To illuminate or not to illuminate: Roadway lighting as it affects traffic safety at intersections

John D. Bullough, Eric T. Donnell, Mark S. Rea

Research output: Contribution to journalArticlepeer-review

104 Scopus citations


A two-pronged effort to quantify the impact of lighting on traffic safety is presented. In the statistical approach, the effects of lighting on crash frequency for different intersection types in Minnesota were assessed using count regression models. The models included many geometric and traffic control variables to estimate the association between lighting and nighttime and daytime crashes and the resulting night-to-day crash ratios. Overall, the presence of roadway intersection lighting was found to be associated with an approximately 12% lower night-to-day crash ratio than unlighted intersections. In the parallel analytical approach, visual performance analyses based on roadway intersection lighting practices in Minnesota were made for the same intersection types investigated in the statistical approach. The results of both approaches were convergent, suggesting that visual performance improvements from roadway lighting could serve as input for predicting improvements in crash frequency. A provisional transfer function allows transportation engineers to evaluate alternative lighting systems in the design phase so selections based on expected benefits and costs can be made.

Original languageEnglish (US)
Pages (from-to)65-77
Number of pages13
JournalAccident Analysis and Prevention
StatePublished - 2013

All Science Journal Classification (ASJC) codes

  • Public Health, Environmental and Occupational Health
  • Safety, Risk, Reliability and Quality
  • Law
  • Human Factors and Ergonomics


Dive into the research topics of 'To illuminate or not to illuminate: Roadway lighting as it affects traffic safety at intersections'. Together they form a unique fingerprint.

Cite this