SUMMARY

This report synthesizes the findings of NCHRP Project 17-97, “Strategies to Improve Pedestrian Safety at Night”; the research and findings from Phases I and II informed the development of a guidance document for transportation professionals to increase safety for pedestrians in darkness for the third and final phase of the project. The first phase consisted of a comprehensive literature review and a state of the practice survey. The second phase consisted of four studies: a macro-level, nationwide analysis of trends in pedestrian fatalities from 2010-2020; a multicity, case-control analysis to investigate characteristics of severe pedestrian injuries in darkness; a driver simulation experiment examining driving behavior in darkness; and focus groups exploring decision-making and behavior while walking and driving at night. Phase II also included practitioner interviews to gain insights into how the guidance developed in Phase III could be most helpful. Overall, the research findings corroborated other studies’ findings and provided new insights to inform the guidance that is the ultimate output of this project. The following are key findings from each analysis.

Literature Review

Overall Trends

• Pedestrian fatalities have occurred disproportionately in darkness since the Fatality Analysis Reporting System (FARS) began collecting data, but the proportion of fatalities in darkness has steadily and alarmingly grown over time.
• Factors influencing pedestrian fatalities at night appear to primarily relate to drivers’ seriously compromised ability to see and therefore detect and react to a pedestrian in the roadway in time to avoid a crash.
  • – Higher speeds significantly negatively impact the ability to slow in time, even when a pedestrian has been detected, and these speeds are significantly associated with arterial classification and multilane roadways.
  • – Co-occurring risk factors like driver and pedestrian intoxication and driver drowsiness increase pedestrian risk at night, as does driver distraction, although the latter factor lacks rigorous data for understanding the extent of the problem.
• The rise in pedestrian fatalities in darkness since 2010 relates to multiple factors that are interrelated in complex ways that researchers are still working to understand.
  • – Primary risk factors like speed and roadway design have always been high-risk, and pedestrian exposure to those risk factors appears to have increased in ways that are not fully understood. Some hypotheses include increased houselessness and the suburbanization of poverty.
  • – Other risk factors include changing vehicle designs and an increase in risky driver behavior due to increasing smart phone use during this time period.

Safe Roads Findings

• Higher-speed, multilane roadways are consistently noted as risk factors for pedestrians, and their influence as risk factors increases in darkness. Arterial roadways and freeways are also risk factors in general and particularly at night.
• Midblock locations are more prevalent than intersections as the sites of pedestrian fatalities at night.
• Roadway lighting can be a key countermeasure to improve pedestrian safety but is not sufficient to address the problem on its own – particularly in higher-risk locations.
• Pedestrian countermeasures that alert drivers to crossing pedestrians can be important parts of a strategy to improve pedestrian safety at night. Lighted countermeasures are particularly effective at night.

Safe Vehicles Findings

• Headlights play an important role in a driver’s ability to see, but adaptive headlighting is not yet standard on U.S. vehicles, and some headlight designs may have unintended negative consequences regarding glare. More research in this area would help clarify the impact of newer headlight designs on glare and driver vision and safety..
• Interior lighting is known to negatively impact driver visibility, but research has lagged with regard to challenges to driver vision associated with smart phones and newer in-car infotainment systems.
• Advanced driver assistance systems (ADAS) features show promise in alerting drivers and providing protection for pedestrians, but these features are not yet routine on cars. Additional research in this area could help determine an optimal way to alert drivers to pedestrian detection.
• Vehicle height, width, and mass have increased dramatically over time, and hood shapes have become taller and boxier. All of these changes increase pedestrian injury severity in the event of a crash. While some research has tied larger vehicles to increasing pedestrian risk overall, research establishing this connection in darkness is mixed; however, as larger vehicles increasingly permeate the fleet, any related risk will become more pronounced.

Safe Speeds Findings

• Higher speeds are indisputably linked to greater pedestrian injury severity and a higher likelihood of death, and this is exacerbated in darkness.
• Roadway design features that facilitate higher speeds are consistently associated with pedestrian injury severity, particularly in darkness.
• Little research has examined the impact of posted speed limits on pedestrian safety generally and particularly in darkness, including the potential for lower nighttime speed limits to improve safety.

Safe Road Users Findings

• The human tendency to receive information visually is significantly compromised in darkness, leading to a significant – but undertreated – impairment for drivers. Headlights alone cannot offset high-risk circumstances.
• Drivers appear to have a particularly hard time detecting pedestrians at night when they approach from the left side, likely due to significantly reduced peripheral vision and the tendency to focus in front of the car, where headlights shine.
• Drivers also cannot rapidly adapt to changes in lighting, which may occur with oncoming headlights, in higher-speed areas with different types of street lighting, and if a driver uses a lighted device (e.g., smartphone).
• Drivers also routinely underestimate how compromised their vision is at night and overestimate the ability of the car or technology to help them see.
• Pedestrians are compromised in their ability to understand how quickly a vehicle is approaching, particularly at night. If they assume that drivers can see them, they may inadvertently place themselves in danger, particularly in a higher-risk circumstance like a higher-speed or multilane roadway.
• Pedestrians are more visible when they wear retroreflective clothing, particularly when the clothing highlights biological motion (i.e., moving arms and legs). However, pedestrian conspicuity campaigns have not been shown to be effective, and they place an undue and often unrealistic burden on pedestrians rather than addressing vehicle speed, the root of the safety problem. An emphasis on conspicuity is therefore a low priority in a Safe System Approach.
• Pedestrian fatalities at night are significantly associated with both pedestrian and driver intoxication, a behavior that significantly impacts judgment and disproportionately occurs at night.

• For drivers, technology to prevent intoxicated driving is an important part of the Safe System Approach, as are alternatives to driving, such as public transit and rides-for-hire.
• Intoxicated pedestrians are not breaking the law but are still at elevated risk due to impaired judgment. Lower-speed infrastructure that discourages risky driver behavior, reduces the likelihood of a crash, and results in lower injury severity if a crash occurs will help protect impaired pedestrians.

Demographic Characteristics Associated with Pedestrian Fatalities

• Research shows that Black, Native American, and Hispanic/Latino pedestrians experience an elevated pedestrian fatality risk on a per-capita and per-trip basis.
• Predominantly Black and Hispanic/Latino neighborhoods tend to be collocated with higher-risk roadways.
• Native Americans are at higher risk due to a lack of pedestrian infrastructure near and within Tribal areas.
• There may be other factors, such as a greater likelihood of working nighttime jobs (and therefore being more likely to travel in darkness), that impact pedestrian risk at night for subgroups within these larger racial and ethnic groups.
  Some research indicates that Black pedestrians may face greater roadway risk given findings that drivers are less likely to yield to Black pedestrians than to white pedestrians.
• Low-income neighborhoods are similarly associated with and negatively impacted by the presence of higher-risk roadway design and operations.
• Pedestrians who are shorter or in a wheelchair may be harder for drivers to see and therefore at greater risk of being hit. Reacting in the face of risk is also more difficult for disabled pedestrians.
• Some research suggests greater risk for unhoused pedestrians; this risk may be particularly acute for people who live on DOT land that surrounds high-speed, high-volume roadways and provides little pedestrian accommodation.
• Rural pedestrians experience a lack of pedestrian infrastructure that increases overall risk.

State of the Practice Practitioner Survey

Risk Factors

• Participants that indicated that pedestrian safety at night was a problem in their community most commonly named street type, crossing location, time of day, and demographics of those impacted as key risk factors.
• Participants tended to agree that pedestrian nighttime risk is higher for certain groups of people, including those who are houseless and Black and Brown pedestrians.
• Participants also tended to agree that pedestrian nighttime risk is elevated on higher-speed roadways, influenced by the presence and quality of street lighting, and higher in lower-income areas in their communities.

Programs and Policies to Improve Pedestrian Safety at Night

• While a majority of the respondents indicated that they implement programs or policies to improve pedestrian safety, fewer respondents – about a quarter of agencies – indicated that they implemented programs and policies specifically to improve pedestrian safety at night.
• Efforts to improve pedestrian safety at night included pedestrian safety messaging campaigns targeted to pedestrians, pedestrian safety messaging campaigns targeted to drivers, bicyclists, or other non-pedestrian road users, and lighting policies.

• Many agencies implemented at least two programs or policies, including the programs or policies listed above as well as others, such as speed limit policy changes and the enforcement of other strategies to deter risky driving behaviors.

Infrastructure and Design Treatments to Improve Pedestrian Safety at Night

• Slightly less than half of the respondents indicated that they install infrastructure and design treatments to improve pedestrian safety and that they installed these treatments specifically to improve pedestrian safety at night. Among those agencies, the following infrastructure types and treatments were used:
  • – Driver-oriented street lighting,
  • – Pedestrian crossing enhancements that increase pedestrian visibility (e.g., advance yield lines, high-visibility crosswalk markings), and
  • – Crosswalk lighting.
• Participants representing agencies that installed infrastructure and design treatments to improve pedestrian safety at night tended to implement multiple types within their jurisdictions, although agencies appeared to focus on just a couple of treatments, rather than a multipronged suite of treatments.
• Less than one-third of the state agency participants indicated that their agency installs infrastructure and design treatments specifically to improve pedestrian safety at night. The most common treatments were street lighting and sidewalks or paths.
• In contrast, over half of the city and town respondents indicated that their agencies install infrastructure to improve pedestrian safety at night, including street lighting and pedestrian crossing enhancements that increase pedestrian visibility (e.g., advance yield lines, high-visibility crosswalk markings).

Pedestrian Safety at Night in Rural Areas

• Many of the respondents who represented state agencies, county entities, metropolitan planning organizations, regional councils of government, or other regional planning bodies indicated specific challenges with pedestrian safety at night in rural village centers, tribal lands, and/or high-speed roadway corridors used by people walking, as well as more generally along state highways that are also small-town main streets, suggesting a need for effort to improve pedestrian safety in those areas.

Other Strategies to Improve Pedestrian Safety at Night

• As might be expected based on their purview, few agencies implemented post-crash care or safe vehicle design strategies to improve pedestrian safety at night.

Macro-Level Pedestrian Fatality Analysis

Overall Trends

• Pedestrian fatalities in darkness steeply increased over time. The total number of U.S. pedestrian fatalities increased by 51% during the 2010s (from 4,302 in 2010 to 6,516 in 2020). During the same period, the proportion of pedestrian fatalities that occurred during darkness increased from 71% to 77%.

Temporal and Seasonal Factors

• Late autumn and winter months have the highest number and proportion of nighttime pedestrian fatalities. December had the greatest number of pedestrian fatalities in the study period (6,392) and the highest proportion occurring in darkness (81%), which likely relates to the fact that winter months have the most hours of darkness in the northern hemisphere.

• Friday and Saturday nights have the highest proportion of pedestrian fatalities at night. This result is highly likely to be associated with exposure since many areas have higher levels of nighttime pedestrian activity on weekends than on weekdays.
• Rainy conditions are overrepresented in nighttime pedestrian fatalities.

Crash Location Factors

• Crash locations in the roadway, away from intersections are overrepresented in pedestrian fatalities at night.
• Pedestrian fatalities in marked crosswalks are significantly less likely to be at night than pedestrian fatalities at other locations. This could indicate lower risk at marked crosswalk locations at night and/or lower pedestrian exposure at marked crosswalks at night.

Roadway Factors

• Higher posted speed limits are associated with significantly higher proportions of pedestrian fatalities at night.
• More roadway lanes are positively associated with pedestrian fatalities at night.
• Arterial roadways are significantly overrepresented in pedestrian fatalities in darkness, even after controlling for posted speed limit and number of lanes.
• Freeways are significantly overrepresented in pedestrian fatalities in darkness.

Vehicle Factors

• Pedestrian fatalities involving passenger cars (e.g., sedans) had significantly higher odds of occurring during darkness than fatalities involving larger vehicles (e.g., large trucks, pickup trucks, vans, sport utility vehicles [SUVs]).

Movement Factors

• Pedestrian fatalities involving drivers going straight are more likely to occur at night than fatalities involving drivers turning.
• At intersections, pedestrian fatalities at night were significantly more likely to involve pedestrians crossing the street than walking along. When the pedestrian was killed while walking along the street, the driver was more likely to have been traveling in the same direction prior to the crash than in the opposite direction.

Behavioral Factors

• Driver and pedestrian drinking are overrepresented in nighttime pedestrian fatalities.
• Pedestrian drug use is overrepresented in pedestrian fatalities at night.
• Hit-and-run pedestrian fatalities are overrepresented during darkness.

Demographic Factors

• Fatalities of child pedestrians and older adult pedestrians are significantly less likely to occur during darkness. This finding is likely significantly related to exposure.
• Pedestrians who are Black or Pacific Islander (most commonly Native Hawaiian) are significantly more likely to be killed at night than pedestrians who are White. These disparities may reflect differences in exposure, but also likely reflect differences in roadway infrastructure and traffic conditions near where people live and work.

• Older drivers are associated with fewer nighttime pedestrian fatalities. This result is likely related at least in part to reduced exposure for older drivers.
• Male pedestrians and male drivers are associated with more nighttime pedestrian fatalities.
• Wheelchair users and other pedestrians with disabilities are an important group to study for improving safety during darkness.

Micro-Level Case-Control Analysis

Overall

• Our findings indicate a gap in protection for pedestrians crossing the street on larger, faster roadways. These gaps appear to be even larger for Black and Hispanic/Latino neighborhoods. Safe crossings near known attractors, as well as the appropriate spacing for those crossings, would broadly improve pedestrian safety, including in darkness, and help create a transportation system that better protects all users.

Roadway Design and Operations Factors

• At a bivariate level, we found significant and positive associations between case-control status and:
  • – posted speed limit (cases more likely to occur at a posted speed of 35 mph),
  • – maximum number of through lanes in one direction (cases less likely to have only one lane in each direction),
  • – maximum number of turn lanes in one direction (cases more likely to have at least one turn lane), and
  • – direction of travel (cases more likely to have two-way travel).
• The maximum number of through lanes remained significant after controlling for land use, broader network measures of walkability, commute behavior, and neighborhood sociodemographics in a multivariate model, underscoring the importance of providing safe crossings along multilane roadways.
• Fewer than 10 percent of sites had any pedestrian-specific countermeasure other than sidewalks, reflecting a general lack of attention to pedestrian crossing needs on even the highest-risk streets in our case cities.
• The presence of a complete sidewalk on both sides was positively associated with case status, likely reflecting the broader land use and roadway conditions than any association with the sidewalk itself.
• Street lighting was more prevalent than other countermeasures but had no association with case-control status for midblock locations and was positively associated with being a case at intersections – likely reflecting the fact that cases were also more likely to occur near commercial land uses and along larger roadways, where intersection lighting may be more common.
• Multimodal network density variable from the EPA’s Smart Location Database was negatively associated with being a case in our model, which may capture some aspect of overall neighborhood multimodal design.

Land Use Factors

• Cases were highly significantly associated with certain land uses, particularly convenience stores and grocery stores, and significantly associated with the presence of a liquor store and general low-density commercial design in both bivariate and multivariate analyses.
• The National Walkability Index score was also significant in the model, likely reflecting the importance of land use to case-control status.
• Low-density residential areas were significantly less likely to be case locations in the multivariate model, likely reflecting roadway design, lower pedestrian demand, and different driver behavior in these areas.

Findings Related to Sociodemographic Factors

• Higher percentages of residents who identified as Black or Hispanic/Latino were positively associated with case status, even after controlling for land use and roadway design characteristics.

Driver Simulation

Speed Limit

• A majority of participants (65%) indicated that they typically drive above the posted speed limit in real life. This was reflected in the simulation’s baseline scenario where the posted speed limit was 25 mph, and the average velocity of participants was approximately 29 mph.
• Higher speeds were associated with higher average stress and lower average total fixation duration (TFD, the amount of time the driver spends looking at something) on a pedestrian or a rectangular rapid flashing beacon (RRFB). In contrast, the lower speed limit allows for a greater mean TFD at unmarked locations in dark conditions.
• Nine of the ten observed crashes in the simulator – where the participant was involved in a conflict with the crossing pedestrian – occurred when the speed limit was posted at 40 mph. This correlates with most participants (79%) indicating that they found it more difficult to recognize a pedestrian when the speed limit was 40 mph.

Lighting Conditions

• Eighty-six percent (86%) of participants indicated in the post-driving survey that additional roadway lighting helped with visibility.
• In the simulator, participants slowed down closer to the crosswalk in darker scenarios compared to brighter scenarios.
• Level of stress was higher in the darker roadway lighting condition.
• Lighting was associated with increased TFD for the pedestrian and the crossing, but lower TFD for the RRFB – perhaps because the RRFB is more apparent in the darker roadway lighting conditions.

Pedestrian Appearance

• On average, participants focused on the pedestrian with skin pigmentation VI (darker) 0.26 seconds longer than the pedestrian with skin pigmentation II (lighter), potentially because the darker-skinned pedestrian was harder to identify.

Midblock Crossing Treatment

• Most participants reported that high-visibility markings (72%) and an RRFB (95%) helped them detect the pedestrian crossing in the simulator.
• Additionally, participants yielded earlier for midblock crossings with high-visibility markings and an RRFB than unmarked crossings for both posted speed limits (25 mph & 40 mph).
• Unmarked crosswalks were also associated with a higher average velocity, while RRFB crossings had a lower average velocity.
• Participants also tended to have a higher level of stress at unmarked crosswalks compared to the other two design treatments.
• Additionally, participants had the highest TFD for pedestrians at unmarked crossings, followed by high-visibility crosswalks and then RRFB crossings. This may reflect the greater need to look for pedestrians where infrastructure does not clearly indicate their potential presence.
• Of the ten simulator crashes, 70% occurred at unmarked crossings and 30% of crashes occurred at high-visibility crosswalks; there were no crashes on segments with RRFBs.

Focus Groups

Overall

• Focus group participants shared an overarching belief that they could not trust other roadway users’ behavior. While walking, they did not expect drivers to see them, obey signals, or drive at the posted speed. While driving, they did not trust pedestrians to cross at intersections or crosswalks, look for vehicles before stepping onto the road, or be intentionally visible by using lights or reflective gear. Participants cited this lack of trust as the justification for breaking the rules when they deemed it necessary for their own safety.

Transportation Safety at Night

• The majority of participants noted that walking at night felt less safe than walking during the day due to a variety of factors, with visibility and reckless driving being the most frequently cited responses.
• People felt less visible walking at night because of the lack of sunlight in addition to the irregular presence or brightness of streetlights. Participants felt that this lack of lighting made them less visible to drivers and more susceptible to being involved in a crash. Darkness was considered even less safe when it was raining.
• Participants believed that driver behaviors such as speeding, driving under the influence, and being on their phone were more likely to happen at night and might contribute to crashes with pedestrians.

Personal Safety

• Safety from transportation-related crashes is only one component of the feeling of safety for pedestrians. Participants, especially women and, even more so, women of color, noted how they felt less safe as a pedestrian at night because of the risk of crime or assault. In contrast, nearly all of the men said that they felt nearly as safe during the night as during the day with respect to personal safety.
• The risks of personal safety seemed to impact people’s decisions regarding where and if to walk much more than transportation safety factors, especially for women.

Crossing the Street at Night

• Most participants said that they use the same general process to cross the street during the day and night, often looking for traffic in both directions and crossing when there was a gap in vehicle traffic.
• Participants said they generally behave as if drivers do not see them at night. Some participants said that they cross at signals or crosswalks in "risky" areas to protect themselves from being legally "at fault" if a crash were to happen.

Impact of Signals on Crossing Behaviors

• Most participants said they were as likely to use a signal or other crossing infrastructure (e.g., crosswalk, RRFB, etc.) at night as during the day. However, there was a general lack of trust that drivers would obey signals (especially those specifically for pedestrians, like RRFBs), so participants did not believe there was a real safety benefit in using pedestrian-specific crossing infrastructure. Instead, participants said they mostly used and were willing to walk slightly out of their way to crossing infrastructure only if it would help create a gap in vehicular traffic flow.
• In the Los Angeles focus groups, many participants described past experiences of crossing at a pedestrian hybrid beacon (PHB) and having drivers fail to stop or swerve around a stopped car. Participants also reported that some PHBs in Los Angeles either did not work when they were first installed or worked erratically, further degrading the experience for all users.

• Participants wanted to feel safe and confident crossing the street. One person reported that pedestrian countdowns at full signals are generally too fast for her typical crossing speed, which makes crossing a wide road uncomfortable and stressful. Another person stated that the "pedestrian scramble" was her preferred crossing location, as she could feel confident that no vehicles would try to go through the intersection or turn during the pedestrian phase at these locations.
• Participants generally preferred a traditional traffic signal for crossing. While they noted that signals are not always safe locations, there was a belief that people knew what to do and were most likely to behave correctly at a traffic signal compared to other types of crossing infrastructure.

Visibility

Visibility from the Pedestrian’s Perspective

• Participants’ feelings of safety related to transportation and personal safety in the dark were closely tied to streetlights and how well they illuminated the streets and, in turn, pedestrians. In areas with less street lighting, people said that they were more cautious but that they did not otherwise change their behavior.
• Many participants believed that it would be safer to wear more conspicuous gear, but that it was not practical to expect them to wear it regularly. The participants who occasionally wore reflective or brightly colored gear did so while out on recreational walks for exercise or while walking dogs.
• Some participants noted that they sometimes carry a flashlight or use their phone flashlight when walking at night to increase their conspicuity.

Visibility from the Driver’s Perspective

• Many participants said that it is harder to see at night, and several mentioned using glasses for night driving.
• Participants discussed that while they make efforts to look for pedestrians when driving at night, they often have other priorities for their focus, such as potholes and lane lines and markings, all of which are harder to see at night and represent the opportunity to damage the vehicle or hurt the driver.
• A few participants who drove smaller vehicles mentioned that when larger vehicles’ headlights shine directly into their line of vision, they make everything harder to see.
• Streetlights were also mentioned as important for driver vision, and this was particularly noticeable when the lights are not maintained or are blocked by trees.

Driving in Pedestrian-Heavy Areas

• The participants named areas in their communities where pedestrians were expected, such as schools, areas with restaurants or bars, college campuses, and transit stops. In general, they said that they try to be more cautious and aware and to slow down when they expect pedestrians to be present, particularly if they expected pedestrians to be distracted or intoxicated. Some participants said that they actively avoided driving in areas when they expected a lot of pedestrians to be present.

Distracted Driving

• Nearly all focus group participants said that they drove with their phone in the vehicle, although none of the participants believed that their driving ability was affected by the light of the screen. However, participants did admit to being distracted by interactions with the screen(s), including texting in the car, changing music, or entering a new destination into the map.

Speed Limit Reductions

• Most participants said that speed limits have little to do with how fast they actually drive, stating that they usually drive 5-10 mph over the speed limit and that more would have to happen in tandem with the speed limit reduction to have any impact.
• Most people said that changes in roadway design would be necessary to encourage slower driving, with many citing countermeasure installations that have slowed speed on nearby streets. Automated enforcement was also named as an effective countermeasure.

Practitioner Interviews

The interview findings suggested that additional research and guidance would be helpful to understand and address pedestrian safety in darkness. The following key findings support our research and inform our guidance approach:

• Despite challenges with data and crash analysis, agencies are identifying pedestrian safety in darkness as a critical problem and are directing attention and resources to addressing the problem.
• Agencies confirmed their high-risk locations (midblock crossings and arterials) and provided information on the types of countermeasures used at these locations. At the same time, the findings revealed that most of the treatments do not specifically aim to address pedestrian safety in darkness.
• Speed is a critical concern; however, specific nighttime speeding interventions have not been employed by the participants’ agencies.
• Impairment is a prevalent behavioral factor agencies are working to address. Related land uses such as liquor stores were identified as focus areas among some participants.
• Street lighting is the primary countermeasure used to improve safety in darkness, which aligns with the prevalence of street lighting found in the micro-level analysis and the feedback from driver simulator participants and focus groups. However, participants expressed concern that the lighting information in crash data may inadequately capture nuance, such as lighting on only one side of the street, which could impact analysis.
• Agencies often combine countermeasures, especially at RRFBs and PHBs. The simulator results suggest that RRFBs combined with high-visibility markings improve attention toward pedestrians and crosswalks. RRFBs are particularly impactful in dark conditions, where high-visibility flashing lights may be necessary to draw a driver’s attention to a crossing pedestrian.
• Few participants discussed high-visibility markings as countermeasures for safety in darkness, although most used them. The simulator results suggest that these treatments improve pedestrian detection and yielding behavior.
• Crashes in darkness have been identified as a problem, but analysis cannot fully examine the problem given current data collection practices. Lighting condition could be a factor in systemic analysis, but this requires more consistent tracking of lighting condition and quality in crash reporting and better inventory of lighting infrastructure.

Triangulated Findings Across Studies

Our Phase II research corroborated findings from Phase I and provided insights and new knowledge to help fill some research gaps, providing a strong foundation for the Phase III guidance. Key findings across the multiple research studies follow.

Higher roadway speed and related design and operations elements create consistent pedestrian risk

• The macro-level analysis found that higher posted speeds and multilane roadways were significantly associated with pedestrian fatalities, corroborating findings from the literature review. In addition to these roadway design and operations factors, designation as an arterial or freeway was significantly related to pedestrian risk. The micro-level analysis and driver simulation study supported these findings, showing increased risk on multilane roadways and at higher speeds.
  • – Additionally, the focus group participants openly discussed driving above the speed limit stating that changes to roadway design would generally be required for them to change their behavior in the absence of speed enforcement.

Darkness is a critical risk factor for pedestrian safety

• The macro-level analysis, driver simulation, and focus groups underscored that darkness itself is a serious risk factor for pedestrian safety, corroborating findings from the literature review. Given that the average human receives 90 percent of their information visually, darkness significantly negatively impacts the ability to detect and react to stimuli in time.
  • – Importantly, darkness is a less serious limitation at slower speeds, when any delay in detection equates to less roadway covered and the car needs less space to come to a stop.
• Lighting alone will not solve the pedestrian safety problem at night. The data from the macro-level analysis are clear that pedestrians are killed in both dark, lit conditions and dark, unlit conditions. Efforts to enhance visibility and reduce crash likelihood through lighting combined with efforts to reduce crash likelihood and injury severity potential by slowing driver speeds can improve pedestrian safety in darkness.

Countermeasures are an important part of the solution

• In the simulation experiment, drivers identified the crosswalk and slowed more quickly when a RRFB or high-visibility crosswalk markings were present.
• However, findings from the focus groups underscore the importance of education and engagement campaigns to help drivers and pedestrians understand how to effectively use and trust these recommended countermeasures.

Understanding the context is key to reducing pedestrian exposure to risk

• The micro-level analysis highlighted the significant positive correlation between the likelihood of a location having experienced a fatal or severe pedestrian crash in darkness and land uses that attract pedestrian traffic, such as convenience stores, liquor stores, grocery stores, and low-density commercial uses.

Nighttime behaviors exacerbate risk by increasing the likelihood of a crash

• The macro-level analysis found several behavioral factors associated with pedestrian fatalities at night, including driver and pedestrian intoxication, pedestrian drug use, driver hit-and-run (associated with intoxication or other illegal behavior), distraction, and speed. These findings corroborate what we found in the literature review.

Traffic safety and demographic groups

• Our macro-level analysis found that Black, American Indian/Alaska Native, Pacific Islander, and Hispanic/Latino pedestrians were significantly more likely to have been killed in darkness than in daylight compared to White pedestrians who are not Hispanic/Latino. These findings corroborate what we found in the literature review.
• The focus groups revealed that the women in the groups, and particularly the women of color, were far more concerned about their safety in darkness than the men in the groups.
• These findings corroborate what we found in the literature review about the risk found in Black, Hispanic/Latino, Pacific Islander, and Native American communities.