Implications for Delivery
To ensure robust analysis is undertaken, consideration must be given to how interventions will be evaluated early in the project development to ensure data is adequately captured to reflect circumstances prior to change as well as capturing data post intervention deployment. This includes the selection of comparator sites.
Resource allocation – time and budget - must be identified for the collection and analysis of data to deliver robust analysis that will verify the effectiveness of interventions.
Background
Evaluating road and roadside interventions is essential for several reasons:
Determining Effectiveness: Evaluation establishes whether interventions actually improve road safety by reducing crashes or injuries, instead of relying on assumptions or popularity.
Avoiding Harm: Without evaluation, well-intentioned interventions might unintentionally cause harm.
Economic Accountability: It allows funders and policymakers to justify investments, report on cost-effectiveness, and make evidence-informed spending decisions.
Knowledge Gain and Learning: Evaluation generates knowledge about what works and why, enabling practitioners to refine future interventions and avoid repeating ineffective or harmful approaches.
Policy and Practice Improvement: Evidence from evaluations guides the selection and scaling of interventions and supports the development of best practices and standards.
Supporting Funding and Wider Implementation: Demonstrating effectiveness with robust data is often required to secure continued or expanded funding and to justify broader adoption.
Transparency and Replication: Evaluation results allow others to understand and replicate successful interventions, helping improve road safety more widely.
What the Research Shows
There are a variety of ways of evaluating the impact interventions have had on road safety.
A familiar approach is by undertaking Before-and-After Studies that consider safety outcomes (e.g., crash frequency or severity) before and after an intervention. These studies should include comparison (control) sites that share similar characteristics but did not receive the intervention. The use of comparator sites helps to understand whether safety outcomes are genuinely as a result of the interventions or whether external factors, such as a general change in trends or unrelated changes have also had an influence.
The Empirical Bayes (EB) approach is considered the gold standard for evaluating road safety infrastructure interventions. It compares the number of crashes expected at a site (if no changes had been made) with the actual crash count following an intervention, using a reference group of similar sites. This approach specifically addresses statistical issues like regression to the mean and external confounding factors (Evaluating Interventions | Road Safety Manual - PIARC (World Road Association), n.d.).
Calculating the estimated First Year Rate of Return (FYRR) is often used to understand the expected economic effectiveness of interventions and where necessary is used as a means of prioritising multiple schemes for which there is a limited budget. However, FYRR calculations rely on accurate capital cost estimates and only calculates the return in the first year of implementation.
The FYRR method also usually yields high values for low-cost schemes but with relatively small crash savings, and for this reason it is less consistent with the Safe System approach (Prioritisation & Assessments | Road Safety Manual - PIARC (World Road Association), n.d.). To improve this, an assessment period aligned to the expected life of the intervention should be used, and maintenance costs over that period also included.
Crash risk mapping can also be used to monitor and visualise changes in crash occurrence and risk over time and may be used to identify high-risk locations before and after interventions have been deployed. However use of crash risk mapping tools should be accompanied by formal evaluation, including statistical analysis.
Sources
Evaluating Interventions | Road Safety Manual—PIARC (World Road Association). (n.d.). Retrieved 28 July 2025, from https://roadsafety.piarc.org/en/planning-design-operation/monitoring-and-evaluation/evaluating-interventions
Prioritisation & Assessments | Road Safety Manual—PIARC (World Road Association). (n.d.). Retrieved 29 July 2025, from https://roadsafety.piarc.org/en/planning-design-operation/intervention-selection/prioritisation-assessments