Highway Worker Safety (2017)

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.

14 chapter two Issues In HIgHway worker safety IntroductIon This chapter provides a review of the overarching issues related to highway worker safety. The information serves as background knowledge for the concepts and terminology presented in subse- quent chapters and understanding the knowledge and concerns associated with worker safety that are required to develop effective highway worker safety management programs. The following sections are based primarily on a review of the literature regarding the topics of worker safety and work site safety in the United States. Focusing on the United States gives the information more relevancy to state DOTs that are subject to the same federal guidelines, such as OSHA or their OSHA-approved State Plan alternatives. Particular emphasis is placed on the following content areas: • Prevalence and causality of injury and fatality incidents for highway workers, • Availability of highway worker safety data, • Existing legal standards and policy recommendations related to highway workers, • Risk and human factors, • Stakeholders in highway worker safety, and • Evaluation of safety programs. These content areas highlight contemporary safety issues associated with highway workers and expose how the issues are being addressed through research and practice and by national and state government policy. Subsequent chapters review existing programs and conditions to explore further the available data and current practices in state DOTs and attempt to determine the consequences of current practices. Prevalence and causalIty of HIgHway worker IncIdences One of the most efficient ways to improve safety programs and reduce highway worker injuries and fatalities is to understand the causes and frequency of incidents associated with highway workers. Highway work sites potentially pose significant risk to motorists and workers; if the risks are not effec- tively controlled, injuries and fatalities for motorists and workers can result. Research has been con- ducted regarding specific types of work site incidents, comparing work site incidents to nonwork site incidents, and possible methods for eliminating or reducing incidents that lead to injuries and fatalities. Some state agencies have also performed research to explore the causes and characteristics of work site crashes within their state. One study investigated crashes that occurred in Kansas work zones to better understand the characteristics of the crashes (Li and Bai 2007). The characteristics included demographic data, such as age and gender, and characteristics relating to the incident, such as day of the week and environmental factors. The researchers compared these characteristics between fatal and nonfatal work zone incidents. The researchers sought to determine the characteristics that may be linked to severity. The researchers argued that this comparison could provide improved guidance regarding how and where crash reduction strategies could be implemented to maximize their effect (Li and Bai 2007). A 2009 study involving states from the Smart Work Zone Deployment Initiative (SWZDI) sought to determine the characteristics of work zone crashes in Iowa, Kansas, Missouri, Nebraska, and Wisconsin (Dissanayake and Akepati 2009). These characteristics included incident frequency infor-

15 mation based on light and weather conditions, as well as frequency information based on speed, driver behavior, and traffic control factors. The report presents the similarities and differences between work zone incidents in the five states. For these states, only slight differences between the incident characteristics were found (Dissanayake and Akepati 2009). This result is possible because all of these states are located in the Midwest region of the United States and are likely to have similar driver demographics, roadway conditions, and physical geography. An Indiana DOT report (Ferreira-Diaz et al. 2009) identified two leading causes of worker fatali- ties in work zones. The first is workers being struck by passing motorists, and the second is workers being struck by moving equipment within the work zone. The study determined that the causes of these types of incidents primarily were the result of “negligence of a third party” and “lack of aware- ness from the injured worker” (Ferreira-Diaz et al. 2009). Causes of work zone incidents are varied; however, high-risk objects, such as fast-moving vehicles and large pieces of equipment, present the greatest risk to highway workers. Lack of awareness by the injured worker is perhaps a cause that can be effectively mitigated through a highway worker safety program that includes worker training, internal traffic control plans, a focus on worker behavior, and improving safety culture. Although highway work sites are a specific and unique form of construction and maintenance sites, they retain many characteristics of other construction and maintenance sites. Therefore, it is also necessary to understand the prevalence and causality of incidents in all places where construction and maintenance work are performed. In his dissertation, Hallowell (2008) summarized construction and maintenance incident causation and analytic models for those incidents. He described five models of incident causation complexity, including the distractions theory, goals-freedom-alertness theory, adjustment stress theory, two-factor theory (unsafe acts and unsafe conditions), and the Swiss cheese model of incident causation. However, Hallowell also indicated that safety program elements that are physically implemented in organizations and work sites often lack connections to specific models of incident causation. That is, to be effective, it is important that safety programs take into consideration the established models of incident causation and target eliminating those behaviors, conditions, and operations that are shown to lead to incidents. Lacking such a connection, a safety program will not necessarily lead to improved safety. The Construction Chart Book, published by The Center for Construction Research and Training, is a synthesis of data relating to all aspects of the construction and maintenance industry (CPWR 2013). It provides visual representations of construction industry statistics and simple, easy-to-comprehend descriptions of industry data. In this format, the data are summarized and presented in figures and tables. Figures 1 and 2 are example figures taken directly from The Construction Chart Book and show how injuries and fatalities are distributed by cause in the construction industry. FIGURE 1 Distribution of leading causes of fatalities in construction in 2010. Source: CPWR 2013.

16 These two figures provide a robust summary of construction incident causes. The data used for these figures relates to the construction industry as a whole and does not specifically target highway workers. The Construction Chart Book is one of the most comprehensive repositories of safety and other statistics in the construction industry. avaIlabIlIty of HIgHway worker safety data Better understanding of the prevalence and causes of injury and fatality incidents for highway workers is possible through evaluation of data available to help quantify such incidents. This section explores the different sets of publicly available data and the background research that has used or sought to improve the data sets. The publicly available data sets are explored in greater depth for the specific information they provide to highway worker safety in chapter four. The data sets include those from the following entities: • Bureau of Labor Statistics (BLS), • Occupational Safety and Health Administration (OSHA), • National Institute for Occupational Safety and Health (NIOSH), • Fatality Analysis Reporting System (FARS), • Strategic Highway Research Program (SHRP 2), and • State DOT data. These data sources publicly available databases report raw data. Chapter four explores the uses for these databases. This section explores the six databases and other published data sources that synthesize data related to highway worker safety. BLS and OSHA are both divisions of the Department of Labor (DOL). BLS is responsible for col- lecting information on worker incidents through the agency’s Injuries, Illnesses, and Fatalities (IIF) program. This information is published on the BLS website and is searchable by industry, allowing the identification of incidents that occurred in highway work sites (BLS 2016). OSHA also reports Inves- tigation Summaries for fatalities and catastrophes that occur in the workplace. (Note: OSHA defines a catastrophe as a workplace incident that results in the overnight hospitalization of one employee.) These summaries are searchable by industry and key word in the online database (OSHA 2016). Hinze et al. (1998) highlighted the lack of specificity in root causes of construction and maintenance inju- ries and demonstrated that by increasing the detail of the database, the search results of relevant inci- dents would be more comprehensive. Since the publication of that journal article, OSHA has increased the specificity of crash causes in the database, particularly with detailed key words. Although the FIGURE 2 Distribution of leading causes of nonfatal injuries resulting in days away from work in construction in 2010. Source: CPWR 2013.

17 ability to search specifics has been improved, it is still difficult to aggregate and quantify the records in OSHA’s online database. NIOSH, a division of the Centers for Disease Control and Prevention (CDC), researches specific workplace incidents and compiles the resulting reports within its Fatality Assessment and Control Evaluation (FACE) program. The reports are detailed and provide a description of the incident. Some of the reports are specifically related to fatality incidents in highway work sites (NIOSH 2016). The FARS database, which is maintained by NHTSA, documents all fatal crashes on public road- ways. The details of the incident that are recorded in the police report are collected and placed into the FARS database (NHTSA 2016). Over the years, the police crash reports have been inconsistent on how they document fatal crashes that occur in work zones. Because FARS is based on these police crash reports, the database coding has some different searchable factors over time. In 2004, a study published in the Transportation Research Record explored individual state reports and how the states quantified work zone incidents (Ullman and Scriba 2004). The study concluded that the database at the time might have been underreporting work zone crash fatalities by as much as 10% based on the variations in forms used between states (Ullman and Scriba 2004). The FARS database does not allow the user to select “work zone” as a filter for records dated before 2009. This limitation may be the result of the discrepancies found in the reporting of work zone incidents identified in the 2004 report (NHTSA 2016). The Strategic Highway Research Program 2 (SHRP 2), which began in 2006 and ended in 2015, was a broad program to uncover data-driven solutions to transportation challenges (NAS 2016). One element of the SHRP 2 program was the Naturalistic Driving Study (NDS), which focused on improving highway safety. The project was administered by the National Academy of Sciences (NAS) and TRB and resulted in a database that, to different extents, can be used by researchers to better understand real-world driver behavior and analyze traffic incidents (Campbell 2012). Contrac- tors for the study were the Virginia Tech Transportation Institute (VTTI), which collected the NDS data, and Iowa State Center for Transportation Research and Education, which collected the roadway information data. This research is ongoing but maintains an online database with some of the data collected to date. State DOTs also maintain their own internal databases of traffic and worker incidents within their states. Each of these databases is different based on the type and volume of data the state chooses to collect and archive. Therefore, the detail and availability of the data from the state archives vary. The Construction Chart Book presents data in the form of charts, figures, and graphs. Although this publication presents analyses of the entire construction industry, some visuals specifically relate to highway construction and maintenance (CPWR 2013). Figure 3 is an example directly from The Construction Chart Book that makes specific references to highway construction. Figure 3 is representative of the type of data available in The Construction Chart Book. Although the data are more useful in determining overall trends and characteristics of the construction industry, select tables and figures provide insight into the nationwide highway construction and maintenance industry. Another resource, “Safety Management in the Construction Industry” (McGraw Hill Construction 2013), attempts to quantify the existence and characteristics of safety policies and programs in the construction and maintenance industry. This resource also focuses on the types of safety practices that are in use, the impact of those safety practices, influence factors, and communication and education. All of these sections provide the reader with visual representations of the practices and how the practices affect (positively and negatively) the construction and maintenance industry. The data are particularly useful in that they show the frequency at which construction and maintenance firms use different pro- grams and education methods (McGraw Hill Construction 2013). The National Safety Council (NSC) publishes Injury Facts, an annual synthesis of injury statistics in the United States and around the world (NSC 2016). The content is not limited to any one industry

18 or occupation. The statistics include a broad exploration of fatalities and injuries that occur from unintentional and intentional (homicide/suicide) incidents in the United States and, to a lesser degree of accuracy, the world. Subsequent sections of the publication divide nationwide statistics into cat- egories, such as occupational, motor vehicle, and home and community incidents. The occupational category of injury and fatality incident statistics pertains most directly to highway workers because their occupation places them at risk of injury or death while on the job. The 2016 edition of the NSC’s Injury Facts section on occupational injuries and fatalities is general in its presentation of available statistics, so it is impossible to isolate specific statistics related to highway workers. However, some of the general information can provide context in the form of potential trends in occupational incidents related to highway workers who are employed by state DOTs. Using data from 2013, the NSC reports more than 1.1 million injuries and ill- nesses that resulted in time away from work that year. Of these, approximately 65,000 (6%) were state government employees. This is not limited to state DOTs and includes all agencies within state government. Approximately 190,000 of the national lost-time injuries and illnesses were recorded for local government employees; the remainder was for private industry employees (NSC 2016). The NSC publication also provides general information on the estimated economic cost of occu- pational incidents. In addition to worker’s compensation insurance, these costs include loss of pro- ductivity and administration expenses. In 2014, the NSC estimated that the cost of each death is approximately $1 million. For every medically consulted occupational injury, the average total cost is $29,000. In addition to the monetary cost, the NSC recorded the lost time. In 2014 in the United States, among all industries, the total time lost was 99 million days, with 65 million of those days lost because of work-related injuries (NSC 2016). Other statistics, such as injury types and causes, are included in the NSC Injury Facts. However, the data categories most closely matching highway workers are the construction industry and gov- ernment employees (NSC 2016). Neither of these provides enough detail to observe the unique risks and incidents associated with highway construction and maintenance workers. Although the data sources mentioned are national databases or publications and are publicly available, individual states also maintain in-state records and data for the safety records for their individual agencies. More discussion on state data sources available to individual state DOTs is provided with the survey results in chapter three. States may choose to rely most heavily on their FIGURE 3 Share of dollar value of public sector construction, by type, in 2010. Source: CPWR 2013.

19 in-agency data because it is likely the most pertinent to safety successes and issues within their state. However, national data can be used to normalize incident rates and find common areas of concern among different states. legal standards and PolIcy recommendatIons related to HIgHway workers Federal agencies have implemented nationwide guidelines to help states improve their safety pro- grams and protect highway workers. One of the most visible documents describing these programs is FHWA’s Work Zone Operations Best Practices Guidebook (FHWA 2013). This report is published at the national level to assist states and provide guidance on common work zone issues. The best practices identified by this guidebook are categorized into 11 topics and 49 key words to allow users to search the practices by a particular topic. States have enacted their own policies and guidance for improving safety in highway work sites. Strategic Highway Safety Plans (SHSPs) are a requirement for state DOTs that are a part of the High- way Safety Improvement Program (HSIP). This requirement has continued under the Fixing America’s Surface Transportation (FAST) Act (FHWA 2016c). These programs direct the decision making of state DOTs based on the governing federal policies. Some of the state SHSPs that specifically outline priorities in work sites are summarized here. • Illinois’s SHSP notes that a particular issue in the state’s work zones are incidents located on the Interstate system that involve heavy vehicles. The strategies for improving work zone safety involve the three Es: engineering, enforcement, and education. The report specifically identifies that Illinois work zone speed laws are weak and do not allow for adequate enforcement (Illinois Department of Transportation 2009). • Massachusetts’s SHSP focuses on the lack of ability to quantify nonfatal incidents in highway work zones. The report indicates options to improve safety that are being explored. In addition, it outlines strategies for improving work zone design and educating the public about safety in work zones (Massachusetts Department of Transportation 2013). • Minnesota’s SHSP presents detailed statistics for incidents that occurred in work zones. The data include types of incidents and the demographics of those involved in the incidents. The report specifically recognizes that, for Minnesota, there is an overrepresentation of commercial vehicles involved in work zone incidents and indicates this is a safety focus area for the state (Minnesota Department of Transportation 2014). • South Carolina’s SHSP identifies that the state has been seeing an increase in work zone fatal and severe injury collisions. The report identifies strategies to meet an objective of preventing a continued increase in work zone-related collisions; the state saw such an increase between 2008 and 2012. Some of the strategies include improved law enforcement and first responder training relating to work zones (South Carolina Department of Transportation 2015). • Washington’s work zone safety issues are handled by the state’s Work Zone Safety Task Force (WZSTF). The SHSP outlines specific strategies of the WZSTF that focus on the visibility of workers and work zones. The report also states a WZSTF priority of maintaining worker training and improving public notification of work zones (Washington State Department of Transportation 2013). State and federal policies drive decision making in terms of protecting highway workers and preventing incidents. These policies vary in effectiveness based on how they are communicated and how they are implemented. In addition, the policies need to be accepted as effective for improving safety to ensure continued compliance with the policies among state DOTs. In addition to legal policies produced by state and federal government agencies, the construction and maintenance industry makes recommendations on programs and policies that have demonstrated the potential to improve safety. The following literature is a collection of documents describing research conducted for the construction and maintenance industry as a whole, elements of which can be directly applicable to highway work sites.

20 Construction and maintenance industrywide guidelines are available that provide broad recommen- dations to improve safety. One example is “Guide to Best Practice for Safer Construction: Principles,” a nationwide guideline for Australia. The text presents six principles the authors think are important to fostering a healthy safety culture on a work site. These principles are (Fleming et al. 2007): • Demonstrate safety leadership, • Promote design for safety, • Communicate safety information, • Manage safety risks, • Continuously improve safety performance, and • Entrench safety practices. These principles are described and promoted similarly in other publications on safety culture in the construction and maintenance industry. Although this document is not detailed and provides only general recommendations that accompany each principles, it is good source for the understanding of the different aspects of a safety program. In addition, because this document is an Australian publica- tion, it can give safety officials in the United States a new perspective on safety programs. As innovative safety initiatives are adopted by industry, it is important that research be produced to discuss and analyze such initiatives for the implementation of these program elements. Instituting safety programs that rely on leading indicators is becoming more accepted and more common in the construction and maintenance industry. Leading indicators represent the conditions and behaviors exhibited in a workplace that provide an indication of the level of safety performance. In contrast, lagging indicators (e.g., injury incidence rates) provide a retrospective assessment of safety perfor- mance. Monitoring leading indicators may be new to some state agencies. When a concept is in its infancy, it is more likely to be implemented incorrectly. To counter this possibility, the Construction Industry Institute (CII 2012) published a report that outlines a nine-step process for implementing active leading indicator safety ideas. Figure 4, which was provided in CII’s report, explains the process FIGURE 4 Flowchart for implementing active leading indicator initiative. Source: CII 2012.

21 in the form of a flowchart. The entire process is based on the idea of improvements and adjustments being made as necessary throughout the course of implementation. Another initiative in the construction and maintenance industry is the “zero injury objective,” whereby no incidents and therefore no recordable injuries occur on construction and maintenance sites. This objective is similar to the Toward Zero Deaths (TZD) vision the FHWA, state DOTs, and municipalities are currently championing. This is related primarily to the elimination of trans- portation user fatalities, as opposed to those of workers, but all are included (FHWA 2016d). Accord- ing to FHWA, the TZD vision “uses a data-driven, interdisciplinary approach that FHWA has been promoting for many years. The TZD approach targets areas for improvement and employs proven countermeasures, integrating application of education, enforcement, engineering, and emergency medical and trauma services” (FHWA 2016d). A report by Hinze and Wilson (2000) examines the progress in the construction and maintenance industry after the zero injury objective gained traction. The authors indicate that safety programs are becoming more common and more effective. Because safety programs are becoming more common, the next logical progression is evaluating the existing programs and making improvements. In some cases, large private construction and maintenance companies draft their own safety programs and policies that are applied on all of their projects. This requirement adds consistency to the safety regimens from project to project, with the intention of improving institutional safety. One such company is the Howard S. Wright Construction Co. (HSWCC). The companywide safety guide includes directives for administration officials and detailed requirements for workers in many common work scenarios; the directives concern issues such as personal protective equipment, fall protection, tools, heavy machinery, and specialty machinery (HSWCC n.d.). Another company with a comprehensive safety plan is Skanska. Skanska’s guidelines include policies for common scenarios, and provisions for subcontractors and various emergency situations, such as natural disasters or terrorist activity (Skanska 2005). Whether the safety measure is a federal, state, or company guideline or policy, it is important that workers follow it during the course of the work. Maintaining consistent safety programs that are easily understood by workers at highway work sites results in safer work sites for workers and motorists. rIsk and Human factors At any work site, there is a degree of risk to the safety of the workers. Understanding this risk and the hazards that create the risk is vital to protecting highway workers to the greatest degree possible. Some of the risk results from the inherent qualities of the work itself, including the geographic fea- tures of the work site. Risk of injury and fatality is also the result of human factors and human error on the part of the worker or motorists traveling within highway work sites. Early investigations into work-related incidents were aimed at quantifying the extent to which human behavior contributes to injury incidents and revealed behavior to be a significant factor. For example, in an analysis of injury incidents, Heinrich (1959) found that 88% of incidents were caused by unsafe acts of persons (Heinrich 1959; Johnson 2011; Manuele 2011). Heinrich (1950) proposed that management personnel in an organization have the best opportunity and ability to initiate the work of prevention, so management assumes the responsibility. Contemporary literature strongly supports this axiom of management involvement and suggests examining the entire organizational and industry spectra to understand fully the causes of injuries and how to prevent them (Rasmussen et al. 1994; Suraji et al. 2001; Haslam et al. 2005; Gibb et al. 2014). The roadway type also contributes to risk for workers near the roadside. The diversity of the struc- ture of state DOTs means that, depending on which agency employs the worker, the worker could be exposed to different functional roadway classifications, each associated with a different level of exposure and risk. Many state DOTs construct and maintain primary routes, which generally have

22 higher speeds. The severity of crashes has been shown to increase as the speed of passing traffic increases (Aarts and van Schagen 2006). Therefore, the risk for employees is higher on routes that the state DOTs are more likely to construct and maintain. According to this literature, to be effective the prevention measures implemented by an organi- zation must address human factors at their foundational level and apply prevention measures at all levels in an organization. The facets of basic human behavior that exist as root causes of incidents include mistake/error, absent-mindedness, lack of caring/indifference, ignorance, poor risk manage- ment, and high-risk tolerance. These behaviors can exist anywhere within an organization and be the cause of an injury incident. For example, a worker at the front line may poorly assess the risk present in a situation and cause an injury incident. In addition, a decision made with indifference at a management level may cause a worker on the site to act in a certain way that leads to an injury. The cause in this case is the result of management’s indifference and is not associated with the worker. The literature argues that safety management programs be designed to address all of these unsafe human behaviors and eliminate such behaviors at all levels within an organization. With regard to worker injuries from crashes, multiple risk factors, including at-fault driver, envi- ronmental condition, crash information, road condition, driver error, and others, have been identified (Li and Bai 2009). The researchers determined how these risk factors affected the severity of work zone crashes in Kansas. Statistical analyses were used to determine which factors were significant, such as driver error, vehicle type, light conditions, age, and gender. The researchers concluded that additional efforts to improve compliance with existing work zone traffic laws could reduce driver risk. In another study, McAvoy et al. (2011) identified risk factors, which included roadway type, traffic density, and work zone type. The researchers used their risk factors to test the performance of alternative work zone configurations in a driving simulator environment. The human factors literature related to crashes in work zones primarily concerns the behavior of motorists operating within work sites. In their respective studies, Reyes and Khan (2008) and Morgan et al. (2010) performed tests in a driving simulator to evaluate alternative work zone con- figurations. Driving simulation provides a safe medium for evaluating driver behavior in work zones. Reyes and Khan performed a study of how drivers changed their behavior based on the presence of different types of barriers that separated the travel way from the work zone, among other variables. One of the other variables, the level of activity in the work zone, influenced the speed of the drivers, with the average speed being lower for work zones with higher activity levels. Morgan et al. tested a lane reduction work zone with a reduced taper length. The study concluded that although the reduced taper width slightly reduced risk because of the shorter work zone, the overall risk to the motorists and workers increased because of the reduced taper. Other research has provided evidence of additional positive impact on safety risk and driver behavior from the use of temporary traffic control devices in work zones, including portable changeable message signs, radar speed trailers, and advisory speed signs (e.g., Zhang et al. 2014; Gambatese and Zhang 2016). Safety risk for highway workers is directly related to human factors, so understanding these factors and the types of hazards they create in work sites is important. Hazards related to construction and maintenance operations and the behaviors of construction and maintenance workers themselves affect the safety of highway workers. One aspect of human behavior consists of risk taking. Some humans are more willing to exhibit risky behaviors than are others. In a 1991 study, evidence suggested that humans who are more prone to incidents are more likely to be risk takers (Dahlbäck 1991). Understanding the workforce at the construction and main- tenance site and the level of risk the workers will accept may be an indicator of the severity and fre- quency of incidents on such a site. One method for potentially reducing risk and taking advantage of natural human behavior is to incentivize safe practices. In a section of the CII’s 2003 report “Safety Plus: Making Zero Accidents a Reality,” the authors indicated that incentives (recognition and reward) for safe behavior are a common strategy in the construction and maintenance industry. The report concluded that although incentives may be effective if applied correctly, the progressive incentive (reward based on injury- free work hours) in particular was not shown to promote improved safety in the long run (CII 2003a).

23 In 2014, Gambatese reported that safety incentives in the workplace can have a positive effect on safety. However, the effect has not always been quantifiably proven, and many factors, such as the manner, type, and frequency of the incentive, can significantly influence the effectiveness of the incen- tive (Gambatese 2014). OSHA has indicated that safety incentive programs may result in employees being discouraged from reporting incidents in which injuries occur, particularly if the incentive is high (Fairfax 2012). Safety culture is a more abstract concept but affects the risk workers are exposed to based on a col- lective awareness among a group or organization. Formal and informal policies, workplace norms and conditions, ethnic makeup, and social interactions, all of which collectively make up workplace culture, affect safety performance. The safest organizations understand that a positive safety culture can make a difference not only in worker safety performance but also in overall project performance. Although safety culture is a topic that has been identified as a factor influencing safety performance (Molenaar et al. 2009; Zou 2011; Gilkey et al. 2012), the specific attributes that make up safety culture, the magni- tude in which these attributes affect safety culture, and the impact of safety culture on worker behavior and overall safety performance require in-depth study, confirmation, and quantitative measurement. As the construction and maintenance industry continues to look for ways to improve safety per- formance, safety culture, with its effect on worker behavior, has been identified as having an impact (Molenaar et al. 2009; Gilkey et al. 2012; Zou 2011). The development of a positive safety cul- ture has been identified as a means for affecting worker behavior to improve safety performance. Many definitions of safety culture have been developed. For example, in the National Occupational Research Agenda (NORA), safety culture is defined as the “organizational principles, norms, com- mitments, and values that relate to how safety and health is operationalized” (NORA 2008). Cultures are patterns of interacting elements and represent the accumulated learning of a group—the ways of thinking, feeling, and perceiving the world that have made the group successful (Schein 1999). Although establishing a positive culture in an organization is desirable, measuring an organization’s culture is difficult, especially in the construction industry. Determining an organization’s culture requires examining deeply held personal beliefs in connection with and in response to organizational principles and values. Approaches to measuring organizational culture have been developed, but there is no consensus regarding the most effective approach (for example, see Cameron and Quinn 1999; Schein 1999; Hofstede and Hofstede 2005). Additional difficulty in measuring culture exists in the construction and maintenance industry because of its fragmentation and project-centered focus. Culture becomes multifaceted with multiple cultures forming around different groups com- posed of individuals with commonalities (Sackman 1997). Cultures may also form at different levels (subcultures): for example, crew, project, organization, and industry (Schein 2004). NORA (2008) defines two levels of safety culture within an organization: top management and supervisor/subunit. Cultures formed around each project site are particularly influential because individuals assess a site culture quickly and intuitively (Hartley and Cheyne 2009). The strength of safety culture is determined by group stability and shared history (Schein 2004). For the construction and maintenance industry, the short-term nature of construction projects, multiple work partners, and high frequency of change in employment make developing and maintaining a positive safety culture problematic. Given the dynamic nature of the partnerships and conditions on construction projects, Maloney (2003) highlights the need for research that investigates culture formation amid a changing environment. Beyond the presence of a positive safety culture, an employee’s interpretation of the safety culture and subsequent actions (behaviors) are important considerations when attempting to optimize safety performance. Worker response/behavior is reflected in the worker’s recognition of safety hazards, comprehension of the implications of the hazards, valuation of the associated risk, selection of an appropriate control, decision to implement a control, and the quality and extent of control imple- mentation. In their study of the creation of an impression of safety culture on construction sites, Hartley and Cheyne (2010) examine the connection between safety culture and worker behavior. The researchers found that construction sites and site personnel create an impression of safety cul- ture, which is quickly assessed by new workers as they enter the site, and that this initial assessment influences their behavior. To ensure safe behavior, a positive safety culture was recommended to be

24 established and in place at the start of a project and maintained constantly. The study was conducted using a small sample of management personnel surveys and two focus group interviews of construc- tion workers. Cautioning that discrepancies may exist between managers’ beliefs about the behavioral consequences of interpreting safety culture on sites and the beliefs of the workers themselves, the authors recommended additional investigation; the type required would target workers and confirm that their actions coincide with their beliefs about the impacts of safety culture. Current construction and maintenance safety research suggests a need to explore the interaction between employee interpretation/response and the safety culture in which the employee works. That is, a shift in thinking is needed to focus on the actions that lead to good safety performance by mea- suring employees’ perceptions of the way safety is in operation on construction sites (i.e., the safety culture) (Choudhry et al. 2009). Safety culture is one part of comprehensive safety management, which also includes consideration of work operations, work site conditions, a dynamic work site, and worker behavior and risk tolerance within a safety management system. Understanding the risks associated with construction, maintenance, and highway work sites and how motorists and workers respond in this environment can be useful in designing more effective work sites that better protect the employees and the public. Also important is having an understand- ing of worker behavior and how safety performance is influenced by behaviors to design effective highway worker safety management programs. stakeHolders In HIgHway worker safety The literature reveals that many different stakeholders have the potential to positively influence high- way worker incidents that occur at work sites. These stakeholders are vital and, whether they realize it or not, have a significant impact on the safety of highway workers. In the construction and mainte- nance industry, the most frequently considered stakeholders are the owners, contractors, and workers. Other stakeholders interested in highway worker safety include insurance companies and transportation professionals. The Indiana DOT’s report Worker Injury Prevention Strategies specifically discusses the owner, contractor, and worker stakeholders for safety in work zones. This study revealed that all three groups have similar views about work zone realities, such as traffic in the work zone posing a high risk to workers. However, these three stakeholders have differing ideas of the effectiveness of prevention measures, such as law enforcement and efforts to improve safety. The study found that the workers, who are directly exposed to the hazards in work zones, were the least satisfied with safety efforts (Ferreira-Diaz et al. 2009). These results demonstrate the competing interests of stakeholders and indicate that workers believe they have the highest risk among all of the stakeholders. The NCHRP report Training and Certification of Highway Maintenance Workers fundamentally acknowledges that state DOTs and workers are vital stakeholders in highway worker safety. Accord- ing to the report, state DOTs understand the need for maintenance workers to understand the aspects of their maintenance job and the need for training in proper safety techniques. Many states incentiv- ize their maintenance workers to partake in safety training (Laffey and Zimmerman 2015). The project owner, which in the case of many highway construction and maintenance projects is the state DOT, can have a significant influence on the safety of the project based primarily on actions taken during the beginning stages of the project. A study by the CII looked specifically at ways that owners can have a role in construction and maintenance worker safety. Some of the findings included practical methods that reduce worker risk. These methods include the following (CII 2003b): • Selecting contractors based on known safety records, • Emphasizing safety requirements in project contracts, • The owners being actively involved in safety throughout the project duration, and • Selecting projects that incorporate characteristics such as design-build delivery methods and a 5-day or less work week.

25 Some of these methods are not as applicable to state DOTs that may use their own employees for a project or have little choice in the contractor that wins a project. However, an understanding that the owner’s role in safety for all projects begins long before construction or maintenance starts and continues until construction or maintenance ends is important to reducing risks to workers throughout the duration of the project. There also is significant literature regarding insurance companies as a vital stakeholder in the construction and maintenance industry. Insurance companies are incentivized to influence design to improve safety. In the report Cost of Highway Work Zone Injuries, the authors quantify the monetary costs associated with incidents in highway work zones (Mohan and Gautam 2000). This research was conducted when it was determined that highway work zone fatalities were not decreasing as were the overall construction and maintenance industry fatality rate and highway transportation fatality rate. The authors determined an average monetary value for different types of incidents that could occur in work zones, focusing on the costs associated with motorists who are involved in crashes in work zones (Mohan and Gautam 2000). Another study looked at the idea of prevention through design (PtD) from an insurance perspective. This historical look at PtD discusses how the cost of construction and maintenance incidents encourages insurance companies to be influential in the design of safety programs and practices to minimize costs in terms of money and human life (Braun 2008). In the textbook Construction Safety Management and Engineering, two chapters specifically address the injury costs and the role of insurance in the construction and maintenance industry. The first chapter (Hill 2014) breaks down the worker compensation system that pays for construction and maintenance worker injuries. Hill (2014) argues that, to be successful, a safety administra- tor must understand the financial aspects of the system and the practical measures of reducing incidents. The second chapter (Ennis 2014) explains the construction and maintenance insurance industry. Although workers’ compensation insurance is a portion of the insurance portfolio, Ennis explains that there are many other types of construction and maintenance insurance. A complete insurance portfolio is likely to include coverage for the workers, such as medical and disability benefits, as well as coverage for the project itself. This coverage can include commercial auto liability, commercial general liability, pollution liability, and insurance to cover the structures and equipment used in the project (Ennis 2014). These sources indicate that it is important to use economic data in the development and evaluation of safety programs. In 2012, Ikpe et al. proposed that there is a positive relationship between prevention costs and subsequent benefits. The report also suggested that there is an inverse relationship between incident costs and prevention costs. Although the figures differ based on the size of the construction firm, the benefits of incident pre- vention outweigh the economic and direct costs by an average of 3 to 1 (Ikpe et al. 2012). Looking at the raw economic costs of incidents could help to bolster the argument for a particular safety program at a state DOT. Insurance companies recognize the institutional benefit of promoting safety in the workplace. Liberty Mutual, an insurance company that is involved with business insurance, published its “The Architecture of Safety Excellence” model. This initiative, designed to bring attention to all aspects of workplace safety, highlights four cornerstones of the risk management process. These cornerstones are as follows (Liberty Mutual Insurance 2003): • Risk Assessment—identify and measure the existing levels of risk present in the workplace. • System Analysis—determine if the system, for all parties involved, is working well. • Integrated Solutions—three solutions, including implementing design solutions, adding educa- tion programs, and enhancing motivation, must be combined to improve the system • Progress Measurement—determine if the programs put in place have had a positive impact on safety. Insurance companies recognizing themselves as a stakeholder is an important step, but other stakeholders (i.e., construction and maintenance companies, owners, workers, etc.) regarding insur- ance companies as a legitimate stakeholder is more critical. Doing so not only protects human life but also reduces the costs related to highway construction and maintenance.

26 In a report on the OSHA standards that are important to professional engineers, Toole and Gambatese (2002) propose that in the construction and maintenance field, professional engineers are another stakeholder in worker safety through their designs. By understanding OSHA safety standards and accounting for them in construction and maintenance designs, professional engineers have the capacity to help reduce work site (including highway work zone) incidents (Gambatese et al. 2003). This concept is often referred to as designing for safety (DfS) or PtD. In the construc- tion industry, safety on the construction site traditionally has been assigned to be the responsibility of the constructor. This responsibility reflects the constructor’s control over the project site, workers, and work practices and procedures. However, engineers who design the project features can pro- actively play a role in site safety because their designs influence the hazards to which workers are exposed (Churcher and Alwani-Starr 1996; Smallwood 1996; Haslam et al. 2003; Driscoll et al. 2004; Gibb et al. 2004; Behm 2004, 2005; Gambatese et al. 2008). The basis for designer involve- ment is the OSH principle that the most effective and reliable means for ensuring worker safety is to eliminate safety hazards from the work site before workers are present. Designing to eliminate a hazard can be more effective than controlling the hazard or protecting the workers from the hazard (Manuele 2013). Designing for safety is founded on the hierarchy of controls, which provides guidance on safety and health problem solving (Andres 2002; Manuele 2013). In order of decreasing effectiveness, the hierarchy is as follows: elimination, substitution, engineering control, administrative control, behav- ior, and personal protective equipment (PPE). If the hazards cannot be eliminated, controls lower in the hierarchy may be suitable with appropriate consideration given to the corresponding lower levels of reliability and effectiveness. When determining how to mitigate safety risk during construction and maintenance operations, eliminating hazards should be the first choice, according to the DfS concept. Designing for safety is one of the foremost methods for eliminating hazards and reducing risk regardless of work industry. By waiting to apply safety controls during construction and main- tenance operations, frequently the most effective approach available is to warn workers of hazards, implement procedures, train employees, and provide personal protective equipment. Research has uncovered examples of designing for safety in practice within the transportation industry and identified recommended DfS processes (Gambatese et al. 1997). A simple example is to design locations on roadway shoulders for construction and maintenance vehicles to park where they are protected from oncoming traffic. These parking spots could be adjacent to roadway light- ing at locations where maintenance work is periodically needed to replace lamps. Another example is to design bridge guardrails such that they meet the guardrail height requirements set by OSHA for exposed edges (42 in. ± 3 in.). Permanent guardrails at this height provide protection for bridge users during construction and maintenance operations and eliminate the need for those conducting construction and maintenance work to install a temporary guardrail. Processes have been developed to efficiently implement the DfS concept in the project develop- ment process (e.g., WorkCover 2001; Angelo 2004; Istephan 2004; Hecker et al. 2005; Weinstein et al. 2005; Toole and Gambatese 2011). In one example, Bovis Lend Lease (BLL), an international design and construction company, established and implemented the program ROAD—Risk and Opportunity at Design (Zou et al. 2008). ROAD aims to eliminate or minimize the risks of injury throughout the life of the product being designed by involving all decision makers who contribute to the life cycle of the product (ASCC 2006). ROAD incorporates the following key elements and considerations: person with control, product life cycle, systematic risk management, safe design knowledge and capability, and information transfer. BLL implements the ROAD process through the following nine steps (BLL 2004; Zou et al. 2008): 1. Building element assessment at the preconstruction phase. 2. Providing trade package assessment at the construction stage. 3. Recording the ROAD document and uploading it into the project management plan. 4. Including a ROAD agenda item on design program meetings. 5. Establishing action and status lists. 6. Updating and reporting status at each design review. 7. Considering actions from ROAD issues before approval for construction.

27 8. Holding environment, health, safety, and quality monthly management meetings to review the reporting of projects including the ROAD status. 9. Having a monthly update of the ROAD document as part of the project review. Behm, in conjunction with Toole and Gambatese, further explored the DfS concept. Behm argues that, in a process known as design for construction safety and health, risk to workers and hazards on the construction site can be reduced or eliminated during the design process. In addition to safety, this process can increase overall project quality and cost effectiveness. However, Behm concedes that there are barriers to implementing this practice, including contractual, legal, and regulatory issues. These barriers arise from design for construction safety and health being less common in the United States than elsewhere, such as Europe and Australia (Behm et al. 2014). Additional research has revealed other barriers, such as a lack of designer training with respect to safety, a fear of third-party liability for worker injuries, additional costs associated with imple- menting design for safety practices, and difficulty in identifying hazards before the start of the work on site (Hinze and Wiegand 1992; Gambatese et al. 2005; Hecker et al. 2005; Toole 2005; Gambatese 2008). Despite these barriers, professional engineers can still influence safety on construction sites (including highway work zones) through their designs, and designing for safety has been identified as a viable intervention in the construction industry (Gambatese et al. 2005). Designing for safety is currently not a common intervention in the construction industry. However, as designing for safety diffuses throughout the industry, industry reactions and changes are expected. Toole (2001) showed how the characteristics of a construction task, process, and industry have caused innovative building products to follow one or more trajectories. The concept of trajectories can also be applied to the implementation of DfS. Toole and Gambatese (2008) identified four specific trajectories that PtD is likely to follow: (1) increased prefabrication; (2) increased use of less hazardous materials and systems; (3) increased application of construction engineering; and (4) increased spatial inves- tigation and consideration. Owner organizations, including state DOTs, have been identified as key stakeholders and drivers of designing for safety on work sites (Toole et al. 2016). Designing for safety is just one element that may be included in a state DOT’s safety program. Although each stakeholder has a different interest in designing for safety, and highway worker safety as a whole, the collective group has the capacity to affect safety in work zones through cooperative action and mutual understanding. It is also clear from safety and health research in the construction industry that the top management within an organization is in a position to positively affect safety performance. In addition to the stakeholders mentioned, many other groups have a role in highway worker safety. These include local agencies that often work closely with state DOTs for large projects. Other organizations, such as emergency medical services, law enforcement, state legislators, and public health officials, can have an impact on worker safety, particularly in setting policy and responding to incidents. Professional groups such as ASCE, ITE, and the Association of General Contractors also can contribute to worker safety programs at state DOTs to keep workers and the traveling public safe. All of these stakeholders have the opportunity to collaborate and combine their efforts to advance the state of the practice in highway worker safety. maIntenance worker Issues As state DOT budgets tighten, the traditional focus on infrastructure development has transitioned to one with a greater priority given to maintenance of existing roadways rather than new construction. Maintenance on roadways presents different challenges and risks to the traveling public and state DOT employees performing maintenance work. In general, maintenance work is more frequently performed on roadways with higher demand, meaning vehicle-worker conflicts may be more preva- lent. In addition, the temporary nature of maintenance work may limit the protections and visibility afforded to construction workers in more established work zones. These differences can result in greater levels of risk to highway workers.

28 FHWA recognizes this difference and the subsequent importance of highlighting unique mainte- nance worker safety issues. In a 2014 brief, FHWA indicated that workers establishing a maintenance work site can best identify the individual safety concerns of that particular site. The brief describes many low-cost mitigations that can be incorporated into a maintenance site based on the safety defi- ciencies identified by workers. These include solutions from signage and pavement markings to the physical road surface and roadside itself (FHWA 2014). The NCHRP Report 500 series is a 23-volume collection of reports that provide guidance for imple- menting AASHTO’s Strategic Highway Safety Plan. Volume 17 of the series, A Guide for Reducing Work Zone Collisions, documents that, although maintenance work zones are generally in place for a short time, 8% of fatal work zone crashes in 2003 occurred in maintenance zones (Antonucci et al. 2005). A 1998 report published in the Transportation Research Record describes the desired role of maintenance as an element of an overall infrastructure development program. The report argues that, to promote the safety of motorists and workers in maintenance sites, maintenance managers should be committed to continuing education so that they understand and implement the latest technology and policy initiatives to promote maintenance site safety (Byrd 1998). NCHRP Report 475, which outlines options for planning nighttime highway construction and maintenance, acknowledges that it is desirable for many maintenance activities to be completed without restricting the full volume of the roadway. One approach to meeting this desire is to conduct nighttime maintenance work when road volumes are much lower, and some of the risks associated with daytime are reduced. However, nighttime work introduces alternative risks, such as reduced visibility and an increased incidence of drivers who are less alert and more likely to be impaired. The report presents potential strategies, such as increased traffic control, for minimizing the risks for nighttime maintenance work (Bryden and Mace 2002). Another source for information relating to maintenance work site issues is the National Work Zone Safety Information Clearinghouse. This information repository is a resource for any stakeholder seeking to improve work zone safety for everyone who interacts with a work site, including highway workers (Clearinghouse 2015). One of the elements of the National Work Zone Safety Information Clearinghouse is a searchable database of publications related to work site safety available on their website (www.workzonesafety.org). This database, when filtered for publications relating to mainte- nance, provides several specific reports relating to maintenance worker safety (Clearinghouse 2015). One of the resources from this database is the “Maintenance Work Zone Safety” pocket guide. This document, which provides excerpts from the MUTCD on temporary traffic control, is to be used in the field by maintenance workers to make their work sites as safe as possible. Because most maintenance operations are short term, this guide focuses on the short-term traffic control require- ments presented in the MUTCD. It provides seven core principles of temporary traffic control that for incorporating into maintenance sites. These principles are as follows (American Traffic Safety Services Association 2008): 1. Plan for traffic safety. 2. Minimize interference with vehicular, bicycle, and pedestrian traffic flow. 3. Provide clear and positive guidance on how to get through the temporary traffic control zone. 4. Perform continuous inspection and maintenance of temporary traffic control zone devices. 5. Maintain roadside safety throughout the project. 6. Make sure workers receive the training that is required. 7. Maintain good public relations. Another study, conducted at the University of Kentucky, sought to identify best safety practices for highway maintenance workers. A survey that solicited ideas from public and private maintenance workers in Kentucky was conducted to determine what practices could be implemented to improve their safety. Some of the recommendations from the report included closed cab tractors for mowers on the roadside, LED stop signs in maintenance sites, and additional lighting in nighttime conditions (Hancher et al. 2007).

29 The literature presented in this section highlights some of the unique safety issues that mainte- nance workers encounter on roadways as opposed to construction workers in more established work sites. These safety issues demonstrate the potential for high-risk exposure for maintenance workers. Therefore, the literature suggests, maintenance worker issues require particular attention, and state DOT safety plans should consider maintenance workers as a special element of their overall safety programs. evaluatIon of safety Programs To keep improving worker safety and determine if improvements are necessary to the existing poli- cies and programs, it is necessary to evaluate the existing effectiveness of these policies and programs. Literature exists that relates to evaluating a state’s record in highway work sites and to barriers to reducing work site incidents. In one study, Minnesota DOT sought to determine, through a simulator study, the safety impacts of the potential implementation of automated speed enforcement (ASE) systems in work zones. The researchers, understanding that speed and distraction are key issues in worker safety incidents, found the ASE system did not significantly affect driver behavior. However, when the ASE system was combined with a dynamic speed display sign, there was improved driver attention, especially among younger and older drivers (Morris et al. 2016). Virginia DOT (VDOT), which historically has used injury and fatality statistics to evaluate work zone performance, published a report outlining potential improvements in the evaluation system of work zones. Using a state-level database that contains information on weather and work zones and combining that information with raw incident statistics, VDOT established a rate measure to normalize the incident statistics based on the frequency of road construction and maintenance. This rate allows VDOT to better evaluate trends in work zone performance based on other factors (Kweon et al. 2016). There are also barriers to the implementation of safety programs. A 2014 report from the Missouri DOT (MoDOT) observed survey response data suggesting that although MoDOT employees thought work zone setups accurately conveyed the hazards to the public, motorists more often perceived that the warning systems were inaccurate or did not provide enough information. Barriers such as driver education and lack of worker compliance with work zone requirements limit the effectiveness of the safety measures placed into the design of work zones (Long et al. 2014). Evaluating safety programs is critical to determining their effectiveness and identifying key areas of success that can be expanded and replicated. A comprehensive NCHRP report outlines ways in which work zone data are collected and how the data are used to evaluate and assess the performance of work zones (Bourne et al. 2010). The report finds that safety programs are established at the state agency level and are based on federal guidance. Detailed programs that recognize and use available data are more likely to consider the effectiveness of work zone safety features during the entire design, construction, and maintenance process. This effectiveness includes the ability to use data to adjust the work zone safety features throughout the process as needed (Bourne et al. 2010). This document highlights how accurately evaluating work zone and highway worker safety programs is a necessary step in improving over- all safety. The literature cited earlier examines how the transportation engineering field evaluates safety pro- grams in work sites and how the construction and maintenance industry explores the effectiveness of safety practices from the viewpoint of the organization, worker, and work site. In their report in The American Professional Constructor, Berryman et al. identified the five main factors that negatively influence safety programs (Berryman et al. 2006): • Inadequate employee involvement, • Lack of upper management enforcement,

30 • Inadequate monitoring and training, • Improper means to measure safety performance, and • Ineffective incentive programs. These factors were determined through a literature search and could be used as metrics for determin- ing if a program’s structure has the potential to improve safety. The use of leading indicators in construction and maintenance safety programs is becoming more widely used in the United States, including for highway work sites. However, it is important to be able to evaluate the effectiveness of programs that use leading indicators. A 2012 report from the CII identifies methods for evaluating safety performance in leading indicator programs. The report poses that the identification and measurement of leading indicators is critical to reaching the goal of zero work site incidents. This includes analysis of active (visible during the duration of the project) and passive (visible before construction or maintenance) leading indicators (CII 2012). It is also important to verify that collected leading indicators are correlated to incidents and properly mapped to lagging indicators. The evaluation of programs is critical because having an ineffective safety program can have high human and monetary costs. conclusIons This chapter has reviewed existing literature regarding current issues in highway worker safety. The collective research on the topic indicates that highway worker safety is a significant concern. Much of the literature discusses the issues related to vehicles and their effect on safety in work sites. A sig- nificant amount of research regarding the hazards of work sites from the perspective of the highway workers also exists that does not pertain to public vehicle crashes. All of this research, combined with the available data sources, provides a summary picture of current issues in highway worker safety. The research included in this chapter has identified several elements that can be effective aspects of an agency’s highway worker safety program. One of these elements is the option for the program to consider literature and data analysis practices from the traffic engineering and construction and maintenance engineering fields. Exploring recent trends in each of these fields can lead to a more comprehensive understanding of the unique challenges in highway work sites. Data regarding highway work sites are collected at the federal and state levels. For state DOTs to be able to implement effective safety programs, it is beneficial to understand the data available so they can be analyzed in some of the ways the research described in this chapter has done. The data can be combined with knowledge of risk and human factors that are intrinsic to highway work sites. This additional knowledge can include driver behavior and worker behavior at the work sites. In addition, the research has shown that identifying all of the stakeholders, including the owners, workers, and insurance companies, can better allow the natural efforts of these groups to improve safety to achieve a more efficient and effective product. Designers also play a key role in ensuring that designs minimize risk for workers who are constructing and maintaining state DOT infrastruc- ture. Finally, evaluating existing programs and continually updating them as trends and data change can ensure the continued effectiveness of program elements.