CHAPTER 1

Background

1.1 Introduction

Pedestrian fatalities in the U.S. have increased dramatically since 2010, with most of the rise occurring in dark conditions. In response to this increase, which the NTSB has called “a public health crisis,” NCHRP Project 17-97, “Strategies to Improve Pedestrian Safety at Night,” was funded as a multi-year, multi-phase research project aiming to investigate and understand key issues contributing to pedestrian risk at night and to develop design and operational strategies to address this risk (Shepardson 2024).

This report summarizes the research from the first two phases of NCHRP 17-97. The first phase consisted of a broad literature review of factors related to pedestrian safety at night, and a practitioner survey to understand the state of the practice for recognizing and addressing pedestrian safety at night. Phase I served to establish a baseline of knowledge regarding pedestrian safety and risk in darkness, as well as to highlight gaps in our understanding that could begin to be addressed through Phase II. Phase II consisted of a series of research studies that built on the Phase I findings, including a national pedestrian fatality analysis, detailed micro-level case-control analysis, driver simulation study, focus groups, and practitioner interviews. Each of these studies aimed to contribute additional knowledge to support future guidance. This report provides an overview of the key findings from Phase I and covers the methodologies and findings from the Phase II research to articulate and then help fill knowledge gaps. The report concludes with an overview of the Phase III tasks.

This report is organized in the following way:

• The Background Chapter (this chapter) includes the:
  • – Literature Review describing key findings from the comprehensive review conducted in Phase I and sets the stage for the remainder of the research.
  • – State of the Practice Survey providing an overview of the survey methodology and highlights from the results that guided the Phase II research.
• Chapter 2 covers the Data Collection and Analysis which includes:
  • – Macro-Level Analysis describing findings from the analysis of nationwide trends in pedestrian fatality data from 2010-2020. This analysis sought to understand both overall trends and changes over time.
  • – Micro-Level Analysis describing findings from a multicity, case-control analysis investigating characteristics of severe pedestrian injuries in darkness. This analysis sought to explore patterns among high-risk locations that could help further explain severe crash occurrence.
  • – Driver Simulation describing the experiment examining driving behavior under varying conditions of roadway design, lighting, and speed. This study sought to explore how driver behavior and stress levels change under varying conditions and in response to various types of stimuli.
  • – Focus Groups presenting the results of qualitative work exploring decision-making and behavior while walking and driving at night. This portion of the research sought to provide additional context for and human perspective about many of the findings from the prior analyses.

• • – Practitioner Interviews give an overview of the key findings from interviews with practitioners at various levels of government who are working to address pedestrian safety at night through their agencies.
• Chapter 3 synthesizes the findings from the various Phase II studies and sets up the guidebook contents.
• Chapter 4 covers the Conclusions and Suggested Research presenting next steps and further research.
• Appendices A-C contain additional data and analysis related to the macro-level analysis.
• Appendix D contains the scripts for the driver and pedestrian focus groups.
• Appendix E contains the script for the practitioner interviews.
• Appendix F contains the infographics to be used to share the results of this research project.

1.2 Literature Review

The literature review completed during Phase I summarized the methods and findings from peer-reviewed journal articles, transportation agency reports, safety organization reports, and other references in a variety of categories, including those emphasized in the Federal Highway Administration’s Safe System Approach (SSA) (2020). The review included a detailed examination of each study, including authors, year, study location, methods, sample sizes, SSA factors associated with pedestrian crashes at night, effectiveness of specific strategies to reduce pedestrian crashes at night, and special categories of findings (e.g., people with disabilities/wheelchair users, members of racial and ethnic minority groups, visibility for pedestrians who are shorter, unhoused populations, pedestrians in rural areas, pedestrians on freeways). While the detailed literature review primarily focused on research studies and reports, there are several citations from important reference books, regulations, and professional guidelines that are relevant to pedestrian safety at night. All of these resources provide useful information about the broader professional context around the studies covered in this document.

Methods

The research team used a multi-step process to gather and review references related to pedestrian safety at night.

1. The research team and panel members provided 48 studies that they knew about. Second, we searched the Transportation Research International Documentation (TRID) Database (i.e., TRB’s Transportation Research Information Services (TRIS) Database and OECD’s Joint Transport Research Centre’s International Transport Research Documentation (ITRD) Database). During the initial search of the TRID database, we used combinations of the following terms:
   1. pedestrian, walk, walking
   2. crash, safety, risk
   3. night, nighttime, dark, darkness, light, lighting
2. After generating an initial list of references, the team identified studies to review in more detail in three stages. The first stage sorted by title and removed duplicate records, resulting in a list of 1,412 references. Because 21 of the 48 references that had been identified by team members were included in this list, our final total was 1,412 + 27 = 1,439 references. Most of the 27 articles provided by team members (but not captured in the original search) provided valuable contextual information about pedestrian safety (e.g., perceptions of safety and other challenges in lower-income communities and communities of color) or insights into relationships between specific behaviors and general traffic safety outcomes (e.g., driving while intoxicated, distracted driving, ride-hailing).

3. The team read the title of each remaining record and skimmed the abstract of a small sample of references. Based on this preliminary classification, we identified 249 references for further review and excluded 1,190 references.
4. The team read the abstracts of the 249 selected references and entered information from each abstract into a literature review spreadsheet.
5. The team then skimmed the full text of the 249 remaining references. For studies that were directly related to at least one of the five SSA categories, covered at least one special category of interest (listed above), or examined the effectiveness of specific strategies to reduce pedestrian crashes at night, we read the full text and entered details about study into our literature review spreadsheet.
6. As we reviewed our preliminary findings from the literature, we identified several issues that would benefit from an even more targeted literature search (e.g., increases in pedestrian fatalities at night over time; nighttime speed limits). We added 12 more references as a part of this targeted search. Adding these studies meant that our literature review spreadsheet included a grand total of 261 references. Because of saturation of certain topics and the lack of robust results from certain studies (mentioned above), we cite 150 of these references in this document.

Key Findings

After we identified a list of studies to read, we categorized the main findings from each study according to the five main categories of the SSA (Table 1). We noted key themes within each of these categories, as suggested by the right-hand column examples. In addition to viewing these five categories independently, we looked for interactions between them. For example, if a pedestrian is crossing a wide roadway (“safe roadways” category) from an approaching driver’s left (safe road users), they may be backlit by vehicle headlights (“safe vehicles” category) traveling in the opposite direction, especially if the overhead roadway lighting (“safe roadways” category) is poor.

Table 1: Five Categories of Safe System Approach for Literature Review

Most studies include nighttime or darkness as one of a large set of potential variables associated with pedestrian crash or injury outcomes rather than the primary variable of interest. Therefore, researchers currently have a limited understanding of how different design and behavior variables interact to influence nighttime pedestrian risk. The literature on pedestrian safety at night provides many insights that will be useful for the following stages of this research project and future research. Knowledge from decades of research on this topic may also be applied directly to practical transportation safety policies and strategies.

Pedestrian Risk at Night: Magnitude and Trends

This first section of findings summarizes what is known about several basic questions related to pedestrian safety at night: What proportion of pedestrian crashes occur at night, and how does this proportion vary by day of week, season, and injury severity level? Why have pedestrian fatalities at night increased in the United States in recent years?

Pedestrian Crash Prevalence and Risk at Night Versus Daytime

Several studies have quantified differences in pedestrian crash prevalence at night versus daytime. In 2019, 76% of U.S. pedestrian fatalities occurred in darkness, and another 4% occurred during dawn and dusk (NHTSA 2021). By comparison, 70% of pedestrian fatalities occur at night in Japan (Hiratsuka et al. 2016), and 57% occur between 4 pm and 4 am in the European Union (Costa et al. 2020). Compared to other roadway system users, pedestrians are particularly vulnerable to the additional crash risk created by darkness. Sullivan and Flannagan (2001) argue that lighting countermeasures targeted toward pedestrian visibility would save nearly twice as many lives as countermeasures targeting collisions with other motor vehicles.

Compared to other roadway system users, pedestrians are particularly vulnerable to the additional crash risk created by darkness. Sullivan and Flannagan (2001) argue that lighting improvements focused on pedestrian visibility would save almost twice as many lives as countermeasures targeting collisions with other motor vehicles. The same authors show that single-vehicle run-off-road crashes are impacted much less significantly by daylight saving time (DST) transitions than pedestrian crashes (Sullivan and Flannagan 2002a).

Current Understanding of the Increase in United States Pedestrian Fatalities at Night

The proportion of pedestrian fatalities occurring during darkness in the United States has increased in recent decades (Schneider 2020; Sandt et al. 2020). Pedestrian fatalities during darkness increased from 63% during 1977-1981 to 73% during 2012-2016 (Schneider 2020). During the 2010s, total US pedestrian fatalities increased by approximately 46%, from 4,302 in 2010 to 6,272 in 2019. Over this period, most of the additional fatalities occurred during darkness (Hu and Cicchino 2018; Retting 2020; Ferenchak and Abadi 2021; Tefft, Arnold, and Horrey 2021). There is speculation about why pedestrian fatalities have increased at night. One possibility centers on the changing demographics of pedestrians. At least some of the shift toward darkness could reflect a broad societal shift in pedestrian travel from children toward adults, likely meaning that a greater proportion of overall pedestrian activity is occurring at night when children are likely to be indoors. Taking a multi-decade perspective, Schneider (2020) showed that pedestrian fatalities simultaneously shifted from daytime toward nighttime and from children toward adults aged 18 to 64. Similarly, Tefft, Arnold, and Horrey (2021) found that adults accounted for all of the additional pedestrian fatalities between 2009 and 2018, while child pedestrian fatalities actually decreased during this period.

Roadway and urban context factors may have also led to the additional number of pedestrian fatalities at night in the last decade. The increase in overall pedestrian fatalities has been connected with urban areas (Hu and Cicchino 2018; Ferenchak and Abadi 2021; Tefft, Arnold, and Horrey 2021), urban arterial roadways (Hu and Cicchino 2018; Sandt et al. 2020; Ferenchak and Abadi 2021; Tefft, Arnold, and Horrey 2021), multilane roadways (Ferenchak and Abadi 2021); 40-45 mph roadways (Ferenchak and Abadi 2021) and non-intersection locations (Ferenchak and Abadi 2021; Tefft, Arnold, and Horrey 2021). The total extent of locations with these characteristics may have increased some during each study period, but the increase in fatalities may have also been due to increases in pedestrian and motor vehicle activity in these types of locations during each study period.

Factors Associated with Pedestrian Risk at Night

Speed and lighting are two central factors influencing pedestrian crash risk at night, but they are impacted by and intertwined with many other factors, like roadway design, pedestrian and driver behaviors, and vehicle characteristics. Figure 1 provides a conceptual representation of the main factors associated with pedestrian crash risk at night. The categories of the SSA are featured prominently as blue boxes in the diagram. The boxes use font size and color to indicate the order within a pathway to impacting the pedestrian crash risk at night (larger font and darker color correspond to earlier in the pathway, whereas smaller font size and lighter color are later in the pathway). For example, the design of a roadway sets the stage for the driver and is the first line of “defense” against pedestrian crash risk. Similarly, a safe vehicle can help mitigate risk in the event that a driver does not see or react to a pedestrian in time. Within the boxes, bold text is used to highlight factors for which previous research has identified links with pedestrian crash risk at night. Other factors have theoretical connections with nighttime pedestrian risk, but there is limited or no evidence of these relationships from the literature. Figure 1 communicates the interrelated nature of various factors associated with pedestrian risk during darkness and also provides a theoretical roadmap for readers as they consider the findings in each subsection below.

Safe Roads

Roadway factors associated with pedestrian risk at night include design characteristics (e.g., width, number of lanes), location type (e.g., intersection versus mid-block), lighting characteristics, and pedestrian crossing facilities.

Roadway Design

An analysis of United States and California pedestrian fatalities found that many of the same high-speed, multilane roadway designs that create high-risk environments for pedestrians during the day are associated with even higher pedestrian risk at night (Sanders, Schneider, and Proulx 2022). Sullivan and Flannagan (2002) explored pedestrian fatalities before and after DST transitions and found similar results: pedestrian fatalities were nearly seven times as likely in darkness as in daylight on limited access roadways, five times as likely on arterials, and three times as likely on local roads. Arterial roadways are called out as high-risk in several other studies (Sullivan and Flannagan 2001; Ferenchak and Abadi 2021; Long and Ferenchak 2021). More than 80% of the increase in United States pedestrian fatalities at night between 2002-2009 and 2010-2017 was on arterial roadways (Ferenchak and Abadi 2021; Tefft, Arnold, and Horrey 2021). Arterial roadways are also associated with high concentrations of pedestrian fatalities and severe injuries at night (Guerra et al. 2020; Long and Ferenchak 2021). Similarly, roadways with more lanes often have high numbers of pedestrian crashes and fatalities at night (Nabavi Niaki et al. 2016; Ferenchak and Abadi 2021; Alogaili and Mannering 2022; Sanders, Schneider, and Proulx 2022). Wider lane widths (Fitzpatrick and Park 2021) and lack of sidewalks may also increase pedestrian risk at night (Long and Ferenchak 2021). In contrast, at intersection locations, traffic signals may decrease pedestrian risk at night (Sanders, Schneider, and Proulx 2022), though more research could help to clarify the effect of signalization on nighttime pedestrian safety along the length of a roadway corridor.

Location type

Previous studies do not show definitive results about nighttime pedestrian crash risk at intersections versus mid-block locations. Ferenchak and Abadi (2021) found that more than 80% of the additional United States pedestrian fatalities at night between 2002-2009 and 2010-2017 occurred at non-intersection unmarked locations. Gu and Peng (2021) found that longer roadway segments in Miami-Dade County, FL are associated with higher numbers of pedestrian crashes at night, which may point toward greater risk at mid-block crossings. Vehicles often travel at higher speeds mid-block and drivers may be less likely to expect and be alert to pedestrian crossings at these locations. After controlling for other temporal, built environment, socioeconomic, and behavioral characteristics, mid-block locations had a higher pedestrian fatality risk at night than intersection locations in California, but this result was only moderately significant and was not significant across the United States (Sanders, Schneider, and Proulx 2022). Examining Florida data, Siddiqui, Chu, and Guttenplan (2006) found that mid-block locations had higher percentages of pedestrian fatalities than intersections during both daylight and dark conditions, but it was unclear if the mid-block risk is actually higher at night.

Lighting

Roadway lighting has been identified as a key roadway design component for reducing pedestrian crash risk at night (Pocus and Katz 1978; Bush 1985; Kim et al. 2010; Haleem, Alluri, and Gan 2015; Kemnitzer et al. 2019; Hennessy and Ai 2021). Compared to daytime, nighttime pedestrian fatalities are two to five times more common, depending on whether streetlights are present (Kim et al. 2010). Broadly, research shows that even small differences in lighting impact pedestrian safety. To illustrate this point, Sivak, Schoettle, and Tsimhoni (2007) found that relatively darker nights with a new moon are associated with 22% more pedestrian fatalities than relatively lighter nights with a full moon. González-Hernández et al. (2020) show a positive relationship between decreasing illumination levels after sunset and decreasing driver yielding rates.

Pedestrian crossing facilities

Pedestrian crossing facilities can make crosswalks more visible and alert drivers to the presence of pedestrians in crosswalks. Several studies have focused specifically on the effectiveness of these facilities at night. We identified one unpublished conference paper that directly evaluated pedestrian risk at marked crosswalks alone, without any other complementary treatments, during nighttime versus daytime (Yuan and Dulaski 2017). This study of crosswalk locations in the Boston, MA, area found that drivers approached three marked crosswalks at a higher speed and yielded less often to staged pedestrians at night (45% day vs. 23% night yield rate). It also found decreased yielding at night at three marked crosswalks with a crosswalk warning sign (77% day vs. 33% night yield rate). Current United States pedestrian crossing safety guidance suggests that marked crosswalks alone may generally be effective on roadways with low speeds and low traffic volumes, but they should be complemented by other treatments on higher-speed, higher-volume, multilane streets (Zegeer et al. 2005; Blackburn, Zegeer, and Brookshire 2018). Yet, these marked crosswalk guidelines have not been evaluated specifically for nighttime conditions. A lack of nighttime versus daytime exposure data may be a core reason for the lack of research on marked crosswalk safety at night.

Safe Vehicles

In addition to roadway factors, vehicle design may contribute to pedestrian risk at night. The mix of vehicle types on the road, as well as their design and technology, has evolved significantly in recent decades. There have also been several studies and proposals for new vehicle technologies designed specifically to improve safety for pedestrians at night.

Headlights

Headlight design and technology are key vehicle safety considerations. Pedestrian safety at night is related to the ability of drivers to visually identify pedestrians in or along the roadway. Generally, brighter illumination improves pedestrian visibility as well as subjective perceptions of their conspicuity. Tyrrell, Wood, and Carberry (2004) found that the use of high beams increased pedestrians’ estimates of their own conspicuity by approximately 69% and that there was a linear relationship between brightness and perceived conspicuity.

The type of lighting technology used can also have significant pedestrian safety implications. High-intensity discharge (HID) headlights increase the level of peripheral illumination, which is important for drivers identifying pedestrians along the edges of the roadway (Bullough and Skinner 2012). In some situations, conventional low beam headlights may actually reduce pedestrian conspicuity. Several adaptive headlighting technologies and other innovative vehicle lighting designs have been studied to determine if they can provide a safety benefit, particularly in locations where pedestrians are not anticipated or where roadway lighting is inadequate. Adaptive headlighting technologies have also been found to have benefits for vehicles traveling on straight roads, which account for a significant number of pedestrian fatalities. Motorway lighting, a headlight configuration designed to optimize visibility on roads with speed limits greater than 45mph by projecting more light, further ahead of the vehicle, could yield a potential 90% reduction in pedestrian fatalities and 72% reduction in total pedestrian crashes in darkness (Sullivan and Flannagan 2007). The type of vehicle lighting also impacts the way that pedestrians perceive vehicles and, in turn, the safety decisions that they make. Conventional headlight designs can make judging the vehicle type, distance, and approach speed difficult for pedestrians in darkness (Balasubramanian and Bhardwaj 2018). Although daytime running lights (DRL) are not intended for nighttime use, the light color of DRLs can have a significant impact on pedestrian perception of vehicle turn indicators (Peña-García et al. 2010).

Interior lighting

Interior vehicle lighting is another factor which can have significant impacts on drivers’ abilities to detect pedestrians in darkness. Studies as early as the 1980s have identified that turning on the interior lights in a vehicle can reduce forward sight distance by as much as 20% and stressed the need for lighting systems that minimize this degradation (Olson 1985). Today, this remains an important consideration as recent trends have generally increased the number of vehicle cabin light sources including entertainment systems, driver assistance displays, portable electronic devices, and ambient cabin lighting designed to enhance in-vehicle atmosphere (Isele, Neumann, and Blankenbach 2018). Despite the understanding of the potential for interior lighting to impact visual performance and the rapid changes in lighting technology, there has been relatively little research on how these factors interact. Overall, the effect of vehicle interior lighting is an important area for future research. In particular, it is important to study drivers’ abilities to visually adapt to darkness immediately after looking at a bright screen.

Automated Pedestrian Detection Systems

Automated pedestrian detection systems integrated into vehicles can be used both to trigger adaptive lighting and to alert drivers to the presence of pedestrians. These technologies have promise: The Highway Loss Data Institute (2017) found that Subarus equipped with automated pedestrian braking systems had a 35% lower rate of insurance claims for pedestrian crashes than the same vehicle models without the system. The technology needed to accurately identify pedestrian changes under different conditions. Multispectral pedestrian detection, using both visible and infrared images, has been proposed as an efficient and effective method for low-light conditions (Hamdi et al. 2020).

When a pedestrian is detected, the mechanism by which the driver is notified also affects driver detection and braking distances. A lighted pedestrian warning indicator on the vehicle dash was found to decrease the ratio of missed pedestrians from 13% to 5% and improve mean detection distance from 34m to 44m (Tsimhoni et al. 2007). A study of six different automatic warning configurations with older drivers aged 65-74 found that tactile and audio warnings performed the best while visual warnings performed the worst (Brown et al. 2010). The addition of tactile and auditory warnings caused drivers to apply the brakes at an average of approximately 160 feet from a pedestrian compared with an average of 80 feet with a visual-only warning system. This research also suggests that the effectiveness of warning systems may vary for different driver characteristics (Brown et al. 2010). Further study on different combinations of warning systems and types of drivers would help inform an effective approach to alerting drivers to the presence of pedestrians.

Size and design

Our literature search uncovered few studies that examined the relationship between vehicle size and design and pedestrian safety specifically during darkness. Sanders, Schneider, and Proulx (2022) found that smaller vehicles were more likely to be involved in pedestrian fatalities that occurred during darkness versus daylight. The authors mentioned the trend of more SUVs and large trucks being involved in pedestrian fatalities overall, so their finding may be due to differences in exposure. Ferenchak and Abadi (2021) found that the prevalence of SUV involvement in nighttime crashes was increasing faster than their share of the vehicle fleet. Further research could examine if the percentage of SUVs and large trucks involved in pedestrian fatalities during the nighttime is higher than during the daytime.

While little previous research has focused directly on how vehicle characteristics are related to pedestrian safety during darkness, several studies have revealed a trend toward larger vehicles which are increasingly involved in pedestrian fatalities in both day and night conditions. The proportion of pedestrian fatalities involving large vehicles (e.g., pickup trucks, vans, SUVs) increased from 22% to 44% between 1977 and 2016 (Schneider 2020).

The fatality risk from large vehicles is compounded by multiple factors related to their design which are direct determinants of injury severity. High-profile vehicles increase the impact height on a pedestrian and have longer stopping distances, resulting in a higher impact speed (Siddiqui, Chu, and Guttenplan 2006). An analysis of vehicle designs and pedestrian frontal impact points determined that lower vehicle front ends (e.g., having lower bonnet leading edges) transfer less impact energy to the upper leg of a pedestrian (Brown, 2001).

However, large vehicles alone are not responsible for increasing pedestrian fatalities. From 2009-2016, vehicle horsepower and power-to-weight ratio increased annually across passenger vehicle classes. Higher power and power-to-weight ratios are typically associated with vehicles having the capability to go fast and drivers using them to exceed the speed limit (Hu and Cicchino 2018). Data from North Carolina also shows that motorcycles are among the vehicles most likely to result in a pedestrian fatality in the event of a crash (Chen and Fan 2019).

Safe Speeds

The positive relationships between vehicle speed and pedestrian crash risk (Sullivan and Flannagan 2007; Guerra et al. 2020) and pedestrian injury severity level (Rosén, Stigson, and Sander 2011; Tefft 2013) has been well established in the literature. However, less is known about how vehicle speeds interact with darkness to impact pedestrian safety outcomes (Pour et al. 2017). Recently, Sanders, Schneider, and Proulx (2022) found that darkness exacerbates pedestrian risk on higher-speed roadways.

Higher vehicle speeds increase the distance necessary for a vehicle to stop safely before hitting a pedestrian. For example, Zhang and Ma (2014) find that minimum safe distances are approximately 14 m at 20 km/h, 61 m at 40 km/h, and 134 m at 60 km/h. The authors emphasize that these distances increase when pedestrians are less visible to drivers (e.g., at night) (Zhang and Ma 2014).

Both vehicle and roadway design factors may affect vehicle speeds. One roadway design factor that has a close relationship with vehicle speeds is posted speed limit. Yet relatively few studies have focused on how speed limits relate specifically to pedestrian safety at night. Sanders, Schneider, and Proulx (2022) emphasize the positive association between higher posted speed limits and the risk of pedestrian fatalities and severe injuries at night. Comparing 2002-2009 with 2010-2017, Ferenchak and Abadi (2021) found a significant increase in the proportion of pedestrian fatalities at night on roadways with recorded posted speed limits of 40-45 mph (54.6% of total known increase in fatalities), but a decrease in the proportions on roadways with posted speed limits of 30-35 and 50-55 mph.

In practice, lower nighttime speed limits have been used to attempt to reduce crash risk during darkness. While night speed limits were used in the United States during parts of the 20th Century (Stein 2015), we identified one recent application of night speed limits in a U.S. city. Tucson, AZ, has night speed limits on more than 20 roadway corridors with limited street lighting (Tinsley 2019). However, we did not find any safety studies of this treatment. Night speed limits have also been tested recently on rural freeways to reduce wildlife crashes. Despite three-quarters of pedestrian fatalities occurring at night, only a few studies have suggested night speed limits as a practical pedestrian safety strategy (Schneider 2020; Sanders, Schneider, and Proulx 2022). We did not find any research that examined their potential effectiveness for pedestrian safety. Given the important influence of posted speed limits on pedestrian safety at night, this is a gap in research and practice.

Safe Road Users

Pedestrian and driver characteristics and behaviors are associated with pedestrian crash risk at night. Much of the nighttime pedestrian crash research in this area has focused on understanding drivers’ abilities to detect pedestrians and pedestrians’ abilities to detect vehicles. This section also describes driver and pedestrian pre-crash behaviors, intoxication, and demographic attributes that are overrepresented in nighttime pedestrian crashes.

Pre-crash behaviors: crash types

Pedestrian crash risk at night appears to be higher when the pedestrian approaches from the driver’s left (or the driver’s right in countries that drive on the left) (Sullivan and Flannagan 2011; Hiratsuka et al. 2016). Comparing Michigan pedestrian crashes between daylight and darkness, Sullivan and Flannagan (2011) found that nighttime pedestrian risk is higher when the pedestrian is crossing from the driver’s left side or when the driver is making a left turn. Similarly, Schneider and Stefanich (2016) found that pedestrian fatalities in Wisconsin are more likely when the pedestrian is crossing from the driver’s left (i.e., struck in the second half of the crossing). The Schneider and Stefanich (2016) result was found when examining crashes occurring at all times of day, but some of the crashes could be due to pedestrians being hidden within opposing headlight glare. Still, the underlying causes of higher risk for pedestrians approaching from the driver’s left are poorly understood. Headlights illuminating areas on the right side of the vehicle more than the left side is another possibility. Drivers might also look more to the right at night to track features on the edge of the roadway or scan the roadway less because there is a smaller visual field available when it is dark.

Drivers’ abilities to detect pedestrians at night

The sooner that drivers can detect pedestrians in the roadway or entering the roadway, the more time they have to avoid a potential crash. A key aspect of driver behavior is the fact that information is primarily obtained visually (Dewar and Olson 2016) – thus, darkness critically reduces a driver’s capacity to detect pedestrians.

At night, drivers detect pedestrians who are directly in front of them better than pedestrians who are at the side of the roadway. Detection distance is reduced by approximately one-third for pedestrians approaching vehicles from the side (Luoma and Penttinen 1998). Drivers also detect pedestrians who are facing them directly better than pedestrians who are in profile (e.g., walking perpendicular to the driver while crossing the street) (Sayer and Mefford 2004). Smaller pedestrians (e.g., children) are also more difficult for drivers to detect at night (Bullough, Rea, and Zhang 2012). As discussed in earlier sections, glare from opposing vehicle headlights also reduces the ability of drivers to detect pedestrians (Wood et al. 2012). Glare is likely to impact older drivers more than younger drivers (Dewar 2016a), and more research could help determine how glare affects pedestrian detection specifically.

Drivers tend to focus on areas where their headlights are shining while driving (Olson et al. 1989). This means that pedestrians entering the roadway—who are already hard to see because of low lighting—may also be on the periphery of drivers’ visual focus, leading to very delayed detection. This finding could suggest that risk may be worse on wider, multilane roadways, but we did not find any studies that tested drivers’ detection abilities for this type of roadway cross section. The difficulty of drivers detecting pedestrians is magnified by the optical illusion created by vehicle speed (Yu, Du, and Wang 2019), such that the faster a vehicle travels, the more drivers perceive that the vehicle is further from a pedestrian than it actually is. This perception reduces drivers’ ability to react quickly enough to avoid a pedestrian. When in motion, drivers overestimate the distance to a pedestrian even more when the pedestrian is a child (i.e., shorter), particularly when it is dark, because their distance estimation is calibrated to adult pedestrian height (Yu, Du, and Wang 2019). Further, drivers’ eyes do not adapt well to rapid, sudden changes in

Pedestrian and driver impairment

Alcohol and drug use increases pedestrian injury risk in general, and these two activities are more common during darkness. Both pedestrians and drivers are more likely to have alcohol in their system in nighttime crashes than in daytime crashes (Sullivan and Flannagan 2007; Sanders, Schneider, and Proulx 2022). Because alcohol and drugs impair a person’s ability to simultaneously engage in multiple tasks, as is required in driving, drivers are less able to respond to pedestrians in real-world scenarios (i.e., in contrast to highly curated lab experiments) (Moskowitz 2016; Dewar 2016b). Specifically, as drivers’ blood alcohol levels increase, their abilities to detect the contrast of shapes and overall detection distances decrease (Hazlett and Allen 1968; Moskowitz 2016).

Impairment also appears to affect driver and pedestrian judgement. For pedestrians, alcohol impairment is associated with a lack of awareness of the impairment, risky crossing behavior, and challenges associated with integrating information related to speed and distance quickly enough to allow for safe decision-making (Sullivan and Flannagan 2001; Oxley et al. 2006; Dewar 2016b), whereas impaired drivers may drive faster than conditions warrant (Moskowitz 2016).

The relative risk of pedestrian intoxication depends on the environment in which it occurs. Pedestrian intoxication is more dangerous when the underlying risk of the roadway itself is higher, which, for example, might be due to a higher design speed. In contrast, operating a vehicle while intoxicated is illegal and can cause harm to others in any situation.

Demographic Characteristics

A few studies have identified personal demographic characteristics that are associated with nighttime pedestrian risk. In this section, we first examine race/ethnicity and nighttime pedestrian risk. Then we explore other demographic characteristics, such as gender and age.

Several studies have found that Native American and Black pedestrians are at a higher risk of being killed in a crash than White pedestrians in general (Smart Growth America 2024; Dumbaugh et al. 2023) and specifically at night (Sanders, Schneider, and Proulx 2022). Other research also suggests elevated nighttime risk for Hispanic/Latino pedestrians (Long and Ferenchak 2021; Schneider et al. 2021). Although the disparate risk to Native American, Black, and Hispanic/Latino pedestrians is documented, there is no single causal factor. Existing literature illustrates a wide range of factors that include the design of the built environment, including multilane roadways and a lack of pedestrian infrastructure, infrastructure characteristics such as higher-speed roadways, and socioeconomic factors. The higher documented risk to these groups extends to the nighttime risk experienced by all pedestrians.

The level of investment in the built environment can contribute to both exposure and to the increased risk to Native American, Black, and Hispanic/Latino pedestrians at night. Several studies have found a significant relationship between pedestrian safety, infrastructure and walking behavior in an area, and income, with lower-income areas -- often the areas with high percentages of racial and ethnic minorities – consistently experiencing more and more severe pedestrian injuries (Dumbaugh et al. 2022; Dumbaugh et al. 2023). Lin et al. (2019) noted that low-income areas are associated with higher pedestrian fatality risk, which is due at least in part to these areas having many high-volume and high-speed arterial roadways.

Researchers have sought to determine whether drivers are more likely to yield to some people over others, thus impacting pedestrian safety. For example, Goddard, Kahn, and Adkins (2015) staged pedestrian crossings with research team members in Portland, OR. Their research showed Black male pedestrians being passed by twice as many cars and waiting 32% longer for a driver to yield at a crosswalk than White

The agencies who responded to the survey may have been more likely to be aware of pedestrian safety at night than agencies that did not respond, so, the results of the survey are not necessarily representative of the practices of all agencies in the United States. The aim of this survey was to understand how agencies are approaching pedestrian safety in darkness; therefore, regardless of the representativeness of the survey, it provides a snapshot of policies and practices used in the industry.

Risk Factors

The agencies that indicated that pedestrian safety at night was a problem in their community were asked about risk factors associated with this issue. The most common risk factors agencies associated with pedestrian risk at night were street type, crossing location, time of day, and demographics of those impacted.

The literature review identified risk factors associated with specific demographic groups, and agencies tended to agree most with the statements that “pedestrian nighttime risk is higher for people who are houseless” and “pedestrian nighttime risk is higher for Black and Brown pedestrians”; 51% and 45% of the respondents, respectively, agreed that these two statements reflected trends in their communities.

The literature review also identified risk factors associated with different roadway environments or conditions. Most agencies agreed that the statements “pedestrian nighttime risk is elevated on higher-speed roadways,” “pedestrian nighttime risk is influenced by the presence and quality of street lighting,” and “pedestrian nighttime risk is higher in lower income areas” were reflective of trends present in their communities.

Programs and Policies to Improve Pedestrian Safety at Night

Fifty-five of the 76 respondents indicated that they implement programs or policies to improve pedestrian safety. Of those 55 agencies, 20 indicated that they implemented programs and policies specifically to improve pedestrian safety at night. Among those 20 agencies, the following trends emerged (respondents could indicate multiple programs and policies):

• 14 agencies implemented a pedestrian safety messaging campaign targeted to pedestrians.
• 10 agencies implemented a pedestrian safety messaging campaign targeted to drivers, bicyclists, or other non-pedestrian road users.
• Nine agencies implemented a lighting policy.

Of those 20 agencies, 14 implemented a combination of at least two programs or policies. The combination included 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

Thirty-four respondents indicated that they install infrastructure and design treatments to improve pedestrian safety and indicated that they installed these treatments specifically to improve pedestrian safety at night. Among those agencies, the following trends emerged (respondents could indicate multiple infrastructure types and treatments):

• 29 agencies installed or improved street lighting (presumed to be vehicular lighting as pedestrian-scale lighting was a separate category).
• 18 agencies installed pedestrian crossing enhancements that increase visibility of pedestrians (e.g., advance yield lines, high-visibility crosswalk markings).
• 16 agencies installed or improved crosswalk lighting.

Agencies that install infrastructure and design treatments to improve pedestrian safety at night tend to implement multiple types within their jurisdiction: 29 agencies install at least two types of infrastructure or treatments, while 11 agencies install five or more types of infrastructure and treatments.

Thirteen state agencies installed infrastructure and design treatments specifically to improve pedestrian safety at night, among them, the two most common treatments were street lighting (85% of respondents) and sidewalks or paths (69% of respondents). Among the 20 cities or towns that make a similar effort, the two most common infrastructure or design treatments were street lighting (85% of respondents) and pedestrian crossing enhancements that increase visibility of pedestrians (e.g., advance yield lines, high-visibility crosswalk markings) (50% of respondents). Due to the breadth of the survey, questions were generally limited to an overview (e.g., “Please select the infrastructure or design projects your agency implemented to improve pedestrian safety at night.”) and did not include a follow-up question to ask about frequency or extent (e.g., “How many locations received this treatment?”).

Pedestrian Safety at Night in Rural Areas

The respondents who indicated that they represented state agencies, county entities, metropolitan planning organizations, regional councils of government, or other regional planning bodies were asked a series of questions specifically to solicit information about pedestrian safety at night in rural areas (n=40). When asked if respondents experienced any specific challenges with pedestrian safety at night in rural village centers, tribal lands, and/or high-speed roadway corridors used by people walking, 40% said “yes” (n=17), 42% said “no” (n=16), and 18% did not respond (n=7). Those same 40 respondents were asked whether they experienced specific challenges with pedestrian safety at night along state highways that are also small-town main streets. Forty-five percent of respondents said “yes” (n=18), 38% said “no” (n=15) and 18% did not respond (n=7).

Other Strategies to Improve Pedestrian Safety at Night

The survey responses indicated that few agencies implemented post-crash care or safe vehicle design strategies to improve pedestrian safety at night. Five agencies implemented post-crash care strategies and two considered implementing or implemented safer vehicle design strategies, although the design strategies appeared to reflect roadway countermeasure design rather than vehicle design.

Conclusion

This survey captured a cross section of transportation practitioners from state and local agencies across all regions of the United States. The results suggest that research gaps identified in the literature review parallel possible gaps in knowledge or experience among practitioners and underscore that nighttime pedestrian safety is an issue in most communities. General trends in the literature review findings were also in alignment with practitioners’ perspectives, including that pedestrian nighttime risk is elevated on higher-speed roadways, is influenced by the presence and quality of street lighting, and is higher in lower-income areas. Agencies also generally agreed that pedestrian nighttime risk is higher for people who are unhoused and for Black and Brown pedestrians.

Nearly half of state, county, and regional agencies were unsure whether pedestrian safety at night was an issue in rural areas. However, the survey provided a glimpse into some of the trends that agencies have noticed and the barriers agencies face when trying to implement strategies to improve pedestrian safety at night, particularly in rural areas, such as a lack of pedestrian infrastructure and unsafe road user behavior.

The survey results also provide insights into some of the programs, policies, and infrastructure treatments that state and local agencies are implementing to improve pedestrian safety at night. Among state agencies, pedestrian safety messaging campaigns were the most common program or policy implemented to improve pedestrian safety at night. Among cities and towns, lighting policies were the most common, followed by pedestrian safety messaging campaigns. Speed limit policy changes were among the least common programs or policies implemented to improve pedestrian safety at night.

In general, and particularly for cities and towns, infrastructure and design treatments were more commonly used to address pedestrian safety at night than policies and programs. Within the category of infrastructure or design treatments, street lighting was the most common treatment, followed by crossing enhancements that increase visibility of pedestrians (e.g., advance yield lines, high-visibility crosswalk markings) and crosswalk lighting. Among state agencies, installing sidewalks or paths was a slightly more common treatment than crosswalk lighting.

Most agencies indicated that they analyzed pedestrian crash data across a variety of injury severity levels on a regular basis. However, few of the agencies indicated that they measure the effectiveness of specific strategies and treatments, suggesting that more information about the effectiveness of different strategies may help guide practitioners’ decisions about how best to invest limited resources. The survey also pointed to how improved communication and policy support between state and local agencies, particularly among local agencies wanting to change roadway conditions or speed limits along state-owned roadways, would be helpful.

The survey results did not provide sufficient information to fill research gaps related to post-crash care and safer vehicles, given the general purview of transportation agencies and the corresponding reality that few have used post-crash care strategies or safer vehicle designs to improve pedestrian safety at night.

The findings from this survey and the literature review informed the analysis plan for the next phase of this project, as described in the next section.