Methodology for safety improvement programming using constrained network-level optimization

•Budget increase seems to converge in terms of cost-effectiveness and crashes saved.•Non-linear relationship between budget increases and benefits of crash reduction.•There is a ceiling to the effectiveness of engineering safety countermeasures.•Non-engineering countermeasures as safety education and enforcement are also vital.•Engineering investments in road safety is not likely to completely eradicate crash problem.

This paper develops and implements a decision support framework that prescribes and prioritizes cost-effective safety improvements at deserving locations under funding constraints, economic and comprehensive crash cost methods, at rural two-lane highway sections. The framework is demonstrated using Indiana’s 7700-mile rural two-lane network. For sections that are both deficient and hazardous under unconstrained budget and economic crash cost method, it was determined that $55 M is needed. Assuming a safety budget of $1 M/year over a 5-year period, it was found that 170–180 crashes (translating to $8–15 M) could be prevented. If the annual budget is increased to $2 M/year over a 5-year period, 244 crashes (translating to $12–26 M) can be prevented. Monetary amounts are in Year 2000 constant dollars. Overall, the results suggest that perpetual increases in highway safety improvements spending are not likely to be accompanied by a commensurate reduction of crashes at rural two-lane highways. In other words, there is a ceiling to the effectiveness of engineering safety countermeasures, and therefore non-engineering countermeasures such as safety education and enforcement must be sought to complement the engineering efforts at rural two-lane highways.