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.

21   The purpose of the survey is to identify, document, and summarize state DOT practices for identifying, prioritizing, and evaluating HSIP projects. The results of the survey highlight DOT practice, with the intent of helping agencies assess and improve their current HSIP practices. A survey was distributed to all 50 states, the District of Columbia, and Puerto Rico, targeting the person listed as the HSIP annual report contact. A total of 44 DOTs responded (42 states, the District of Columbia, and Puerto Rico), for a survey response rate of approximately 85%. The following sections provide a summary of the results. The structure of this chapter reflects the flow of the survey, with the following four general areas: • General safety management approach (spot, systemic, systematic) • Project identification • Project prioritization and funding allocation • Project evaluation For the purposes of the survey, the term “local” represents roads that are not owned by the state DOT (e.g., roads and highways owned or operated by a county, city, or township agency). Appendix A presents the complete list of survey questions, and Appendix B presents the responses to the survey. General Safety Management Approach This section provides a summary of general definitions and approaches to safety manage- ment. The survey presented the following definitions of three common approaches to safety management: 1. Spot: identifying locations based on crash experience (e.g., a high number or rate of crashes) and addressing the unique safety issues at each location. 2. Systemic: addressing crash types that result in fatalities and serious injuries by identifying risk factors for those crashes and implementing countermeasures widely across the network at locations where the risk factors are present. 3. Systematic: treating all eligible locations by incorporating safety countermeasures in design policies or implementing a proven countermeasure at all feasible locations. Table 6 presents answers to the question, “Does your state distinguish between spot, systemic, and systematic projects and, if so, are the definitions consistent with those provided above?” A majority of state DOTs (36) distinguish between spot, systemic, and systematic projects, with consistent definitions to those above. Three respondents indicated that their state DOT does distinguish between these approaches, but the definitions are different. Five respondents indi- cated that their state DOT does not distinguish between spot, systemic, and systematic projects. C H A P T E R 3 Survey of State DOT Practices

22 Practices for Balancing Safety Investments in a Comprehensive Safety Program Of the three state DOTs that indicated some difference in definitions, none of them reported differences in definitions for the spot method. Two state DOTs, North Carolina and Georgia, noted the following differences for the systemic definition: • North Carolina utilizes a hybrid version of systemic programs by combining risk factors and crash data, including pattern recognition, to determine specific countermeasures. The state DOT also uses field reviews to confirm roadway characteristic data. • Georgia uses other data, such as crash data, to support the systemic approach and prioritize projects. One state DOT, Alaska, noted that it does not specifically address or define systematic projects. The following two subsections provide further details on the systemic and systematic approaches. Systemic Approach The systemic approach to safety management includes addressing crash types that result in fatalities and serious injuries by identifying risk factors for those crashes and implementing countermeasures widely across the network at locations where the risk factors are present. Table 7 presents the number of state DOTs with or without a systemic safety component and whether it is documented. Thirty-seven state DOTs indicated a systemic safety component as part of the HSIP. Of those with a systemic component, 21 (approximately 57%) are documented. These results are consistent with the findings from the literature review, which showed that only five state DOTs do not specify a systemic method on the state system, and only six do not specify a systemic method on the local system. If state DOTs selected “Yes, and it is documented,” they were asked to provide a link or refer- ence to the document. Appendix B provides a list of reference material by state DOT. If state DOTs were not able to share a weblink, they were invited to upload attachments for documenta- tion. Four state DOTs provided documentation via upload. If state DOTs indicated “Yes, but it is not documented,” they were asked to briefly explain the systemic component. Refer to Appendix B for detailed responses to the survey. As shown in Appendix B, many state DOTs noted that they consider systemic safety through an informal Answer Choice Number of State DOTs Yes, and the definitions are consistent. 36 Yes, but the definitions are different. 3 No 5 Total 44 Answer Choice Number of State DOTs Yes, and it is documented. 21 Yes, but it is not documented. 16 No 5 No response 2 Total 44 Table 6. Methodology definitions (question 3). Table 7. Systemic safety component (question 4).

Survey of State DOT Practices 23   process, whereas others are pilot testing the use of a systemic approach. Several state DOTs indicated that they are working to formalize and memorialize systemic safety as part of their HSIP documentation. Systematic Approach The systematic approach to safety management includes treating all eligible locations by incorporating safety countermeasures in design policies or implementing a proven countermeasure at all feasible locations. As shown in Table 8, 31 of 44 state DOTs include a systematic safety com- ponent, of which 13 (45%) have it documented. The remaining 13 state DOTs indicated that they do not include a systematic safety component. Note that this is different than the results from the literature review, which only identified six state DOTs with a documented systematic process. The difference is likely due to the fact that the literature focused on state safety manuals and HSIP annual reports, which tend to focus on spot and systemic processes. Systematic pro- cesses may be documented in other manuals, such as design policies or engineering instructions. If state DOTs selected “Yes, and it is documented,” they were asked to provide a link or reference to the document. Appendix B shows the links and references to documentation for systematic safety. If state DOTs were not able to share a weblink, they were invited to upload attachments for documentation. Two state DOTs provided documentation via upload. State DOTs that selected “Yes, but it is not documented,” were asked to briefly explain their systematic component. Refer to Appendix B for detailed responses to the survey. As shown in Appendix B, some state DOTs noted similarities in their systemic and systematic methodologies and guidance. Other state DOTs have “preapproved” countermeasures that can be implemented on an as-able basis. General Documentation State DOTs were asked if they have documented HSIP policies, practices, manuals, and/or protocols for identifying, prioritizing, and evaluating HSIP projects. As shown in Figure 10, the majority of state DOTs (36 of 44) do have documentation. These documents were included in the literature review, and links are provided in Appendix C. Project Identification This section focuses on methods for identifying spot and systemic projects. The initial ques- tions reflect the methods to identify potential project locations, noting differences between methods used for state and local road systems. The questions then transition into the methods used to identify specific countermeasures, again noting differences between methods used for state and local road systems. The final questions in this section identify specific types of systemic countermeasures. Answer Choice Number of State DOTs Yes, and it is documented. 13 Yes, but it is not documented. 18 No 13 Total 44 Table 8. Systematic safety component (question 5).

24 Practices for Balancing Safety Investments in a Comprehensive Safety Program Identifying State System Projects Table 9 summarizes how the 44 responding state DOTs identify potential locations for spot and systemic HSIP projects on the state roadway system. For spot projects, the basic crash-based approach (e.g., crash frequency, rate, and severity) is the most common identification method, as noted by 36 state DOTs. For systemic projects, the basic risk-based approach (e.g., crash summaries, crash trees, and SHSP emphasis areas) is the most common identification method, as noted by 35 state DOTs. Note that respondents could select multiple responses to this question under each approach. Four state DOTs indicated “other” identification methods for HSIP projects on the state system, and three of those provided the following responses to explain (note that there is overlap in the “other” responses for spot and systemic): • Indiana uses substantive versus nominal index calculations. • Michigan focuses on fatal and serious injury crashes as well as crash types and patterns. This can be based on number of fatal and serious injury crashes at spot locations or using the approved systemic project list. • Pennsylvania uses tort settlement data to identify options for safety improvements. This is especially helpful for issues that do not fit in the category of a reportable crash. These include a bike-only crash, pedestrian-only-related incidents, and others. One of the things we are working on now based on torts is inlet grates. We know from our tort data that these are an issue when it comes to bikes and pedestrians but would not be able to figure out how big of an issue they are by using traditional crash-based methods or even HSM-based methods, because the predictive methods relate to vehicular-involved related crashes only. No 8 Yes 36 Figure 10. HSIP documentation (question 22). Method Spot Systemic Basic crash-based approach (crash frequency, rate, severity) 36 20 Basic risk-based approach (based on crash summaries, crash trees, SHSP emphasis areas) 24 35 Advanced crash-based approach (predicted, expected, excess expected crashes such as those included in the HSM) 20 11 Advanced risk-based approach (based on predictive crash models) 8 8 Other 3 2 Table 9. Identification methods on state system (question 6).

Survey of State DOT Practices 25   Identifying Local System Projects State DOTs were asked if they have a different process for identifying spot or systemic HSIP projects on the local system versus the state system. As shown in Figure 11, 32 of the 44 responding state DOTs (approximately 73%) do not have a different identification process for HSIP projects on their local system. If state DOTs indicated a different identification method for HSIP projects on the local system, they were asked to indicate those methods. Table 10 shows the responses from those 12 state DOTs. Similar to the state system, the most common identification method on the local system for spot projects was a basic crash-based approach (e.g., crash frequency, rate, and severity). For systemic projects, the results are also consistent with those from the state system where the basic risk-based approach (e.g., crash summaries, crash trees, and SHSP emphasis areas) is the most common identification method. Note that respondents could select multiple responses to this question under each approach. Four state DOTs indicated “other” identification methods for HSIP projects on the local system and provided the following responses to explain (note that there is overlap in the “other” responses for spot and systemic): • Kansas uses the HRRR Program for the local system and is developing LRSPs to identify projects for the program. The program includes both spot and systemic projects using different levels of analysis—some simple and some complex. • Maryland addresses the local system mostly through systemic improvements. The projects are submitted by local agencies to the state DOT for review and selection. No 32 Yes 12 Method Spot Systemic Basic crash-based approach (crash frequency, rate, severity) 9 6 Basic risk-based approach (based on crash summaries, crash trees, SHSP emphasis areas) 5 9 Advanced crash-based approach (predicted, expected, excess expected crashes such as those included in the HSM) 1 1 Advanced risk-based approach (based on predictive crash models) 0 2 Other 3 3 Figure 11. Different identification process for state versus local system (question 7). Table 10. Identification methods on local system (question 7).

26 Practices for Balancing Safety Investments in a Comprehensive Safety Program • New York uses the rate quality control method, but traffic volumes and rates are often not available on the local system. Although there are some exceptions, local governments perform their own analysis to identify systemic and spot HSIP projects. The state DOT is in the process of implementing a new Safety Management System (SMS), which will incorporate many of the more advanced predictive models in the HSM. The system was scheduled for production in fall 2021 and will be available for state and local governments. • Vermont uses the critical crash rate method for federal-aid roads only. Identifying State System Countermeasures Table 11 details how state DOTs identify potential countermeasures for spot and systemic HSIP projects on the state system. For spot projects, detailed site analysis is the most common method, as indicated by 42 of the 44 responding state DOTs. Alternatively, for systemic projects, the presence of risk factors or specific site characteristics is the most common method, as indi- cated by 35 of the 44 responding state DOTs. Note that respondents could select multiple responses to this question under each approach. Three state DOTs indicated “other” countermeasure identification methods for HSIP projects on the state system and provided the following responses to explain: • Delaware collaborates with others in Delaware DOT, law enforcement, MPOs, Office of Highway Safety, and other stakeholders to discuss annually in site review meeting. • Mississippi utilizes the CMF Clearinghouse for countermeasure evaluation when multiple options are presented. • Wyoming uses its SMS to identify spot and systemic treatments at each given location. Identifying Local System Countermeasures State DOTs were asked if they have a different process for identifying spot or systemic counter- measures on the local roadway system. As shown in Figure 12, 35 of the 43 responding state DOTs (approximately 81%) do not have a separate countermeasure identification method for the local system. If state DOTs indicated a different countermeasure identification method for HSIP projects on the local system, they were asked to indicate those methods. Table 12 shows the responses from those eight state DOTs. One state DOT, Maryland, indicated “other” countermeasure identification methods for HSIP projects on the local system. In Maryland, most improvements on the local system are systemic. The projects are submitted by local agencies to the state DOT for review and selection. Common Systemic Countermeasures Systemic countermeasures can be used to address specific focus crash types and risk factors. Table 13 presents the responses to state systemic countermeasure focus areas (note that respon- dents could select multiple answers). The majority of state DOTs (36 of 44) include roadway Method Spot Systemic Detailed site analysis (e.g., road safety audit) 42 17 Presence of risk factors or specific site characteristics 29 35 Preapproved list of countermeasures 19 26 Other 3 3 Table 11. Countermeasure identification methods on state system (question 8).

Survey of State DOT Practices 27   departures as a focus area for systemic improvement, followed by intersections (27 of 44), pedes- trians (19 of 44), and bicycles (10 of 44). Eight state DOTs noted “other” systemic countermea- sure focus areas that, for example, include wrong-way, corridor, and wildlife crashes. In addition, state DOTs were asked to identify typical risk factors and countermeasures related to each focus area. Table 14 presents specific risk factors that state DOTs use to imple- ment systemic roadway departure countermeasures. Some of the more common risk factors include horizontal curve cross-section (lane, shoulder, and median width), posted or advisory speed, traffic volume, and target crashes. Table 15 presents specific countermeasures that state DOTs use to address systemic roadway departure risk factors. Common countermeasures include rumble strips, enhanced curve signs, high-friction surface treatment, and enhanced pavement markings. Table 16 presents specific risk factors that state DOTs use to implement systemic intersection countermeasures. Some of the more common risk factors include traffic control, intersection geometry, traffic volume, and posted speed. No 35 Yes 8 Figure 12. Different countermeasure identification methods on state and local roads (question 9). Method Spot Systemic Detailed site analysis (e.g., road safety audit) 6 2 Presence of risk factors or specific site characteristics 3 5 Preapproved list of countermeasures 3 4 Other 1 1 Answer Choice Number of State DOTs Roadway departures 36 Intersections 27 Pedestrians 19 Bicycles 10 Other 8 Table 12. Countermeasure identification methods on local system (question 9). Table 13. Systemic countermeasure focus areas (question 10).

28 Practices for Balancing Safety Investments in a Comprehensive Safety Program Table 17 presents specific countermeasures that state DOTs use to address systemic inter- section risk factors. Some of the more common countermeasures include enhanced signing, enhanced striping, retroreflective backplates, and traffic signal upgrades. Table 18 presents specific risk factors that state DOTs use to implement systemic pedestrian countermeasures. Pedestrian risk factors are more varied among state DOTs. Some of the more common risk factors include posted speed, location (midblock versus intersection), pedestrian volume, pedestrian crashes, area type, and traffic control. Table 19 presents specific countermeasures that state DOTs use to address systemic pedes- trian risk factors. Some of the more common countermeasures include pedestrian crossings, Risk Factors Number of State DOTs Horizontal curve 19 Shoulder width 13 Posted or advisory speed 11 Traffic volume 10 Target crashes 10 Number of lanes or facility type 9 Median width or lack of separation 8 Lane width 7 Edge risk 2 Countermeasures Number of State DOTs Rumble strips 19 Enhanced curve signs 12 High-friction surface treatment 10 Enhanced pavement markings 9 Median barrier 5 Shoulder widening 4 Clear zone 4 Enforcement 4 Raised pavement markers 3 Safety edge 2 Risk Factors Number of State DOTs Traffic control 12 Intersection geometry (number of approaches, turn lanes, and skew) 11 Traffic volume 9 Posted speed 6 Sight distance 5 Proximity to curve/intersection/railroad 5 Functional class 3 Target crashes 3 Capacity 2 Table 14. Roadway departure risk factors (question 10). Table 15. Roadway departure countermeasures (question 10). Table 16. Intersection risk factors (question 10).

Countermeasures Number of State DOTs Enhanced signing 13 Enhanced striping 8 Retroreflective backplates 8 Traffic signal upgrades 6 Left-turn protection 4 Flashing yellow arrow 3 Transverse rumble strips 3 Channelization 3 Convert to all-way stop. 3 Improve sight distance. 3 Convert to roundabout. 3 Left-turn hardening 2 Lighting 2 Red light cameras 1 Dual left-turn lane 1 Improve left-turn offset. 1 Risk Factors Number of State DOTs Posted speed 6 Location (midblock versus intersection) 6 Pedestrian volume 5 Pedestrian crashes 5 Area type 5 Traffic control 5 Crossing distance 4 Land use and pedestrian generators (transit stops, shopping centers) 4 Traffic volume 4 Sidewalk and crosswalk presence and condition 4 Intersection sight distance 4 Socioeconomics and demographics 3 Lighting 3 Functional class 2 Median type 2 Countermeasures Number of State DOTs Pedestrian crossings 9 Rectangular rapid-flashing beacons 6 Pedestrian warning signs 5 Pedestrian signals and signal phasing 5 Lighting 5 Bulb-outs 4 Median refuge islands 4 Leading pedestrian intervals 3 Pedestrian hybrid beacon 3 Sidewalk 2 Countdown timers 2 Prohibit right turn on red. 2 Road diet 2 Left-turn hardening 1 Red light cameras 1 Table 17. Intersection countermeasures (question 10). Table 18. Pedestrian risk factors (question 10). Table 19. Pedestrian countermeasures (question 10).

30 Practices for Balancing Safety Investments in a Comprehensive Safety Program rectangular rapid-flashing beacons, pedestrian warning signs, pedestrian signals and signal phasing, and lighting. Table 20 presents specific risk factors that state DOTs use to implement systemic bicycle countermeasures. Fewer state DOTs identified bicycle risk factors, but those who did included factors such as bicycle crashes, bicycle volume, land use, socioeconomics, demographics, and roadway characteristics. Table 21 presents specific countermeasures that state DOTs use to address systemic bicycle risk factors. Some of the more common countermeasures include bicycle lanes, bicycle warning signs, bicycle crossings, bicycle route signing, and bicycle signals. A few state DOTs identified “other” typical risk factors used to implement systemic counter- measures for other safety concerns. For instance, Colorado identified risk factors and counter- measures to address animal crashes, Illinois identified risk factors and countermeasures to address horizontal curve crashes, and New Jersey identified risk factors and countermeasures to address wrong-way crashes. Massachusetts developed risk factors for nearly all SHSP emphasis areas, including intersections, roadway departure, pedestrians, bicycles, motorcycles, truck involved, younger drivers, older drivers, speeding, impaired driving, distracted driving, and occupant protection. Refer to Appendix B for detailed responses to the survey. Risk Factors Number of State DOTs Bicycle crashes 4 Bicycle volume 3 Land use and bicycle generators (transit stops, shopping centers) 2 Socioeconomics and demographics 2 Functional class 2 Posted speed 2 Roadway cross-section 2 Bike lanes 2 Lighting 2 Area type 1 Access density 1 Traffic volume 1 Countermeasures Number of State DOTs Bicycle lanes 6 Bicycle warning signs 6 Bicycle crossings 3 Bicycle route signing 3 Bicycle signal or interval at signalized intersections 3 Bicycle box 1 Buffer between bike lanes and vehicle lanes 1 Cycle track or bicycle boulevard 1 Left-turn hardening 1 Lighting 1 Prohibit right turn on red. 1 Red light cameras 1 Table 20. Bicycle risk factors (question 10). Table 21. Bicycle countermeasures (question 10).

Survey of State DOT Practices 31   Developing Spot and Systemic Projects The next step in the HSIP process is transitioning HSIP locations and countermeasures into projects. Table 22 describes various state DOT requirements for developing spot and systemic HSIP projects. Of the 44 responding state DOTs, a majority require preliminary engineering for all spot HSIP projects (29 state DOTs) and all systemic HSIP projects (18 state DOTs). For systemic projects, there are nearly as many state DOTs (16) that do not require preliminary engineering for countermeasures with standard drawings. Note that respondents could select multiple responses to this question under each approach. If a state DOT selected “other,” they were asked to explain. A summary of the responses follows. Refer to Appendix B for the explanations of other project development requirements for spot and systemic projects. The following are examples of spot project requirements: • All spot projects require plans of proposed countermeasures, although plans do not have to be highly detailed or complete. • All HSIP projects require a feasibility study as the first step. • Preliminary engineering is not required for countermeasures with standard drawings. • Preliminary engineering is almost always required, but for a small set of countermeasures, construction projects may be let with proposal-only plans developed in-house. • A basic concept of the proposed project is required, which involves a basic analysis with different project options. The following are examples of systemic project requirements: • All systemic projects require plans of proposed countermeasures, although plans do not have to be highly detailed or complete. • All HSIP projects require a feasibility study as the first step. • Some systemic projects can be created using a log approach, but this still requires funding for a preliminary engineering (PE) phase to create the project. • Preliminary engineering is required for all systemic HSIP projects. • The No Plans Contract (NPC) method can be used to deliver systemic projects under several programs if the projects are easily implemented and require no acquisition of ROW. State DOTs were asked if they streamline project planning and development in the interest of implementing spot or systemic HSIP projects more quickly (e.g., not requiring all steps of the typical planning and development process for certain safety projects, purchasing products to be installed by maintenance forces to skip going out to bid for construction). Of the 44 state DOTs, 24 do not streamline any spot or systemic HSIP projects, whereas 20 streamline either spot only, systemic only, or both, as presented in Table 23. State DOTs that selected any of the “yes” responses were asked to provide a brief descrip- tion of how they streamline the respective projects (spot and/or systemic). Approaches include Answer Choice Spot Systemic Preliminary engineering is required for all HSIP projects. 29 18 Preliminary engineering is not required for countermeasures with standard drawings. 8 16 Preliminary engineering is required but only for certain projects (e.g., those that impact ROW or exceed a dollar threshold). 6 9 Other 6 8 Table 22. Project development requirements (question 11 and question 12).

32 Practices for Balancing Safety Investments in a Comprehensive Safety Program providing a list of preapproved countermeasures, implementing projects on an emergency basis, using force accounts, or implementing low-cost projects, such as signing or pavement markings with in-house forces. Appendix B provides further details on state DOT responses. Project Prioritization and Funding Allocation This section focuses on the various methods state DOTs use to prioritize and allocate funding for spot and systemic projects for implementation. The initial questions focus on the quantita- tive factors considered in project prioritization, including economic measures and the methods used to estimate safety benefits. Questions then transition to more qualitative and non-crash- based factors used in project prioritization. Finally, this section identifies methods that state DOTs use to allocate funding between spot and systemic projects. Throughout, the questions seek to identify differences between methods for spot and systemic projects as well as differences for projects on the state and local system. Project Prioritization Methods Figure 13 presents a summary of how state DOTs prioritize HSIP projects on the state system. Of the 44 responding state DOTs, benefit-cost ratio based on total crashes is the most common prioritization method on the state system. Systemic prioritization on the state system is more varied, with “other” being the most common, as indicated by 16 state DOTs. The next most common systemic prioritization methods on the state system were “benefit–cost ratio Answer Choice Number of State DOTs No, we do not allow for any shortcuts. 24 Yes, for both spot and systemic projects 8 Yes, for systemic projects 8 Yes, for spot projects 4 Total 44 Table 23. Streamline project planning and development (question 13). Spot Systemic0 5 10 15 20 25 30 Be ne fit- Co st Ra tio (to tal cr ash es) Be ne fit- Co st Ra tio (s ev ere cr ash es on ly) Co st- Eff ec tiv en ess (to tal cr ash es) Co st- Eff ec tiv en ess (s ev ere cr ash es on ly) Ne t B en efi ts (to tal cr ash es) Ne t B en efi ts (se ve re cra sh es on ly) Co st- Jus tifi ca tio n A na lys is Ot he r Figure 13. Prioritization methods on state system (question 14).

Survey of State DOT Practices 33   based on total crashes” (12 state DOTs) and “cost-effectiveness based on total crashes” (12 state DOTs). Note that respondents could select multiple responses to this question under each approach. State DOTs that selected “other” were asked to explain. Other project prioritization consider- ations by state DOTs include project readiness, ease of implementation, and other variations on benefit-cost analysis. Refer to Appendix B for more details on state DOT responses. State DOTs were then asked if they have different processes for prioritizing HSIP projects on the local system. As shown in Figure 14, a majority of state DOTs (33 of 44) do not have a different prioritization method. This is consistent with the results reported in the Chapter 2 literature review (33 of 52 state DOTs do not have a different prioritization method). If state DOTs selected “yes,” they were asked to select the applicable process(es) they use to prioritize HSIP projects on the local system. Figure 15 indicates the responses for these 11 state No 33 Yes 11 0 1 2 3 4 5 6 7 8 9 Spot Systemic Be ne fit- Co st Ra tio (to tal cr ash es) Be ne fit- Co st Ra tio (s ev ere cr ash es on ly) Co st- Eff ec tiv en ess (to tal cr ash es) Co st- Eff ec tiv en ess (s ev ere cr ash es on ly) Ne t B en efi ts (to tal cr ash es) Ne t B en efi ts (se ve re cra sh es on ly) Co st- Jus tifi ca tio n A na lys is Ot he r Figure 14. Different prioritization methods on state and local roads (question 15). Figure 15. Prioritization methods on local system (question 15).

34 Practices for Balancing Safety Investments in a Comprehensive Safety Program DOTs. For spot projects, the most common method is “other” (six state DOTs), followed by the benefit-cost ratio based on total crashes (five state DOTs). For systemic projects, the most common method is “other” (eight state DOTs). Note that respondents could select multiple responses to this question under each approach. State DOTs that selected “other” were asked to explain. “Other” prioritization methods are typically reliant on local jurisdictions for prioritization. Refer to Appendix B for more details on state DOT responses. Quantifying expected safety benefits can be a key metric for prioritizing HSIP projects. State DOTs were asked if they quantify expected safety benefits for spot and systemic HSIP projects, as shown in Figure 16. Of the 44 responding state DOTs, a majority (37) do quantify expected safety benefits. If state DOTs selected “yes,” they were asked to indicate which method(s) they use to quantify the expected safety benefits of proposed projects. As shown in Table 24, the most common method is using CMFs, as noted by 32 of 37 state DOTs for spot projects and 27 of 37 state DOTs for systemic projects. The next most common method is crash history, as noted by 28 of 37 state DOTs for spot projects and 18 of 37 state DOTs for systemic projects. Note that respondents could select multiple responses to this question under each approach. Six state DOTs indicated “other” safety benefit estimation methods and provided the following responses to explain (note that there is overlap in the “other” responses for spot and systemic): • Kansas: For systemic projects, this only applies to lighting. • Massachusetts: It depends on the project and if there are state-specific or HSM SPFs or CMFs that apply. In other cases, Massachusetts uses crash history or some other method. Yes 37 No 7 Figure 16. Quantify expected safety benefits (question 16). Method Spot Systemic CMFs 32 27 Crash history 28 18 SPF (HSM) 13 5 SPF (state developed) 8 5 Other 3 4 Table 24. Safety benefit estimation methods (question 16).

Survey of State DOT Practices 35   Massachusetts tried quantifying expected safety benefits using the Safe System Approach, but that is relatively new, and it is looking into various options. • Michigan: Created state-specific SPFs for rural roadways and developed an HSM spreadsheet for analysis. • New York: Moving toward state-developed SPFs. • Ohio: Does not estimate for systemic projects. • Vermont: Does not quantify expected benefits for systemic projects. State DOTs were asked if they consider non-crash-based factors in prioritizing HSIP projects, beyond project costs and crash-based benefits. As shown in Figure 17, slightly more than half (23 of 44 state DOTs) responded yes, they do consider non-crash-based factors. If state DOTs selected “yes,” they were asked to select the applicable non-crash-based factor(s) they consider in prioritizing HSIP projects. Table 25 shows the responses from these 23 state DOTs. The most common non-crash-based prioritization factors include SHSP priority (17 state DOTs) and project readiness (14 state DOTs). Note that respondents could select multiple responses to this question. If state DOTs noted that they use “other” non-crash-based prioritization, they were asked to explain. Other non-crash-based prioritization factors include opportunity for project bundling, tort liability issues, local support, and alignment with planned projects and other priorities (e.g., asset condition, mobility and connectivity, economic access, resiliency, environment, and health access). Refer to Appendix B for detailed responses. Yes 23 No 21 Answer Choice Number of State DOTs SHSP priority 17 Project readiness 14 Geographic equity (region/district or urban/rural split) 9 State/local equity 8 Social equity 6 Other 13 Figure 17. Non-crash-based prioritization factors (question 17). Table 25. Applicable non-crash-based prioritization factors (question 17).

36 Practices for Balancing Safety Investments in a Comprehensive Safety Program Funding Allocation Methods This subsection focuses on funding allocation between spot and systemic projects. Specifically, state DOTs were asked how they balance HSIP funds among spot, systemic, and systematic projects. As shown in Table 26, of the 43 state DOTs responding to this question, most (27 state DOTs) noted that all projects compete for the same funds, whereas 11 indicated that there is a set-aside for specific programs. This is consistent with the results from the Chapter 2 literature review, which indicated that 12 of 52 state DOTs use a set-aside for spot or systemic programs. Note that the literature review also uncovered other set-asides by agency (state and local) and initiative (e.g., roadway departure, intersection, pedestrian). Five state DOTs indicated “other” HSIP funding allocation methods and provided the following responses to explain: • Massachusetts: Uses project readiness to prioritize and allocate funds. • Michigan: Distributes HSIP funds as follows: – 75% state and 25% local – For state roads, each region needs to use 25% of funds on systemic projects. – For local roads, a certain amount is not set aside, but the application process is streamlined for systemic projects. • New Jersey: All projects compete for the same funds, but there is an ongoing effort to strategi- cally align HSIP funds in the future to delineate the systemic and spot HSIP funding. • Oregon: Has the following statewide funding goals: spot = 50% and systemic = 50%. The systemic portion is further split into roadway departure = 50%, intersection = 35%, and bike/ pedestrian = 15%. • Vermont: Spends approximately $400,000 annually for systemic safety projects on local roads. This is the set-aside amount that was established by SAFETEA-LU for the HRRR Program. Vermont continued to program the same amount after MAP-21 eliminated the set-aside. For the 11 state DOTs that indicated the use of set-asides, Table 27 describes the methods they used to determine the set-aside amount for each program. Of the 11 state DOTs, seven noted that they use a qualitative approach to set-aside funding, whereas only one indicated a quantitative approach. State DOTs were asked to explain “other” set-aside funding methodologies. Some state DOTs noted that they establish approximate funding goals and then adjust based on circumstances. Answer Choice Number of State DOTs All projects compete for the same funds. 27 There is a set-aside for each program type, and projects compete within the designated program. 11 Other 5 Answer Choice Number of State DOTs Qualitative (e.g., based on professional judgment) 7 Quantitative (e.g., based on past performance of the programs, percentage of fatal and serious injury crashes at high-crash locations [HCLs], or percentage of fatal and serious injury crashes in rural and urban areas) 1 Other 3 Table 26. HSIP funding allocation (question 18). Table 27. Set-aside funding methodology (question 18).

Survey of State DOT Practices 37   Delaware funds short-term spot improvements first and then allocates the remaining funds to systemic projects. North Carolina developed a funding allocation model after reviewing crash data trends, SHSP emphasis areas, existing HSIP subprograms and processes, and historical countermeasure selections. Refer to Appendix B for detailed responses. Table 28 indicates the number of state DOTs that have a certain goal for funding systemic projects. Of the 44 responding state DOTs, the majority (30) do not have a funding goal for systemic projects. Of the 10 state DOTs that do have a funding goal, only three are based on a documented formula or process. This is consistent with the results from the Chapter 2 literature review, which identified few state DOTs with a documented formula or process for establishing funding goals. The 10 state DOTs that selected one of the “yes” options and the four that selected “other” are explored further in the discussion. Three state DOTs selected “Yes, based on documented formula or process,” and provided the following responses to explain: • North Carolina allocates funding based on the following distribution: 50% roadway departure (80% systemic, 20% responsive), 35% intersection (40% systemic, 60% responsive), and 15% pedestrian & bike (40% systemic, 60% responsive). • Oregon uses a 50/50 split by project approach (spot and systemic). Although it is recognized that systemic projects are more proactive and may result in more cost-effective projects, there are situations when systemic projects are not as desirable. For example, without careful planning and combining of projects, the lower-cost systemic projects may not make the best federally funded projects because of the high overhead costs and the difficulty of bundling projects. Spot projects are typically higher cost and require more time and effort to deliver, which seem better suited for HSIP funding. • Wyoming determines the set-aside for systemic and spot projects on an annual basis. Seven state DOTs selected “Yes, but not based on documented formula or process,” and provided the following responses to explain: • In Indiana, the Safety Asset Team seeks to approach a 50/50 split of spot and systemic project funding allocation. Systematic improvements are achieved via road standards. • Louisiana proposed a target split in the HSIP Implementation Plan based on a balanced program and not focused on only spots or systemic. • Maine uses a set-aside for pavement marking retroreflectivity, with the remainder used for systemic and spot improvement projects. Maine targets 50% of the remaining funds for systemic and spot improvements, but it can vary each year depending on the candidate projects in each grouping and an assessment of priority by the Safety and Mobility Committee. • Maryland determines the percentage of systemic projects by the projects selected for HSIP support. • In Minnesota, each region has different priorities but nothing official. However, the following is a typical distribution: Metro state highway HSIP is largely reactive (+75%), with the remainder as proactive; Metro local highway HSIP is approximately a 50/50 split; Greater Minnesota Answer Choice Number of State DOTs No 30 Yes, but not based on documented formula or process 7 Yes, based on documented formula or process 3 Other 4 Total 44 Table 28. Funding goal for systemic projects (question 19).

38 Practices for Balancing Safety Investments in a Comprehensive Safety Program state highway HSIP is approximately 60% reactive and 40% proactive; and Greater Minnesota local highway HSIP is approximately 90% proactive and 10% reactive. • In Nebraska, the HSIP Implementation Plan focuses on systemic projects due to the need to improve many miles of rural roads as a key category with crashes. • Washington sets preliminary targets for different strategies. For the seven state DOTs that selected “Yes, but not based on documented formula or pro- cess,” Table 29 indicates the approximate funding goals. Percentages for systemic ranged from 0% (note that the other percentages are 50% each for spot and systematic) to 75%. Louisiana had the highest spot project funding goal with 80%, whereas Nebraska had the highest systemic project funding goal with 75%. Indiana, Maine, and Maryland have relatively balanced funding goals at approximately 50/50. Four state DOTs selected “other” set-aside funding methodology and provided the following responses to explain: • Connecticut is working toward a higher percentage of systemic and systematic projects. • Massachusetts would like to implement more systemic projects but needs to figure out how to overcome current challenges (i.e., ROW and survey requirements) to make these projects worthwhile. For instance, low-cost projects can end up with extremely high design costs when following the current project development process. • Michigan is focusing on systemic projects to cover more ground on safety improvements. Aside from the goal of allocating at least 25% on the state system on a 5-year rolling average, there is not a specific goal for implementing systemic projects. • South Carolina does not have an official split but is targeting a 50/50 split. Project Evaluation (Post-Implementation) Evaluating spot and systemic projects is the final step in the HSIP process but provides a critical feedback loop to future project selection decisions. Table 30 shows the ways state DOTs evaluate spot and systemic HSIP projects on the state system post-implementation. Of the State Spot % Systemic % Systematic % Indiana 50 50 0 Louisiana 80 20 0 Maine 50 50 0 Maryland 46 52 0 Minnesota 50 0 50 Nebraska 25 75 0 Washington 30 70 0 Table 29. Not documented funding goals (question 19). Method Spot Systemic Simple before–after 32 23 Empirical or full Bayes before–after 10 9 Comparison group before–after 5 8 Regression cross-section 1 1 Other 4 6 We don’t evaluate. 5 9 Table 30. Evaluation methods on state system (question 20).

Survey of State DOT Practices 39   44 responding state DOTs, a majority conduct simple before–after evaluations for spot projects (32 state DOTs) and systemic projects (23 state DOTs). Note that respondents could select multiple responses to this question under each approach. Seven state DOTs selected “other” evaluation methods for the state system and provided the following responses to explain (note that there is overlap in the “other” responses for spot and systemic): • Colorado has not determined a methodology for systemic projects yet. • Connecticut plans to hire a consultant to conduct before–after studies for all safety projects. • In Illinois, it depends on what data are available and how much. • In Maryland, spots are evaluated on a case-by-case basis. • Massachusetts uses the EB method when possible but looks for other options when it is not feasible. As an example, some systemic projects did not have accurate traffic volumes, so they used alternative methods, such as before–after, with adjustment for reporting. In some cases, Massachusetts does not evaluate projects because of implementation type. For instance, Every Day Counts initiatives and other countermeasures with proven effectiveness measures are not evaluated. • Michigan approves systemic projects based on previous research (national or Michigan-specific). Michigan performs before–after studies separately for state and local projects. • Ohio performs before–after studies with traffic volume correction. State DOTs were asked if they have different processes for evaluating the safety performance of spot and systemic projects on the local system post-implementation. Of the 44 responding state DOTs, the majority (34) noted that they do not have a different evaluation process for the local system, as shown in Figure 18. If state DOTs selected “yes,” they were asked to select the applicable process(es) used to evaluate the safety performance of spot and systemic projects on the local system post-implementation. Table 31 shows the responses for these 10 state DOTs. For spot projects on local roads, the most common method is the simple before–after (four state DOTs); however, an equal number of state DOTs (four) also indicated they do not evaluate local spot projects. For systemic projects on local roads, the most common response was “other.” Note that respondents could select multiple responses to this question under each approach. Yes 10 No 34 Figure 18. Different evaluation process for state and local roads (question 21).

40 Practices for Balancing Safety Investments in a Comprehensive Safety Program Four state DOTs selected “other” evaluation methods for the local system and provided the following responses to explain: • Maryland is developing a program to allocate HSIP fund to locals and plans to support local projects starting in FY 2022. There have been no evaluations done yet. • Michigan performed an analysis of the FY 2013 local agency programs to assess the effective- ness of the program as a whole and the specific safety countermeasures implemented as a part of these programs. The comprehensive post-installation study examined the preinstallation existing conditions at each project location, the safety countermeasures implemented, and the relevant historical crash and traffic volume data. The impacts of each project, program, and specific countermeasures were assessed using two analytical techniques: simple before– after and EB before–after. • New York has a Post-Implementation Evaluation System that is available only to evaluate local roads. The new SMS was scheduled for implementation in fall 2021 and will be able to evaluate state and local road projects. • Washington measures systemic projects on a larger scale (often agencywide) in a simple before–after comparison. Chapter Summary The survey results are generally consistent with those presented in Chapter 2 based on the literature review. The following are key program-level takeaways from the 44 survey responses, followed by key takeaways related to each specific area (project identification, prioritization, and evaluation): • A majority of state DOTs (36 of 44) distinguish between spot, systemic, and systematic projects with definitions that are consistent with industry standards. • The majority of state DOTs (36 of 44) document the process through HSIP policies, practices, manuals, and/or protocols for identifying, prioritizing, and evaluating HSIP projects. • Although 37 state DOTs include a systemic safety component as part of the HSIP, only 21 (approximately 57%) document the process. • A majority of state DOTs (31 of 44) include a systematic safety component as part of the HSIP, of which 13 (approximately 30%) document the process. • Much of the existing documentation focuses on the spot approach. Project Identification Of the 44 responding state DOTs, all have some method to identify potential locations for spot and systemic HSIP projects. The following are key takeaways related to specific methods, which are consistent with the results presented in Chapter 2 based on the literature review: • For spot projects, the most common method is the basic crash-based approach (e.g., crash frequency, rate, and severity), as noted by 36 state DOTs. Method Spot Systemic Simple before–after 4 3 Empirical or full Bayes before–after 2 1 Comparison group before–after 1 1 Regression cross-section 0 0 Other 2 4 We don’t evaluate. 4 3 Table 31. Evaluation methods on local system (question 21).

Survey of State DOT Practices 41   • For systemic projects, the most common method is the basic risk-based approach (e.g., crash summaries, crash trees, and SHSP emphasis areas), as noted by 35 state DOTs. • The methods are generally consistent on state and local roads where 32 state DOTs (approxi- mately 73%) use the same identification process for HSIP projects on the local system. • Detailed site analysis is the most common countermeasure identification method for spot projects, as indicated by 42 state DOTs. • The presence of risk factors or specific site characteristics is the most common countermeasure identification method for systemic projects, as indicated by 35 state DOTs. • The countermeasure identification methods are relatively consistent on both state and local roads, as indicated by 35 state DOTs (approximately 81%). Note that there were only 43 responses to this question. Project Development State DOTs have various requirements for developing spot and systemic HSIP projects; however, there appears to be more flexibility in the requirements for systemic projects. The following are key takeaways related to specific project development requirements based on the 44 survey responses: • A majority of state DOTs (29) require preliminary engineering for all spot projects. • For systemic projects, 18 state DOTs require preliminary engineering for all systemic proj- ects, but 16 do not require preliminary engineering for countermeasures with standard drawings. • The majority of responding state DOTs (24, or approximately 54%) do not allow for any shortcuts in the planning and project development process, eight (approximately 18%) allow for some streamlining of both spot and systemic projects, eight (approximately 18%) allow for streamlining in systemic projects, and four (approximately 9%) allow for streamlining in spot projects. Project Prioritization The following are key takeaways from the 44 survey responses related to project prioritization methods, which are consistent with the results presented in Chapter 2 based on the literature review: • The benefit-cost ratio is a common method. • For spot projects, the majority of state DOTs (29) use some type of benefit-cost ratio, and most of these (25) include total crashes as part of the benefit-cost analysis. • For systemic projects, the use of benefit-cost ratio is less common with only 15 state DOTs indicating some type of benefit-cost analysis. • Cost-effectiveness is just as common for systemic project prioritization, as indicated by 16 state DOTs. • Project prioritization methods are relatively consistent on both state and local roads, as indi- cated by 33 of the 44 responding state DOTs (75%). • A majority of state DOTs (37 of 44) quantify the expected safety benefits for use in project prioritization, and the most common methods include CMFs (32 for spot and 27 for systemic) and crash history (28 for spot and 18 for systemic). • More state DOTs quantify the expected safety benefits for spot projects compared to systemic projects. • Beyond project costs and crash-based benefits, more than half of the state DOTs (23 of 44) also consider non-crashed-based factors in prioritizing HSIP projects, including SHSP priority, geographic equity (region/district or urban/rural split), and project readiness.

42 Practices for Balancing Safety Investments in a Comprehensive Safety Program Funding Allocation The survey results related to funding allocation are relatively consistent with those presented in Chapter 2 based on the literature review, but there are a few differences. The following are key takeaways related to funding allocation methods based on the 44 responses: • The majority of state DOTs (27) noted that all HSIP projects compete for the same funds, regardless of project type (spot, systemic, or systematic). • Although only 11 state DOTs indicated the use of set-asides for specific programs, it became apparent that more state DOTs use this type of funding allocation approach when reviewing the specific responses and HSIP documentation. For example, four of the state DOTs that indicated all HSIP projects compete for the same funds indicated in another response that the agency has a goal for funding systemic projects, but it is not based on documented formula or process. • Some state DOTs are moving toward the use of set-asides to delineate funding for spot and systemic programs or for focus areas, such as roadway departures, intersections, pedestrians, and bicycles. • Of the 11 state DOTs that indicated the use of set-asides, the majority (seven) use a qualita- tive approach to establish the set-asides, whereas only one indicated a quantitative approach. Similarly, of the 10 state DOTs that indicated funding goals for systemic projects, only three have a documented formula or process. • The results from the literature review and survey are nearly identical in terms of state DOTs that do not have different processes for prioritizing HSIP projects on the local system (33 of 44 from the survey and 33 of 52 from the literature review). • The literature review identified the use of more set-asides than the survey, but the literature review included allocations by project type (spot and systemic), agency (state and local), and initiative, whereas the survey focused on allocations by project type only (spot and systemic). Considering the results of the literature review by project type only, the results are nearly identical (11 of 44 from the survey and 12 of 52 from the literature review). The results are also consistent in that few of the state DOTs use a formula to allocate funding between projects or initiatives. Project Evaluation Post-implementation evaluation is the final step in the HSIP process, providing a critical feedback loop to future project selection decisions. The following are key takeaways from the 44 survey responses related to project evaluation methods, which are consistent with the results presented in Chapter 2 based on the literature review: • A majority of state DOTs conduct simple before–after evaluations for spot projects (32 state DOTs) and systemic projects (23 state DOTs). • Relatively few state DOTs indicated the use of more advanced (EB or comparison group) before–after studies, but it appears that these advanced studies are equally popular for spot and systemic project evaluations (12 state DOTs each). • State DOTs typically evaluate all projects in a similar framework regardless of how the projects were identified or prioritized (e.g., identified by spot, systemic, or systematic approaches or implemented on a state or local road). For instance, 34 state DOTs (approximately 77%) use the same evaluation process for HSIP projects on the local system.