countdown Signals Estimated to Reduce Crashes by 8 Percent, Resulting in an Average Expected Annual Benefit of $12,900.
Before-After Empirical Bayes Analysis of Crash Data in Charlotte, NC and Philadelphia, PA Estimated the Expected Change in Crash Frequency after Installation.
Made Public Date

Safety Evaluation of Pedestrian Countdown Signals

Summary Information

A pedestrian countdown signal (PCS) displays a real-time numerical countdown of how many seconds are left in the flashing “DON’T WALK” interval, to provide pedestrians with more information on remaining crossing time. Researchers investigated the effectiveness of PCS in reducing pedestrian crashes, and also studied impacts on other types of crashes. The research team obtained geometric, traffic, and crash data from signalized intersections in Charlotte, NC and Philadelphia, PA to evaluate the safety effect of PCS installations.

Researchers performed a before-after empirical Bayes (EB) analysis using data from 115 treated intersections in Charlotte, NC and 218 treated intersections in Philadelphia, PA. The evaluation also included 136 reference intersections in Charlotte, NC and 597 reference intersections in Philadelphia, PA. For Philadelphia, treatment and reference sites were only considered if they were within 1,500 feet of a point from which the city had collected a pedestrian count. For both locations, crashes were associated to the study sites via spatial proximity using a 200-foot radius. Crashes that occurred in parking lots or were related to driveways were also removed.


Safety Performance Functions (SPF) were developed in combination with the number of crashes at a site in the before period to estimate the number of crashes expected in each year of the before period. These estimates were then modified to account for the length of the after period, and differences in traffic volumes between the before and after periods. The resulting estimates reflect the expected number of crashes that would have occurred in the after period without the installation of the PCS at a site.

Crash Modification Factors (CMF) were computed as the ratio of the Observed Number of Crashes in the After Period (With PCS Treatment) to the EB Estimate of Expected Number of Crashes in the After Period (Without PCS Treatment). The research team also conducted economic analysis based on the estimated CMF values. The benefit-cost ratio was calculated as the ratio of the annual crash savings to the annualized treatment cost. The project team used average comprehensive crash costs based on crash severity and location type to estimate crash savings. Treatment cost data were not available from Philadelphia and Charlotte, but were estimated at $4,000 based on other cities that provided cost information. Due to the smaller sample of treatment intersections and lower rate of pedestrian-involved crashes in Charlotte, the researchers combined the data sets from the two cities.


  • The CMF for total crashes was estimated at 0.921, representing an 8 percent reduction, which was statistically significant at the 95 percent confidence level.
  • The CMF for rear-end crashes was estimated at 0.875, representing a 12 percent reduction, which was also statistically significant at the 95 percent confidence level.
  • The CMF for pedestrian-involved crashes was estimated at 0.912, representing a 9 percent reduction. This was statistically significant at the 90 percent confidence level, which was considered by the researchers as reasonable for rare crash types.
  • The expected benefit due to PCS installation was estimated as a reduction of 0.03 fatal / injury crashes and 0.37 Property Damage Only (PDO) crashes per intersection per year.
  • Using average comprehensive crash cost values based on crash severity and location type, the expected annual benefit due to the reduction in crashes after PCS installation was $12,900.
  • Comparing the annual estimated benefit to annualized treatment cost, the economic analysis revealed an average benefit-cost ratio of 23, with individual values ranging from 13 to 32.
Goal Areas
Results Type
Deployment Locations