Practices for Balancing Safety Investments in a Comprehensive Safety Program (2022)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

43   This chapter provides four case examples based on the literature review, survey responses, and follow-up interviews. The case examples highlight both basic and advanced practices for identifying, prioritizing, and evaluating HSIP projects. Basic practices employ data that are typically available to most agencies and methods that do not require calibration or develop- ment of statistical models. Advanced practices utilize multiple integrated data sets and predic- tive methods. The four case examples were selected to represent different state DOT practices for identifying, prioritizing, and evaluating HSIP projects. Table 32 provides a summary of the four case examples. The case examples highlight differences between spot and systemic projects as well as differences between state and local projects, with a particular focus on funding allocation between spot, systemic, and systematic; prioritization of spot and systemic projects; and evaluation of systemic projects. The chapter is structured by state, calling out specific examples within the state DOT that relate to the scope of this project. Specifically, each case example describes the general approach to safety management and includes sections that summarize project identification, project prioritization, funding allocation, project evaluation, highlights, and opportunities to overcome C H A P T E R 4 Case Examples HSIP Practice Maine North Carolina Oregon Pennsylvania Practices for identifying HSIP projects (basic or advanced) Advanced where possible and basic otherwise Basic warrants but helps focus on overexposed crash types Basic measures: crash frequency, crash rate, and crash severity Advanced where possible and basic otherwise Practices for prioritizing HSIP projects (single program or set- asides) Set-asides for spot, systemic, and systematic projects Set-asides based on quantitative approach (funding allocation model) Set-asides to five regions based on proportion of fatal/incapacitating injury (K/A) crashes. Even split between spot and systemic, but systemic is split by emphasis area based on proportion of K/A crashes. Set-asides based on quantitative and qualitative approaches Practices for evaluating HSIP projects (spot versus systemic) Limited to simple before–after studies Advanced evaluation practices Limited to simple before–after studies Advanced evaluation practices Table 32. Summary of case examples.

44 Practices for Balancing Safety Investments in a Comprehensive Safety Program challenges. The highlights are based on a combination of information obtained from the litera- ture review, survey, and interviews. The challenges and related opportunities are primarily based on the interviews. Maine The Maine HSIP involves the identification of safety issues, analysis of problems and counter- measures, and prioritization and scheduling of improvement projects. HSIP projects target locations or corridors with a substantive safety problem as indicated by location-specific data on severe crashes and where it is determined that the proposed project is expected to produce a measurable and significant reduction in the number and/or consequences of severe crashes (13). To achieve the maximum benefit, the program focuses on the cost-effective use of funds allocated for safety improvements. Maine ranks eligible projects in descending order and funds as many as possible until all funds are depleted. The HSIP is also holistic, considering all safety program needs identified in the SHSP. The following is a brief overview of the Maine HSIP, which includes spot, systemic, and systematic components that are similar on state and local roads. Project Identification Maine DOT identifies sites with promise using both spot and systemic approaches. Using the spot approach, Maine DOT identifies and reviews locations based on excess expected crashes (2) or the high-crash location (HCL) list. • The “excess” method focuses on intersections and ranks sites using simple ranking based on the total excess expected crash frequency with EB adjustment, excess expected crash costs, and excess expected-to-predicted crash ratio. From the ranked lists, analysts perform evalua- tions on the top 50 sites based on crash costs, top 50 sites based on expected-to-predicted crash costs ratio, top 20 sites based on rural crash costs, and top 20 sites based on rural expected- to-predicted crash costs ratio. The focus on the 20 top rural locations is intended to develop a more balanced listing of urban and rural sites because urban locations tend to dominate the systemwide ranked lists. • The HCL list focuses on intersections and roadway sections with a significant crash history that includes fatal or serious injury crashes. Maine uses the critical crash rate method to assess significance and develop an annual HCL list based on the most recent and complete 3-year period of crash reports. Maine defines HCLs as follows: – Crash rate is greater than expected for similar locations (same federal functional class and urban/rural setting), adjusted for traffic volume. – At least eight crashes in the 3-year reporting period Maine uses systemic and systematic approaches to complement the spot approach and address risk factors across the entire system. The systemic and systematic approaches use basic methods, such as crash summaries and crash trees, to identify risk factors and target SHSP emphasis areas. As an example of a systemic application, Maine deploys centerline rumble strips to address head- on collisions on priority, high-volume, high-speed corridors (i.e., rural corridors with posted speed limits greater than or equal to 45 mph and traffic volumes greater than 8,000 vehicles per day). The area type, traffic volume, and posted speed were selected as risk factors based on an analysis of head-on crashes. Maine also employs a hybrid approach that includes both spot and systemic components. As an example, Maine used the “excess” crash methodology with EB adjustment to identify intersections with potential for safety improvement, prioritized a list of similar locations (rural, three-leg, stop-controlled intersections with the highest excess crashes), and then programmed enhanced systemic signage packages at these locations.

Case Examples 45   Project Prioritization The purpose of project prioritization is to develop a portfolio of projects that maximize the expected safety benefit (i.e., lives saved and injuries prevented) for the amount of funds invested. Maine prioritizes spot projects using the benefit-cost ratio, which incorporates the estimated project cost, expected project effectiveness, and expected project service life. To quantify the expected safety benefits of proposed spot projects, Maine uses crash history, SPFs from the HSM, and CMFs. For systemic projects, Maine prioritizes locations based on the risk factors associated with the focus crash type and selected countermeasure. Beyond project costs and crash-based benefits, Maine also considers non-crash-based fac- tors in prioritizing HSIP projects. This includes a review by the Project Safety Review Team, comprising Maine DOT staff from the Bureau of Project Development, Bureau of Maintenance and Operation, Office of Safety, and Bureau of Planning. Members of the Project Safety Review Team evaluate the feasibility of projects in terms of scope, cost, constructability, eligibility, and alignment with SHSP. Other factors may include level of municipal support and/or legislative interest. The Project Safety Review Team then either moves the project forward, refines the scope, requests regional maintenance forces to make repairs, or, in some cases, discontinues the project. Funding Allocation Maine balances HSIP funds among spot, systemic, and systematic projects through set-asides. There is a set-aside for each program type, and projects compete within the designated program. First, the HSIP includes an annual set-aside for pavement marking retroreflectivity improve- ments. The remaining funds are split among systemic and spot improvement projects, targeting 50% for each program; however, the actual split can vary each year depending on the candidate projects. The funding split among spot and systemic projects is a qualitative approach based on professional judgment. Project Evaluation The Safety Office conducts post-implementation evaluations of safety projects and the HSIP as a whole. Maine uses the simple before–after methodology for project and program evalua- tions. The annual HSIP report documents the results of the analysis. Highlights Maine’s HSIP is structured and funded to make significant progress in reducing highway fatalities and serious injuries on all public roadways. The following are some highlights of the program. Project Identification • Maine uses a team approach to identifying problems and potential engineering solutions. Historically, the Bureau of Planning managed the HSIP, but this process changed with the formation of the Office of Safety. In addition, Maine established a Safety and Mobility Com- mittee, which serves as a community of practice for people to work together. • Maine uses more advanced methods where possible and relies on less data-intensive methods to supplement the advanced analysis, which provides coverage of the entire system for network screening. – The advanced network screening analysis utilizes excess expected crashes (the “excess” method) as the performance measure. This accounts for regression to the mean, changes in

46 Practices for Balancing Safety Investments in a Comprehensive Safety Program traffic volume, and the nonlinear relationship between crashes and traffic volume (2). The “excess” method also establishes a threshold to indicate sites experiencing more crashes than expected compared to sites with similar characteristics. – The HCL list is a basic analytic approach to network screening, allowing more complete coverage of the entire network, especially for those sites that lack data to support the “excess” method. • Maine is explicit in the HSIP documentation that “it is not enough to select a location from the ‘excess’ method or HCL list” and “a good location does not directly translate into a good project” (13). – The guidance emphasizes the need to properly diagnose the safety problem and better understand the underlying crash contributing factors before developing an appropriate project. – Although the spot projects implement countermeasures to address the type(s) of crash(es) at the particular location, the systemic approach deploys countermeasures based on high-risk roadway features correlated with specific severe crash types. Project Prioritization • Maine uses benefit-cost analysis to prioritize spot projects. This helps to maximize the expected safety benefit for the amount of funds invested. Maine adapted the HSM Part C spreadsheets to perform the quantitative benefit-cost analysis for the different facility types based on total crashes. • Maine uses CMFs from Part D of the HSM and/or the CMF Clearinghouse to estimate the crash reduction potential of countermeasures for spot projects. Project Implementation • Maine allows for some streamlining to implement systemic projects. For instance, Maine only requires preliminary engineering for certain systemic projects (e.g., those that impact ROW). • Maine has developed practical ways to implement systemic improvements. For example, Maine has procured hardware (e.g., flashing LED stop signs or speed feedback signs) and used main- tenance and operations forces to install the hardware. Maine has also used state DOT mainte- nance and operations forces to convert rural, four-leg, two-way stop-controlled intersections to all-way stop-controlled intersections. For local municipalities, Maine has a program to procure and distribute hardware (e.g., rectangular rapid-flashing beacons for midblock crosswalks) to the locals for installation and maintenance. Project Evaluation • Maine is considering the use of advanced analysis methods to estimate the effectiveness of common countermeasures (e.g., sinusoidal centerline rumble strips) and develop Maine- specific CMFs. Opportunities to Overcome Challenges Maine’s current HSIP documentation is relatively short and includes limited details for certain areas. This has raised some questions regarding the eligibility of certain projects for HSIP funding. To overcome these and other challenges, Maine is developing a more comprehensive HSIP manual to reflect current processes, formalize the business process, and explicitly discuss HSIP project eligibility. To guide the development of the new HSIP manual, Maine looked to peer state DOTs for ideas on what to include. Another challenge within the HSIP is related to project identification. Specifically, the “excess” method currently only applies to intersections, and some intersection types are not currently

Case Examples 47   available (e.g., roundabout). To overcome these challenges, Maine continues to enhance the project selection process. For instance, data collection efforts are under way to gather the infor- mation needed to employ the “excess” method on urban and rural roadway segments. Through an internal partnership with the Highway Management Department, Maine is collecting geo- metric data in both directions of travel to obtain more accurate cross-section data that will sup- port segment- and curve-related network screening. Finally, Maine described challenges related to the use of commercial off-the-shelf software to implement methods in support of the HSIP. Specific challenges include the time and cost associated with converting and maintaining data in the format required for the use of the software. Instead, Maine decided to develop a homegrown geographic information system (GIS) solution using internal staff. The GIS platform allows for integration of existing data sets, custom safety analysis, and visualization of results through the map viewer. The map viewer provides easy access to different layers (e.g., network screening and HCL lists) by analysts throughout the state DOT, not only in the Safety Office. This also provides consultants with direct access to data for analysis instead of requesting data through the state DOT. North Carolina The ultimate goal of the North Carolina HSIP is to reduce the number of traffic crashes, injuries, and fatalities by reducing the potential for and severity of these incidents on public roadways. North Carolina uses the HSIP to provide a continuous and systematic process for identifying, reviewing, and addressing traffic safety opportunities throughout the state. The program is structured in several phases that follow the traditional safety management process: problem iden- tification, diagnosis, countermeasure selection, economic analysis, project prioritization, and safety effectiveness evaluation. The following is a brief overview of the North Carolina HSIP, which includes spot, systemic, and systematic components that are similar on state and local roads. Note that there are relatively few local roads in North Carolina because the DOT is responsible for approximately 80% of public roadways in the state. Project Identification North Carolina uses safety warrants and crash pattern recognition to identify sites with poten- tial for safety improvement. Current warrants relate to intersections, sections, bicycles and pedestrians, and bridges. For locations that meet the warrant criteria and have more severe and correctable crash patterns, detailed crash analyses and collision diagrams help to diagnose crash contributing factors. Regional Traffic Engineering staff then perform engineering field investi- gations to further diagnose contributing factors and develop safety recommendations. Regional staff utilize benefit-cost analysis to support countermeasure decisions and recommendations before submitting proposed projects for funding consideration. North Carolina focuses systemic efforts within specific program emphasis areas (lane depar- ture, intersection, and pedestrian/bicycle). The systemic approach generally follows a similar process to spot projects where the state DOT analyzes crash and roadway data to identify sites with potential for safety improvement and then develops economically effective countermea- sures for broad application across the network. The following is a summary of this hybrid- systemic process: • Use safety warrants to identify sites with potential for safety improvement. • Identify cost-effective countermeasure(s) that address common crash contributing factors. • Review candidate locations to determine feasibility of specific countermeasure(s).

48 Practices for Balancing Safety Investments in a Comprehensive Safety Program For example, North Carolina identified more than 3,000 intersections statewide as potentially hazardous locations based on the intersection-related warrants and crash pattern recognition factors. In parallel, North Carolina determined that converting two-way stop-controlled inter- sections to all-way stop-controlled intersections is a cost-effective safety strategy based on post- implementation evaluations of past projects and supporting literature. Although it would not be feasible or appropriate to convert all 3,000 intersections to all-way stop-controlled, the state DOT recognized that there were likely hundreds of locations where this countermeasure could be effective. As such, the state DOT first narrowed the list of 3,000 potential locations to 1,200 potential locations based on number of lanes, traffic volume, and other basic factors to identify sites that could be appropriate for all-way stop-controlled. From there, analysts per- formed detailed investigations using online imagery to confirm site characteristics (e.g., number of lanes and traffic control) and determine if other measures had already been implemented (e.g., “vehicle entering when flashing” system, roundabout, or traffic signal). After further inves- tigation, 380 sites remained as viable locations that also met the safety warrant criteria and where the state DOT had not completed a recent project. This list served as the basis for targeted systemic intersection improvements (i.e., converting to all-way stop-controlled). As another example, North Carolina is identifying potential sections to address lane departure crashes based on the section-related warrants and crash pattern recognition factors. The state DOT selected rumble strips as the preferred countermeasure to address these crashes based on past project evaluations and the potential cost-effectiveness. To identify candidate locations, the state DOT is using risk factors, such as traffic volume, posted speed, and severe lane departure crash frequency and density. To help avoid noise-sensitive areas, the state DOT used a GIS building layer to identify areas with a high density of nearby buildings. As more of a systematic example, North Carolina is installing long-life pavement markings on secondary roads. Again, based on post-implementation evaluations of past projects, North Carolina estimated that new 4-in. long-life pavement markings could reduce lane departure crashes by up to 13%, and new 6-in. long-life pavement markings could reduce lane departures by up to 18%. Rather than screening for specific locations and performing a site-by-site benefit- cost analysis, the state DOT performed a planning-level economic analysis to estimate the average expected benefit and cost per mile. These analyses helped to demonstrate that long-life pave- ment markings are cost-effective at reducing lane departure crashes. Project Prioritization Based on the scale and scope of the recommended countermeasures, the investigations may trigger one of the following actions to implement the countermeasures: • Request division maintenance forces to make improvements or repairs. • Develop Spot Safety project. • Develop Hazard Elimination project. • Adjust current transportation improvement plans (TIPs). • Utilize other funding sources to initiate countermeasures. The Spot Safety Program develops smaller improvement projects (less than $400,000) to address safety, potential safety, and operational issues. North Carolina supports this program through state funds at approximately $12 million per state FY. Other monetary sources (e.g., Small Construction or Contingency funds) can assist in funding Spot Safety projects; however, the maximum allowable contribution of Spot Safety funds per project is $400,000. A Safety Oversight Committee (SOC) reviews and recommends Spot Safety projects to the Board of Transportation (BOT) for approval and funding using criteria such as the frequency of correct- able crashes, severity of crashes, delay, congestion, number of warrants met, effect on pedestri- ans and schools, division and region priorities, and public interest.

Case Examples 49   The Hazard Elimination Program (HEP) develops larger improvement projects ($400,000– $1 million) to address existing and potential safety issues. North Carolina supports this program through 90% federal funds and 10% state funds. Similar to the Spot Safety Program, the SOC reviews and recommends HEP projects to the BOT for approval and funding. The SOC priori- tizes these projects using a safety benefit-cost ratio, where the safety benefit is based on the esti- mated crash reduction using observed crash history and state-approved CMFs. Once approved and funded by the BOT, these projects become part of the Statewide Transportation Improve- ment Program (STIP). Historically, systemic projects have competed with spot projects on the basis of benefit-cost analysis. Because the state DOT uses safety warrants as one approach to identify potential loca- tions for systemic projects, these projects compare relatively well to spot projects with respect to benefit-cost ratio, even on the basis of observed crash history. Systematic projects such as long-life pavement markings do not compete on the basis of benefit-cost analysis. Instead, the state DOT promotes countermeasures that target focus crash types and provide a cost-effective return on investment. A project selection team reviews projects to ensure alignment with the goals of each specific systematic countermeasure. Once reviewed and approved by the team, the state DOT funds systematic projects as submitted up to the quarterly funding goal. North Carolina also has a separate prioritization process for projects that address vulnerable users. Evaluation and prioritization factors include total pedestrian crossing distance, vulnerable user exposure, and conflicting vehicle speeds. The state DOT uses this process to prioritize and select vulnerable user projects quarterly for HSIP funding. In addition to crash-based factors, North Carolina considers non-crash-based factors in prioritizing HSIP projects. These factors include SHSP priority, social equity, geographic equity, and project readiness. For bicycle and pedestrian projects, North Carolina also con- siders risk factors. Funding Allocation North Carolina is transitioning to a quantitative approach to determine the set-aside amount for each program. In 2020, North Carolina developed an HSIP Implementation Plan based on a review of crash data trends, SHSP emphasis areas, existing HSIPs and processes, and historical countermeasure selections. As part of this effort, the state DOT established funding allocation goals to align with the proportion of fatal and serious injury crashes. This resulted in funding goals for roadway departure, intersection, and pedestrian and bike programs, with suballoca- tions for systemic and responsive (spot) projects, as shown in Figure 19. Within each pro- gram area, projects compete for funding using the data-driven and objective selection processes described previously. Although there are specific funding goals (e.g., 50% for roadway depar- ture), there is also a range (e.g., 40%–60% for roadway departure) to allow for more flex- ibility in project prioritization based on the estimated cost-effectiveness of proposed projects. In addition, projects are not necessarily mutually exclusive, and the state DOT encourages applicants to address multiple concerns in the same project. For instance, a project might be identified and justified based on intersection crash warrants and patterns, but the project could also address pedestrian safety issues. Project Evaluation North Carolina has a robust evaluation component to the HSIP (4). The Traffic Safety Unit conducts safety evaluations for all completed safety projects, common countermeasures, and the safety program as a whole. To evaluate individual project effectiveness, North Carolina uses the simple before–after method and collision diagrams with a focus on the change in target crashes.

50 Practices for Balancing Safety Investments in a Comprehensive Safety Program Focusing on target crashes helps to determine if the project achieved the initial objective (i.e., to address a specific crash type or crash contributing factor). If a project does not address the target crashes, it may be appropriate to consider alternative or supplemental countermeasures. If the project achieved a reduction in target crashes but total crashes increased or remained the same, it may be appropriate to consider alternative measures to address the other crash types. As the state DOT completes a particular type of project at multiple locations, there is an opportunity to conduct large-scale studies using data from locations across the state. For these countermeasure-level evaluations, including the application of a countermeasure at multiple sites, North Carolina uses more rigorous before–after evaluations (e.g., comparison group and EB) to estimate safety effectiveness. North Carolina performs the countermeasure-level evalua- tions and shares the results with project and program managers, including the regional offices, to provide more objective and definite information regarding the expected safety benefits of similar future projects. Although the state DOT is in the early stages of systemic project evalu- ations, the plan is to handle these types of projects similar to spot projects. Specifically, the more rigorous before–after methods are useful for evaluating systemic applications to help account for regression to the mean and sites with few or no crashes before implementation (although this is not yet a concern in North Carolina, as the project identification process is based on safety warrants). For program evaluation, North Carolina evaluates individual HSIP projects and then aggre- gates the results to estimate the overall benefit-cost ratio of the HSIP as a whole. By insti- tutionalizing HSIP evaluation, safety staff have been successful in retaining safety project funding and even obtaining increases in funding. The safety evaluations have also impacted other areas of DOT operations, leading to increased confidence in investments. Highlights North Carolina has integrated several efficiencies in the HSIP project development process and implemented several noteworthy practices related to project and program evaluation. The following are some highlights of the program. Project Identification • The state DOT is not yet ready to treat locations with no crash history because of the preva- lence of locations with a substantial crash history. As such, North Carolina uses a hybrid approach to identify systemic project locations, starting with the safety warrants used in the Roadway Departure 50% (40%–60%) Systemic 80% Responsive 20% Intersection 35% (30%–40%) Systemic 40% Responsive 60% Pedestrian & Bike 15% (10%–15%) Systemic 40% Soft Target Responsive 60% Soft Target Figure 19. 2020 North Carolina HSIP Implementation Plan funding goals.

Case Examples 51   spot approach and then identifying sites with common characteristics for similar treatment. This hybrid approach helps to identify projects with a demonstrated safety issue that can be treated with systemic countermeasures that can also compete with spot projects on the basis of benefit-cost ratio. • North Carolina uses the results of past project evaluations to identify cost-effective counter- measures that can be implemented systemically to address common site characteristics and crash contributing factors. Project Prioritization • North Carolina uses a quantitative approach to prioritize projects and identify those with the greatest potential to cost-effectively improve safety. To this point, nearly all systemic projects compete with spot projects on the basis of benefit-cost analysis. Although system- atic projects, such as long-life pavement markings, do not compete on the basis of benefit- cost analysis, the state DOT uses planning-level economic analysis to estimate the average expected benefit and cost per mile as justification for these investments. • Knowing that it can be challenging to justify pedestrian and bicycle projects on the basis of crash history, North Carolina developed a separate prioritization process for projects that address vulnerable users. This process uses risk factors to prioritize and select projects for HSIP funding. Project Implementation • North Carolina allows for some streamlining to implement HSIP projects. Specifically, pre- liminary engineering is not required for countermeasures with standard drawings. Instead, North Carolina only requires preliminary engineering for certain projects (e.g., those that impact ROW). • North Carolina utilizes and considers the following methods to improve project delivery times and reduce the overall cost of delivering HSIP projects: – Combine multiple safety improvements in a single contract. – Use design–build delivery mechanisms to fast-track project delivery for projects with a well-defined scope. – Use on-call contractors to facilitate immediate delivery of identified projects. Project Evaluation • North Carolina’s HSIP emphasizes the importance of focusing on the change in target crashes rather than total crashes when evaluating projects. This helps in situations where total crashes increase or remain unchanged and target crashes decrease. • North Carolina publishes individual project evaluations on the web, including a description of the project location, project background, summary of improvements, and results and discus- sion of simple before–after analysis for both total and target crashes. The web-based project evaluation documents allow others within the state to access the results, inform future deci- sions, and demonstrate the benefits of past projects when justifying proposed projects. • North Carolina uses relatively simple methods for project-level evaluations and applies more advanced methods for countermeasure-level evaluations. • North Carolina has developed a spreadsheet-based approach to facilitate the more rigorous EB analysis method. The spreadsheets are binned by countermeasure, and as new sites are added to the spreadsheet, the CMFs are updated automatically. Opportunities to Overcome Challenges One challenge in North Carolina is related to data. Specifically, North Carolina does not have an intersection inventory, which makes it challenging to perform systemwide systemic analyses

52 Practices for Balancing Safety Investments in a Comprehensive Safety Program and identify risk factors. To overcome this challenge, North Carolina uses a hybrid approach, employing safety warrants to first identify sites with high potential for safety improvement and then reviewing these sites for potential application of systemic countermeasures. Although this still involves a manual review of numerous locations, the initial screening helps to make the task more manageable. Another challenge is related to the prioritization and implementation of systemic projects. Part of the prioritization score is based on local support, and it can be challenging to obtain the necessary resolution from a county commissioner to support several individual projects. To overcome this challenge, the state DOT may bundle several similar projects into one project. This helps to facilitate the process of garnering local support and helps with project manage- ment. North Carolina does not bundle systemic projects to improve the benefit-cost ratio for specific locations (e.g., sites with low or no crash history); all sites included in the bundled project could stand alone on the basis of cost-effectiveness. Although North Carolina is preparing for systemic project evaluations, including the poten- tial use of more rigorous before–after methods, there are some challenges in preparing for these evaluations. For instance, the state DOT performs detailed crash analyses and comparisons for each spot improvement project; however, this may not be feasible for systemic project evalu- ations because there are many more locations. A related challenge is identifying the spe- cific project locations for systematic projects (e.g., long-life pavement markings and changes in posted speed) and systemic projects that are bundled. For example, the application for a bundled systemic project may indicate “multiple locations” without listing the specific sites to be improved. The state DOT is overcoming this challenge through follow-ups with the regions to identify the specific list of sites where these types of improvements are implemented. North Carolina’s project and countermeasure evaluations have been instrumental in com- municating the value of certain countermeasures. For instance, it can be challenging to convince some people (internal and external) of the potential benefits of systemic countermeasures that are new to their jurisdiction (e.g., converting two-way to all-way stop-controlled). Based on results of HSIP evaluations of past projects in other jurisdictions, the state DOT is able to help convince people of the feasibility and expected effectiveness of these projects. North Carolina also develops and delivers webinars and materials to help demonstrate the value of these projects to internal stakeholders. This combination of extensive evaluations and effective communica- tion of the results has helped to change the mindset on what is feasible and effective. The next step is to evaluate the effectiveness of countermeasures as prevalence increases. This will help to address the question of whether or not a countermeasure remains effective at sites with fewer crashes (i.e., those lower on the list of sites with potential for improvement). Oregon The mission of Oregon’s Highway Safety Program, which comprises mainly HSIP projects, is to carry out highway safety improvement projects to achieve a significant reduction in traf- fic fatalities and serious injuries (14). Oregon documents the program philosophy and project selection process in a Highway Safety Program Guide. Oregon requires the use of these guide- lines for all highway safety infrastructure improvement projects programmed in the STIP. The general program guidelines are as follows: • All projects shall address specific safety problems that contribute to fatal and serious injury crashes. • All projects shall use only countermeasures from the Oregon DOT-approved countermeasure list: https://www.oregon.gov/odot/Engineering/Docs_TrafficEng/ARTS_Countermeasure- Search.zip

Case Examples 53   • Only the most recent available 5 years of Oregon DOT-reported crashes shall be used for crash analysis. • Projects shall be prioritized based on Oregon DOT-approved prioritization method, such as benefit-cost ratio. • Oregon DOT regions will be responsible for developing and delivering projects. Oregon’s safety program, the All Roads Transportation Safety (ARTS) Program, addresses safety on all public roads. The Traffic-Roadway Section is responsible for developing program guidance, developing tools for identifying and analyzing highway safety problems, and preparing annual HSIP reports. The regions are responsible for diagnosing safety issues, selecting projects for the STIP, managing safety funds allocated to their region, and gathering information to support the annual HSIP reporting requirements. The ARTS Program employs a data-driven approach with a focus on addressing fatal and serious injury crashes to maximize the benefits and cost-effectiveness of the HSIP. The data- driven approach uses crash data, risk factors, and other data-supported methods to identify the best possible locations to achieve the greatest benefits in crash reduction. As documented in Oregon’s Highway Safety Program Guide, HSIP funds should target locations or corridors where a known problem exists, as indicated by location-specific data on fatalities and serious injuries, and/or where it is determined that the specific project is expected to produce a measur- able and significant reduction in such fatalities or serious injuries. The following is a brief overview of the Oregon HSIP, which includes spot and systemic components that are similar on state and local roads. Project Identification For spot safety projects, Oregon uses a variety of basic crash-based approaches to identify locations with potential for safety improvement. Although Oregon is pilot testing more reliable statistical methods for network screening from the HSM, the current method—the Safety Priority Index System (SPIS)—combines crash frequency, crash rate, and crash severity to identify poten- tial sites. For the ARTS Program, a spot location must have at least one fatal or serious injury crash within the last 5 years. Potential countermeasures are identified based on detailed site analysis (e.g., road safety audits) and a preapproved list of countermeasures. Spot projects focus primarily on intersections or short segments. An Intersection Safety Action Plan helps to guide intersection projects. In addition, Oregon’s crash reduction factor (CRF) list includes counter- measures specific to intersections (15). For systemic safety projects, Oregon uses a dual approach to identify potential locations. First, regions can use the SPIS list to identify locations with common characteristics and then target those locations with a common countermeasure. Another option is to use one of three focused implementation plans that identify potential locations where investments may yield good returns in terms of reducing fatal and serious injury crashes. The three plans relate to roadway depar- ture, intersections, and pedestrian/bicycle crashes. Although some of the emphasis areas lend themselves particularly well to analyzing crash data (such as roadway departure and inter- sections), others (such as pedestrian and bicycle) do not. Those that might not have sufficient crash frequency to reliably identify unusual patterns or high-risk area may use surrogates such as risk factors. Oregon provides a list of proven low-cost countermeasures that can be widely imple- mented and allows the application of these countermeasures at locations where there is evidence that it would be appropriate (i.e., related crash history or presence of related risk factors). The systemic program is subdivided into three subprograms: roadway departures, intersections, and pedestrians/bicycles. The following is a brief description of these subprograms: • The State Roadway Departure Plan identifies crash types with specific countermeasures that are designed to address these crashes. The state DOT selects clusters of locations that have

54 Practices for Balancing Safety Investments in a Comprehensive Safety Program targeted crashes at or above a designated threshold level. Countermeasures include curve treatments, rumble strips, delineation, high-friction surface treatment, tree management, and alcohol/speed enforcement. • The State Intersection Plan incorporates a comprehensive approach that includes a spot approach (described previously), a systemic approach (application of large numbers of cost-effective, low-cost countermeasures), and a corridor approach (application of low-cost infrastructure improvements coupled with targeted education and enforcement initiatives on corridors). Countermeasures include signal upgrades, flashing yellow arrow, left-turn protection, sight distance improvements, high-friction surface treatment, enhanced signs and pavement markings, and traffic calming. • The State Bicycle and Pedestrian Safety Implementation Plan identifies specific risk factors and targeted countermeasures. The plan follows the steps outlined in NCHRP Research Report 893: Systemic Pedestrian Safety Analysis (16). • Pedestrian risk factors include the following: – Roadway data: functional class (principal arterial), number of lanes, high-access density, presence of sidewalk, and posted speed – Context: zoning (mixed use or other general), proximity to schools, and transit stops – Demographics: high population over the age of 64 – Other risk factors not used in screening due to lack of availability, including high turning volumes at intersections, left-turn phasing, lighting, propensity for midblock crossings, and exposure Bicycle risk factors include the following: – Roadway data: functional class (principal/minor arterial), number of lanes, high-access density, presence of sidewalk, posted speed (≥ 35 mph), and presence of bike lanes – Context: zoning (mixed use or other general), proximity to schools, and transit stops – Demographics: high population over the age of 64 – Other risk factors not used in screening due to lack of availability, including scenic bike- ways, time of day, and lighting To support countermeasure selection, Oregon compiled a list of countermeasures and asso- ciated CRFs from the HSM, CMF Clearinghouse, and FHWA Desktop Reference for Crash Reduction Factors (15). Countermeasures are listed as either “spot” or “systemic” countermea- sures. Any countermeasures listed in the Oregon DOT CRF list can be used for spot projects; however, only countermeasures listed as “systemic” may be used for systemic projects. Oregon updates the CRF list periodically as new countermeasures or better studies on existing counter- measures become available. Oregon’s HSIP also includes a documented systematic safety component that addresses selected safety countermeasures in design policies. This helps to implement proven safety counter- measures at all feasible locations. The following are examples: • Rumble strips: Longitudinal rumble strips shall be installed on STIP projects (17). • Safety Edge: On paving projects with shoulder widths of 6 ft or less and new pavement thickness of 2 in. or more, Safety Edge will be included in the project and shown on the typical sections (18). Project Prioritization Oregon uses slightly different processes for selecting spot and systemic safety projects for the ARTS Program. The following is an overview of the spot portion of the program: • Oregon DOT prepares region-specific draft 300% spot project lists (i.e., list that represents enough projects to spend three times the available funding) using the typical six-step roadway safety management process.

Case Examples 55   • Regions share the 300% draft lists with the local agencies and seek local agency input. • Local agencies submit proposals for additional projects for inclusion in the draft list. • Regions refine the draft 300% list based on local agency inputs and proposals. • Regions finalize draft 150% lists for field scoping. • Regions perform field scoping and prepare final 100% lists. • Regions submit project prospectuses and request approval from the Oregon Transportation Commission to include in the STIP. The systemic component of the ARTS Program is an application-based process in which Oregon DOT regions and local agencies within the region may submit applications for systemic improvements under the three emphasis areas: roadway departure, intersection, and pedestrian/ bicycle. Systemic projects typically consist of more than one location (segments or intersections) where the same countermeasure is proposed. Oregon does not require locations to have any observed fatal or serious injury crashes in the project corridor for systemic projects. The following is an overview of the systemic portion of the program: • Regions and local agencies submit applications for systemic projects. • Regions review applications for program purpose and correctness of the applications. • Regions develop draft 150% lists based on the applications received. • Regions finalize draft 150% lists for field scoping. • Regions perform field scoping and prepare final 100% lists. • Regions submit project prospectuses and request approval from the Oregon Transportation Commission to include in the STIP. Oregon uses benefit-cost analysis as the primary method for project prioritization except for bicycle and pedestrian projects. To prioritize bicycle and pedestrian projects, the state DOT uses cost-effectiveness analysis. The following are a few key points of the benefit-cost analysis process: • Requires benefit-cost ratio of 1.0 or greater. • Uses most recent 5 years of crash data. • Benefits represent economic value of expected reduction in target crashes over service life of countermeasure(s). • Project costs should include all costs associated with installing proposed countermeasure(s) (e.g., environmental mitigation or additional improvements to meet ADA requirements). • Project costs should include annual maintenance and operation cost as appropriate. For cost-effectiveness analysis, Oregon uses the Cost-Effectiveness Index (CEI) to prioritize projects. The CEI represents the estimated cost to reduce one crash. The lower the CEI value of a project, the higher it will rank in the prioritized list. To calculate CEI, Oregon predicts pedes- trian and bicycle crashes using the HSM Part C Predicted Method. Comparing the predicted crashes with the observed crashes, Oregon uses the higher of the two in the CEI analysis. The expected crash reduction is based on the CRFs for proposed countermeasures, and CEI is calculated by dividing the expected crash reduction by the estimated total project cost. Once the final list of projects is approved and included in the STIP, regions work with the appropriate local agencies to determine the delivery methods, delivering agency, and timelines. Oregon does not require preliminary engineering for countermeasures with standard draw- ings, such as signs, pavement markings, rectangular rapid-flashing beacons, and rumble strips (19). The ARTS Program typically results in stand-alone projects; however, these projects may be combined as appropriate. For projects involving local agencies, regions work with these to develop an intergovernmental agreement. The delivering agency is then accountable for timely and fiscally responsible delivery.

56 Practices for Balancing Safety Investments in a Comprehensive Safety Program Funding Allocation The ARTS Program allocates funds to the five Oregon DOT regions based on the proportion of fatal and serious injury crashes within the past 5 years in each region. For a given region, total funding is divided equally between the spot and systemic components. For the systemic component, it is recommended that regions split the available funding between the emphasis areas identified in the Oregon TSAP based on the proportion of fatal and serious injury crashes in those areas within the past 5 years. Figure 20 shows an example distribution of funds among regions and emphasis areas. Project Evaluation Historically, Oregon has performed limited evaluations. The focus was on simple before–after studies to meet the HSIP annual reporting requirements. Oregon is developing a process to evaluate safety effectiveness of the projects built under the ARTS Program. Highlights Oregon updated the Highway Safety Program Guide in 2021 and included several noteworthy practices. The following are some highlights of the program. Project Identification • Local agencies have the opportunity to suggest alternate countermeasures for selected proj- ects or to submit new potential projects that are not included in the 300% draft list; however, the local agencies are required to use the same methodologies used to develop the initial list. • Oregon conducts field scoping for all proposed projects using a multidisciplinary assessment to verify the solution. A multidisciplinary team ensures the countermeasure is appropriate to mitigate fatal and serious injury crashes. The team also scopes each project to verify and revise estimated costs and identify possible impediments to implementation. Project Prioritization • By starting the spot project prioritization process with enough projects to spend three times the available funding (i.e., 300% lists), there is a better opportunity to find the optimal com- bination of projects to maximize the cost-effectiveness of the program. Source: Figure 2-2 from Oregon Department of Transportation (14). Note: Key corresponds left to right to bars in figure. 31% 51% 65% 67% 74% 50%48% 35% 25% 26% 19% 36% 21% 14% 10% 7% 7% 14% 0% 10% 20% 30% 40% 50% 60% 70% 80% Region 1 Region 2 Region 3 Region 4 Region 5 Statewide Pe rc en t F un di ng Roadway Departure Intersection Pedestrian/Bicycle Figure 20. Approximate funding splits between systemic emphasis areas.

Case Examples 57   • Oregon developed a Benet/Cost Analysis Worksheet to calculate benet-cost ratios for the ARTS Program (see Figure 21). e user selects up to four countermeasures from a predened drop-down list, and the spreadsheet automatically populates the target crash type, target crash severity, CRFs by crash type or severity (as applicable), and service life. e spreadsheet combines the CRFs by converting to CMFs and using the multiplicative method to estimate a composite CRF. Based on user-entered data for observed crashes, the spreadsheet applies the composite CRF to estimate the number of preventable crashes, converts the crash benet to an economic value, and computes the present value benet over the service life. Based on user-entered data for construction and maintenance costs, the spreadsheet computes the total present value cost and benet-cost ratio over the service life. • If an application includes multiple countermeasures, Oregon requires the primary counter- measure (i.e., the countermeasure matching the application type) to account for at least half of the benet. is allows projects to address multiple issues (e.g., adding pedestrian improve- ments to an intersection project) while focusing on the target crash type and maintaining a relatively high benet-cost ratio. Figure 21. Sample of Oregon benet-cost worksheet.

58 Practices for Balancing Safety Investments in a Comprehensive Safety Program • Based on the HSM Part C spreadsheets, Oregon developed a Microsoft Excel workbook titled Cost-Effectiveness Analysis for Pedestrian/Bike Systemic Improvements. The user enters appli- cable roadway, traffic, and observed pedestrian and bicycle crash data on separate tabs for segments and intersections as appropriate. The user also specifies the proposed pedestrian and bicycle countermeasures with the corresponding CRF from the Oregon CRF list. The spreadsheet automatically computes the predicted pedestrian and bicycle crashes per year and composite CRFs for use in the analysis. Based on user-entered data for total project cost, the spreadsheet computes the CEI. • Oregon developed crash costs that represent comprehensive economic values and differ by area type and facility type (see Table 33). Project Implementation • Although Oregon does not allow for any shortcuts to implement spot or systemic HSIP projects, preliminary engineering is not required for countermeasures with standard drawings, such as signs, pavement markings, rectangular rapid-flashing beacons, and rumble strips. Project Evaluation • Although Oregon uses a simple before–after analysis for project evaluation, there is con- tinuous effort to evaluate and improve the program. After each round of funding, Oregon distributes a survey to determine what is working and what is not. The survey includes a sample of those that participated and sample of some that did not participate in the most recent round of HSIP funding. This type of feedback has been instrumental in identifying opportunities to continually enhance the program. Opportunities to Overcome Challenges Although Oregon has made great strides to update the HSIP process through the ARTS Pro- gram, there are some specific challenges with respect to project identification, prioritization, and evaluation. Timely data are one of the primary challenges related to project identification. Specifically, Oregon’s most current crash data are up to a couple of years old, and the monthly fatal crash summaries do not provide the level of detail needed for this type of analysis. In addi- tion, the timeline for project development and the TIP process can take several years. With the combination of these factors, it can be several years after the issue arises by the time projects are constructed to mitigate the issue. To address this challenge, Oregon is actively working to Crash Severity Highway Type Urban Rural Property damage only (PDO) All facilities $21,800 $21,800 Moderate (B) and minor (C) injury Interstate $77,800 $89,200 Moderate (B) and minor (C) injury Other state highway $80,800 $91,900 Moderate (B) and minor (C) injury Off system $81,300 $93,200 Fatal (K) and serious (A) injury Interstate $1,530,000 $2,260,000 Fatal (K) and serious (A) injury Other state highway $1,490,000 $2,140,000 Fatal (K) and serious (A) injury Off system $1,110,000 $1,940,000 Table 33. Example of comprehensive crash costs by facility type.

Case Examples 59   identify other sources of data that could be used to identify priority location and other funding opportunities to help address these locations. One initiative is to fund safety projects from smaller pots of money that do not require the same time frame as the TIP. This has helped to improve timeliness, but this type of funding does not stretch very far. Oregon uses the CEI to help prioritize pedestrian and bicycle projects, as described previ- ously; however, there are some challenges related to the use of the associated spreadsheet. Specifically, some have found the spreadsheet cumbersome, and Oregon is working to simplify it. In addition, the spreadsheet incorporates a predictive method but uses the greater of two values: (a) predicted crashes or (b) observed crashes. Because the predicted crashes are generally on the magnitude of a fraction of a crash, even one observed crash will tend to outweigh the prediction in this approach. Oregon is working to overcome this challenge and consider both predicted and observed crashes while considering project prioritization through an equity lens. Another challenge is related to the funding allocation for local roads and the management of federal funds. The ARTS Program allocates funds to the five Oregon DOT regions based on the propor- tion of fatal and serious injury crashes, which is further divided for spot and systemic projects. To improve the efficiency of managing federal funds, Oregon implemented a funding exchange program. This was well received by the local agencies and appeared to be successful until budget constraints impeded the funding exchange. As another measure to focus on projects that could be more efficient for management of federal funds, Oregon implemented a minimum project size of $500,000; however, this created additional challenges, as evidenced by fewer local project applications. One such challenge is the local matching requirement because it can be difficult for smaller agencies to come up with even a percent match given the minimum project size of $500,000. Oregon continues to innovate and identify opportunities to implement HSIP projects on both state and local roads. To maintain a competitive process while allocating funds for local projects, Oregon recommends that regions further split the available funding approximately equally between state and local roads (51% state/49% local) based on a statewide analysis of the proportion of fatal and serious injury crashes on these roads. Oregon will evaluate this funding split after each round to see if it works or if further adjustments are needed. Finally, there are challenges related to project evaluation. As previously noted, Oregon uses the simple before–after method to evaluate projects but is looking to develop a tool or process for enhancing project evaluations. As part of this, the state DOT will be exploring how best to evaluate systemic projects. One concern from stakeholders has been the applicability of CRFs from other state DOTs. Although Oregon’s CRF list supports a quantitative approach to project prioritization, some have challenged that CRFs from other state DOTs may not be applicable to Oregon. Enhancing the project evaluation process and developing Oregon-specific CRFs will help convince people that the countermeasures in question can be effective in Oregon as well. Pennsylvania Pennsylvania’s Highway Safety Program goal is to reduce average fatalities and serious injuries to support the national effort of ending fatalities on our nation’s roads within the next 30 years. The HSIP is one component of Pennsylvania’s Highway Safety Program, which is focused on implementing infrastructure improvements at sites with potential for reducing average crash frequency. HSIP projects on state roads fall into three approaches: • Traditional (spot) approach: Identify specific at-risk locations and determine cost-effective countermeasures for each location.

60 Practices for Balancing Safety Investments in a Comprehensive Safety Program • Systemic approach: Identify promising cost-effective countermeasures and then identify sets of locations where it is cost-effective to apply the countermeasure. • Corridor approach: Identify sections of highway that have significant numbers of severe crashes, of either all or specific types, and apply a coordinated set of engineering, enforce- ment, and education initiatives to address or mitigate the problem. The program is structured in three components: planning, implementation, and evaluation. The planning component is further disaggregated into problem identification, countermeasure identification, project prioritization, and categorization. The following is a brief overview of the Pennsylvania HSIP, which includes spot and systemic components that are different on state and local roads. Project Identification As part of the problem identification process, the Pennsylvania DOT, MPOs, and regional planning organizations (RPOs) identify a list of candidate sites using crash, traffic volume, and roadway data as well as local input. In general, analysts use the SHSP emphasis areas to guide HSIP problem identification and then look to the SHSP strategies for countermeasure ideas. For spot projects, Pennsylvania employs more rigorous network screening performance measures where possible (i.e., excess expected crashes) and allows for the use of traditional crash cluster or crash rate measures to justify safety needs where data are limited. The more rigorous methods typically apply to state roads, and the traditional methods typically apply to local roads. Following network screening, the sites are analyzed to identify applicable countermeasures and estimate project costs. For spot and systemic HSIP projects on the state system, analysts may identify applicable countermeasures based on a combination of detailed site analysis, pres- ence of risk factors, and a preapproved list of countermeasures. On the local system, spot and systemic projects are limited to a preapproved list of countermeasures. Pennsylvania administers systemic safety projects under the HSIP, including cable median barriers, high-friction surface treatment, wrong-way exit ramp countermeasures, elimination of substandard cable guardrail, and rumble strips. To guide systemic projects, Pennsylvania developed safety plans, including the following: • Roadway Departure Implementation Plan (RDIP) • Intersection Safety Implementation Plan (ISIP) • Speed Management Action Plan (SMAP) As an example, the SMAP guides systemic improvements by first identifying the speed-related focus area (i.e., intersections), then identifying risk factors (i.e., signalized intersections with posted speeds of 40 to 45 mph on the approaches), and finally implementing targeted treatments to improve those types of locations. The improvements are not a blanket approach (systematic), but systemic, because the treatments are tailored to the specific locations. Other examples of systemic efforts are based on systemic safety crash lists, such as the cross median crash and wrong-way crash priority lists, to identify opportunities for systemic safety improvements. For both spot and systemic HSIP projects on the local system, the MPO/RPO or engineering district submits municipal project applications. Systematic-type improvements, such as Safety Edge for pavements and proper end treatments for guide rails, are part of construction standards and not a specific component of the HSIP. Project Prioritization Pennsylvania establishes the maximum federal HSIP funding that a planning region or dis- trict can receive in each HSIP call for projects. If a planning region or district submits multiple

Case Examples 61   applications with a combined federal funding request that exceeds the established regional allo- cation, the region’s applications with the highest return on safety are included in the planning process until reaching the region’s maximum federal HSIP budget. Pennsylvania distributes HSIP funding in the following three major categories: • Provide $500,000 base HSIP funding for each of the 23 RPOs and one independent county. • Reserve approximately $35  million for HSIP set-aside projects (various statewide safety initiatives). • Allocate remaining HSIP funding to planning organizations based on the proportion of fatal and serious injury crashes in the region and the total number of reportable crashes in the region. Pennsylvania requires economic appraisals for spot projects and sets a minimum benefit- cost ratio of 1.0. Economic appraisals are based on the most current 5 years of reportable crash data and CMFs from the CMF Clearinghouse, the HSM, or the Pennsylvania CMF Guide. Pennsylvania publishes annual crash costs and provides other guidance and resources for per- forming economic analysis, including FHWA’s Highway Safety Benefit-Cost Analysis Guide (20) and the HSM, Volume 1, Chapter 7 (2). Pennsylvania also allows the use of the CEI (cost per crash severity). Pennsylvania has an extensive list of factors for evaluating and prioritizing proposed projects. The following is a list of common considerations, including both crash-based and non-crash-based factors for prioritizing all HSIP projects (21): • Does the project scope fall within an SHSP focus area? • Does it meet the eligibility criteria listed in the FAST Act? • Is it a non-infrastructure request (e.g., road safety audit, data collection, or planning activities)? • What are the results of HSM predictive analysis and network screening? • What are the results of benefit-cost analysis using CMFs? – Must include a crash resume and summary reports. – Prefer itemized cost breakdowns for countermeasures. • What is the systemic safety improvement value (benefit-cost analysis may not apply)? – What is the common CMF for this networkwide systemic improvement? • Is a clear project timeline provided? – All projects must show an estimated construction let date and open-to-traffic date. • Are there any project complexities that could disrupt project delivery (e.g., complex ROW, utilities coordination, environmental concerns)? • Are HSIP funds available for distribution? • Could the proposed project be completed using other funds? In addition to the factors for general HSIP projects, Pennsylvania considers the following factors in the selection of HSIP set-aside project applications: • Timeliness of the application submission • Highest benefit-cost ratios • HSM analysis results and crash history • Advancement of a previously approved safety project to earlier construction with the addi- tional HSIP funding • Timeliness of project delivery (within the proposed time frame) • Open-to-traffic date • Availability of set-aside HSIP funds • Regional classification of the project as a local or state road • Regulation or special requirement by FHWA

62 Practices for Balancing Safety Investments in a Comprehensive Safety Program Project Evaluation Pennsylvania has a robust evaluation component to the HSIP (21). The HSIP annual report includes the results of simple before–after crash analysis to evaluate HSIP project effective- ness with respect to reducing crash frequency and severity. The annual report also reports on the implementation of systemic safety improvements, such as rumble strips, cable median barrier, and high-friction surface treatments. The Highway Safety and Traffic Operations Divi- sion evaluates systemic safety improvements as data allow. For more common countermeasures and strategies that have been implemented across the state, the Highway Safety Section per- forms evaluations using rigorous methods such as EB and comparison group before–after study designs. The more rigorous before–after methods are useful for evaluating countermeasures, particularly systemic applications, to help account for regression to the mean and sites with few or no crashes before implementation. Pennsylvania shares the results of these safety effective- ness evaluations with District Highway Safety staff through analysis reports. Highlights Pennsylvania has extensive documentation of the Highway Safety Program, which includes the HSIP. In addition, Pennsylvania is working with university partners to expand training related to components of the HSIP. The plan is to develop Khan Academy–style videos on topics such as “What is network screening and how do I use it?” The videos will be available on Pennsylvania DOT’s internal HSM website within the safety webpage. The following are other notable high- lights of the HSIP. Project Identification • Pennsylvania promotes the use of crash data and the HSM predictive method to identify opportunities for cost-effective improvement. • Pennsylvania supports project identification in each county by conducting network screening, performing detailed investigations for the highest ranked intersection and highest ranked segment in each county, and sharing the results with the local agencies. To perform the inves- tigations in 67 counties, the state DOT contracted support from consultants. This approach helps to get the counties through the initial stages of project identification and helps to dis- tribute funds among the counties. The resulting projects range from small signing and marking projects to multimillion-dollar interchange projects. • Pennsylvania recognizes that simply identifying locations with the highest number of crashes (or even the highest number of expected crashes) does not necessarily lead to cost-effective countermeasures. The guidance emphasizes the need to properly diagnose the problem and better understand the underlying crash contributing factors before developing appropriate cost-effective countermeasures. • In addition to crash data, Pennsylvania uses tort settlement data to identify potential locations for safety improvement. This is especially helpful for issues that do not fit in the category of a reportable crash (e.g., bike-only crash, pedestrian-only incident). As an example, Pennsylvania used tort data to identify inlet grates as an issue for bikes and pedestrians. This issue would not have been uncovered using traditional crash-based methods or even HSM-based methods. Project Prioritization • Pennsylvania uses benefit-cost analysis to prioritize spot projects and the CEI (cost per crash severity) to prioritize systemic projects. This helps to maximize the expected safety benefit for the amount of funds invested. • Pennsylvania developed a state-specific CMF guide, publishes annual crash costs, and pro- vides other guidance and resources to support benefit-cost analysis. This helps to promote consistency in the use of CMFs across the state.

Case Examples 63   • Pennsylvania uses an extensive list of factors for assessing and prioritizing proposed projects, including both crash-based and non-crash-based factors. Project Implementation • Although Pennsylvania does not allow for any shortcuts to implement spot or systemic HSIP projects, preliminary engineering is not required for countermeasures with standard drawings, such as signs, pavement markings, rumble strips, guide rails, median barrier, and lighting (21, 22). • Pennsylvania has made strong connections with the design community. Safety specialists with a background in design were able to speak a common language and explain the benefits of incorporating safety analysis in the design process. The HSM is now referenced in the design manual, and the predictive methods are commonly used to support design exceptions and general design decisions. • Pennsylvania is working to disseminate basic safety information to county road managers and maintenance personnel. For instance, the state DOT is providing information to help local stakeholders understand the safety benefits of basic maintenance activities, such as improving shoulder areas on curves. With this information, county managers can better understand where basic, low-cost curve improvements could provide a large safety benefit. Project Evaluation • Pennsylvania uses relatively simple methods for project-level evaluations and then applies more advanced methods for systemic- and countermeasure-level evaluations. Opportunities to Overcome Challenges One challenge related to Pennsylvania’s HSIP is the planning and development of local proj- ects. The current project application process requires a relatively basic concept of the proposed project (i.e., what the district or MPO would like to do). This involves a basic alternatives analy- sis for different project options using the HSM predictive method. There is no requirement (e.g., dollar value threshold or project impact) for a detailed preliminary engineering study. As such, the district or MPO may or may not complete a detailed preliminary engineering study. The lack of a preliminary engineering study can lead to issues such as underestimating project costs (i.e., initial project cost estimates are significantly different than the actual project costs). The lack of roadway segmentation and a complete data set for local roads has also been a chal- lenge. This is a particular concern for network screening and project identification where cur- rent methods are based on crash clustering by municipality or street name rather than individual segments. To overcome this issue, Pennsylvania has been working to complete the All Road Network of Linear Referenced Data (ARNOLD), which will provide the same segmentation as the state roads and allow for better screening. As these data become available, Pennsylvania will move toward a more automated process for local road screening. Another challenge related to HSIP projects on local roads is that there is not a formal policy and process for identifying, prioritizing, and implementing HSIP projects on the local system. There is a local Technical Assistance Program that can help to identify candidate project locations and develop appropriate countermeasures; however, the current process for contracting work on local roads is a challenge. Specifically, there were legal contracting issues and a lot of paper- work involved in the process. This led to a limited use of HSIP funds on local roads. To overcome this challenge, the state DOT is testing a new program that involves a local roads force account. MPOs can apply for the funds, and then the local agencies can use their own forces to implement low-cost improvements, such as signing and markings. Although this will result in relatively small-scale projects, it will be a quick-response opportunity to address safety issues on the local system, which has been an area in need of safety improvement in the state.

64 Practices for Balancing Safety Investments in a Comprehensive Safety Program Finally, Pennsylvania noted challenges related to systemic project evaluation. The primary issue is that it takes a lot of time to properly evaluate systemic projects. This is due to the large number of treated locations and the need to confirm the associated data (e.g., construction date, locations, crash data). Chapter Summary This chapter provided four case examples based on the literature review, survey responses, and follow-up interviews. The four case examples were selected to represent a diversity of basic and advanced state DOT practices for identifying, prioritizing, and evaluating HSIP projects, as summarized in Table 34. These case examples highlighted differences between spot and systemic projects as well as differences between state and local projects. HSIP Practice Maine North Carolina Oregon Pennsylvania Practices for identifying HSIP projects (basic or advanced) Advanced where possible and basic otherwise Basic warrants, but helps focus on overexposed crash types Basic measures: crash frequency, crash rate, and crash severity Advanced where possible and basic otherwise Practices for prioritizing HSIP projects (single program or set- asides) Set-asides for spot, systemic, and systematic projects Set-asides based on quantitative approach (funding allocation model) Set-asides to five regions based on proportion of K/A crashes. Even split between spot and systemic, but systemic is split by emphasis area based on Set-asides based on quantitative and qualitative approaches proportion of K/A crashes. Practices for evaluating HSIP projects (spot versus systemic) Limited to simple before–after studies Advanced evaluation practices Limited to simple before–after studies Advanced evaluation practices Table 34. Summary of case examples.