Improving Safety Culture in Public Transportation (2015)

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.

23 C H A P T E R 2 Introduction Examining the state of safety culture within the public transportation industry was accomplished using a multilayer, multistep research methodology that included: 1. A transit stakeholder survey to identify key components of transit safety culture and transit agencies perceived as having a positive safety culture for further investigation; 2. Selection of transit mini–case studies based on stakeholder suggestions, leading and lagging indicators, and other measures; and 3. Mini–case studies of nine transit agencies considered to have a positive safety culture, involving interviews with employees at all levels and analysis of relevant documentation. This chapter summarizes the research design, methodolo- gies, and findings that each of these research steps yielded. Industry Stakeholder Survey Survey Design The transit safety culture industry stakeholder survey was designed to identify: • Industry stakeholder perceptions of key factors that contrib- ute to a positive safety culture within public transportation organizations, and • Transit organizations perceived by stakeholders to have an overall positive safety culture. The stakeholders surveyed hold key positions in the transit industry and represent public transportation agencies of a variety of sizes and modes, including fixed-route buses, light rail, heavy rail, commuter rail, and paratransit. The stake- holders’ positions give them important knowledge of and insight into transit safety and safety culture. The following organizations were contacted to identify stakeholders for the survey: • The American Public Transportation Association (APTA) and its online member directory, which were used to iden- tify urban transit system chief executive officers (CEOs), board members, and members of safety committees. • The Community Transportation Association of America (CTAA) list of Certified Safety Professionals, which pro- vided a broad roster of safety specialists at small and rural transit providers. • The two largest national transit unions, Amalgamated Transit Union (ATU) and Transport Workers Union (TWU), which identified and distributed surveys to union representatives. • APTA, state safety oversight agencies, the National Trans- portation Safety Board (NTSB), the FTA, the FRA, and CTAA, which were contacted to identify oversight agency and industry association representatives. To limit the sample to a manageable size, participants were randomly selected from two groups, safety professionals and board member committees, which could be expected to have multiple representatives from individual transit agencies, espe- cially larger transit agencies. The transit CEO and local union president groups, although large, contain one representative from each transit agency or union and, therefore, were not further scaled down. Using a stratified random sampling method provided a good balance of transit agency size and the number of transit agencies and unions involved. Surveys were mailed to a total of 718 stakeholders from among the ranks of transit agency CEOs, board members, transit agency safety professionals, national and local unions representing transit employees, federal agency and transit industry association representatives (such as APTA), state oversight agencies, NTSB, FTA, FRA, and CTAA. Data from this survey allowed researchers to gain a sense of the state of safety culture and practices across the transit Safety Culture Within Public Transportation

24 industry. The survey also identified public transportation sys- tems recognized by peers for their positive safety culture that could be further examined through case studies and inter- views. The survey did not solicit participants’ perceptions of the effectiveness of safety culture within their own organiza- tions; rather, it inquired about peer transit agencies thought to have positive safety cultures. The initial draft of the survey instrument was sent to the project panel for review and was revised based on panel recom- mendations. The survey was carried out as an Internet survey through SurveyGizmo. Once the contact list was complete, the link to an online survey site was distributed to 718 potential respondents by e-mail. Multiple e-mail and phone reminders were conveyed to the stakeholders in the months that followed. Response Rate and Sample Description By the conclusion of the survey period, 137 complete responses had been received, representing a 19.1% overall response rate. Another 102 contacts, or 14.2%, responded to a number of the introductory questions without completing the survey, yielding a total response rate (complete and incom- plete) of 33.3%. All analysis, figures, and tables are based on complete responses only. This 19.1% response rate was lower than was hoped for, but it is within the expected range for online surveys found in TCRP Synthesis 69: Web-Based Survey Techniques (Spitz et al., 2006). Figure 5 shows the number and percentage of complete responses. National and local labor representatives account for 26% of the responses; transit CEOs, 25%; safety profes- sionals, 22%; safety oversight agency and industry associa- tion representatives, 11%; and transit board members, 4%. Figure 6 displays the number of respondents by transit mode, with some indicating more than one mode. Fixed-route bus and paratransit represent the two largest groups. Of the 122 respondents who indicated any transit agency affiliation, 98 (80%) represent multimodal agencies. Of the 118 respon- dents who indicated the ridership category of their transit agencies, 45 (38%) represent transit agencies with less than 20 million in annual ridership. The statistical confidence limit for the overall population is plus or minus 7.61%, which means that it is 95% likely that the results from the survey are within plus or minus 7.61% of reality. Statistical confidence limits for the seven constitu- ent subgroups, however, are much wider and therefore much less useful: CEOs, 14.9%; transit board members, 39.0%; safety professionals, 15.8%; labor representatives, 15.9%; and safety oversight agency and industry representatives, 17.2%. The drop in confidence limits is because of the much smaller target populations and correspondingly smaller number of completed survey responses within each subgroup. As a result, the recorded differences in perspective between these subgroups are not statistically significant enough to support any conclusions from the survey results alone. To be credible, such differences have to be supported by interviews, surveys, and other sources of information. With these considerations in mind, the research team decided to use the initial survey results subject to confirmation from the transit agency mini–case studies and in-depth interviews. In determining key components of safety culture in transit, the transit stakeholder survey and the case study interviews turned out to be mutually reinforcing. This mutual reinforcement confirms the value of combining different methodological approaches in social science research, particularly the value of combining quantitative and qualitative methodologies to address related questions from different perspectives. Social science researchers have long held the consensus that this type of “integration leads to maximizing the strengths of the quan- titative and qualitative data and minimizing their weaknesses” (Creswell et al., 2011). CEOs 25% Board Members 4% Safety Professionals 22% Labor Representatives (Local) 22% Labor Representatives (National) 4% Oversight Agency and Industry Representatives 11% Others 12% Figure 5. Number and percent of respondents by role.

25 In statistics and probability theory, the standard deviation measures the amount of variation or dispersion from the average score. A low standard deviation indicates that the data points tend to be very close to the mean; a high stan- dard deviation indicates that the data points are spread out over a large range of values. The largest differences in ranking concern three elements: the importance of accident investi- gation focusing on preventing recurrence instead of blame, reporting near misses without fear of blame, and the effec- tiveness of employee safety reward programs. These rank- ings were used, in conjunction with literature search results, to draft mini–case study interview questions. Additional Key Elements or Components At the end of the survey, participants were asked to write in as many as four additional key elements of safety cul- ture. These additional comments confirm, improve on, and enrich many of the original 15 key elements and were later built into the case study interview questions. Common themes are listed here in the order of the frequency of related comments: • Employee and union involvement, through mechanisms such as safety committees, peer communication, joint problem solving, and joint determination of training (four management, five union, and four unidentified comments). • Systems for accountability that emphasize procedural justice, reinforced by organizational structure, employee evaluations, and rewards (four management, one safety oversight agency, and two unidentified comments). Key Elements of Positive Safety Culture Survey respondents were asked to rate the importance of 15 potential key elements or components (see full text in Table 2) of a positive safety culture, particularly within the transit environment. The list of key elements was generated based on preliminary findings from the literature review and additional project panel input. As the literature review progressed, a number of key elements also evolved. Many of these changes are also suggested by the survey respondents in their open-ended comments. Table 2 shows the 15 key elements ranked from most important to least important as rated by the survey respon- dents. There were 13 components of safety culture listed in this section of the survey, with the respondent allowed to add up to four more. Each statement was ranked as “very impor- tant,” “important,” “moderately important,” “of little impor- tance,” “unimportant,” or “don’t know.” Numerical scores were assigned, with 1 for “very important” through 5 for “unimportant.” The smaller the average score, the higher the ranking, with an average of 1 representing a perfect score. As the chart demonstrates, the three factors considered most important to a positive safety culture in this survey are: • Management and employee commitment to safety as the top priority; • Adequate training; and • Open, effective communication. The standard deviation column in Table 2 gives some indi- cation of disagreement among respondents to each question. 105 23 41 25 86 25 0 20 40 60 80 100 120 Fixed-Route Buses Heavy Rail Light Rail Commuter Rail Paratransit (Demand Response) Other Figure 6. Number of respondents by mode.

26 • Effective information or data management to measure performance and track goals, using key performance indi- cators (two management and two union comments). • Improve safety communication using a myriad of technolo- gies and tools (three management and one union comment). Based on these additional comments, the top-ranked key element was modified to read “Safety as a core value to which management and employees are committed.” The updated key elements can be found in Table 3. Some respondents used this section to voice particular con- cerns with how safety is perceived and carried out within their • Safety as an organizational value consistently emphasized by leadership and recognized as everyone’s role and respon- sibility (four management, one safety oversight agency, and one union comment). • Risk reduction through a systematic approach to identify, analyze, control, monitor, and report hazards, rather than legal compliance or discipline (five management, one safety oversight, and one union comment). • Training and retraining, including mentoring (retraining following an accident is integral to progressive discipline) and educating the public (4 management and 2 union comments). Key Elements Ranked by Total Average Importance Total Responses Average Ranking Standard Deviation of Ranking (1) Safety is recognized as the highest organizational priority, and both management and employees are committed to that priority. 1.32 0.64 (2) Adequate training is provided so employees have an understanding of how to perform their job duties in a safe manner. 1.66 0.92 (3) There is open, frequent, and effective communication on safety. 1.71 1.00 (4) Adequate financial and human resources are dedicated to ensure the safety and health of employees. 1.81 0.88 (5) Management and employees are willing to interrupt schedules and service for safety reasons. 1.82 1.02 (6) There is competence within the organization to draw appropriate conclusions from safety information. 1.83 1.00 (7) The organization takes action visible to employees on all reported safety issues. 1.87 1.04 (8) The organization collects and analyzes relevant data and actively disseminates safety information. 1.90 0.96 (9) There is significant employee involvement in the continuous improvement of safety policies and rules. 1.90 0.98 (10) The culture of safety is deeply ingrained in the organization, and no leadership transition from either management or union will likely change that commitment to safety. 1.90 1.07 (11) Accidents are reviewed from the perspective of future prevention rather than the focus being exclusively on finding someone to blame. 1.96 1.14 (12) There is a high level of trust between management and frontline staff. 1.99 1.09 (13) Employees are encouraged to report near misses and other safety events without fear of blame. 2.00 1.18 (14) Where there is union representation, the union is continually involved in safety processes, including joint safety data collection, analysis, and problem solving. 2.03 1.04 (15) Employees are rewarded for reinforcing safety at work. 2.23 1.14 Average for all elements 1.86 1.01 Table 2. Survey respondents’ ranking of key safety culture elements.

27 select transit agencies for follow-up interviews and mini–case studies. Respondents were not allowed to nominate their own transit agencies. A list of 20 agencies was generated from those that were mentioned most frequently. To complement information provided by industry stakeholders, key performance indi- cators of safety culture were investigated at these 20 transit agencies. Anecdotal evidence, such as existing industry infor- mation collected through presentations at APTA and other industry conferences, and industry safety award entry docu- ments (such as APTA’s Bus Safety and Security Excellence Award), were entered into a database. Much of this evidence pertains to indicators such as organizational safety commu- nication and learning, safety training, labor and manage- ment cooperation around safety problem solving, and safety reporting practices. To gauge transit agency safety perfor- mance using publicly available indicators, safety statistics from the National Transit Database (NTD) were extracted, including annual fatalities minus suicides, accidents minus own organizations or the organizations they oversee. For exam- ple, an oversight agency representative said that “transit agencies are selective, not transparent in regard to information sharing with state safety oversight,” “transit agencies are more focused on achieving minimum compliance,” and “transit agencies real- ize that there are no mandates that keep them systematically accountable.” Several union representatives expressed concerns about the lack of effective involvement of the union and rank- and-file employees when it comes to safety training, implemen- tation of safety procedures, and employee discipline. One union respondent observed that the safety department is constantly overruled by the operations department in that transit agency. Identification of Transit Agencies with Positive Safety Cultures At the end of the stakeholder survey, respondents were asked to identify transit agencies that they believed had a posi- tive safety culture. This identification process was used to help Mini–Case Study Reporting Categories Corresponding Key Elements in Rank Order from Transit Stakeholder Survey A. Strong leadership, management, and organizational commitment to safety 1. Safety as a core value to which management and employees are committed 4. Adequate resources dedicated to safety 5. Management and employees willing to interrupt service for safety 10. Safety culture stable through leadership transitions B. Employee/union shared ownership and participation 9. Significant employee involvement 14. Union involvement in safety process C. Effective safety communication (informed culture) 3. Open and effective safety communication D. Proactive use of safety data, key indicators, and benchmarking (informed/learning/flexible culture) 8. Using safety metrics and leading and lagging indicators to gauge safety performance E. Organizational learning (learning/flexible culture) 2. Adequate safety training provided 6. Organizational safety competence (recruitment and succession planning) F. Consistent safety reporting and investigation for prevention (informed/reporting/learning/just culture) 7. Visible action on all reported safety issues 11. Accident focus is preventing recurrence 13. Near-miss accident reporting in place (and data collected) G. Employee recognition and rewards (just culture) 15. Employee safety performance rewards H. High level of organizational trust (just/informed culture) 12. High level of trust between management and workers Table 3. Mapping of 15 key elements against eight categories.

28 nine locations. Input by union representatives and frontline employees was used either to confirm or qualify the substance of information provided by management representatives. Overview of Findings Detailed mini–case study reports were prepared for each of the nine locations for confidential internal use and for possible identification of potential best practices. Due to the confidentiality of the mini–case studies, the detailed mini– case study reports containing transit agency names and iden- tifying persons interviewed are not included in this report. Additional information using transit agency code names is found in Appendix B. Instead, anonymous summaries and comparisons are presented here to establish major findings. Transit agencies are coded from “A” to “I.” The following sec- tions present (1) an overview of the current safety culture sta- tus at the nine mini–case study locations, drawing from views of the diverse groups of management, employee, and labor representatives interviewed, and (2) a summary analysis of transit agency characteristics across the key elements identi- fied from the literature review and the stakeholder survey. Safety Culture in the Mini–Case Studies Safety culture is a journey, not a destination. Although those managers and union representatives interviewed gen- erally considered the safety culture within their own organi- zations to be positive or somewhat positive, all agreed that there was ample room for improvement. One person said of the journey to safety culture: “It’s a way of life. Maintaining a safety culture requires vigilance and stick-to-itiveness and an ongoing commitment and realization that your work will never be done.” In a number of locations, tragic accidents involving passen- gers and the public were the prime motivators for a renewed focus on safety and safety culture improvements. The memory of these events serves as a constant reminder of the paramount importance of safety in public transportation for everyone in the organization, from top management, the board, and union leadership to supervisors, operators, and shop-floor mechan- ics. All have come to realize that safety needs to be more deeply rooted in the organizational culture and supported by initia- tives such as safety stand-downs, more effective communica- tion, training and retraining, public awareness campaigns, vehicle redesign, and new technologies that help alert pedes- trians and operators. Within these transit agencies, there is a sense of vulnerability, as is frequently the case in high-reliability organizations—the sense that, even if the current safety cul- ture has reached its highest historical level, it is an ongoing effort, and they can never assume they are done. suicides, and incidents from 2008 to 2010. Consideration was also given to the size and geographic distribution of the tran- sit agencies. Using a combination of these factors, the list was narrowed down to 11 potential mini–case study candidates. Transit Industry Mini–Case Studies Mini–Case Study Design A major focus of the research involved mini–case studies of transit agencies identified by industry stakeholders as pos- sessing a positive safety culture. Interviews were sought with CEOs, safety directors and officers, mid-level managers from operations and maintenance, local union presidents and safety representatives, and frontline employees. Based on the litera- ture review and the earlier survey results, a semi-structured questionnaire was developed to collect the perspectives of those interviewed. Two agencies were not responsive after several attempts, and therefore mini–case studies were not pursued for these agencies. The final mini–case study locations were: • Three large transit agencies: two on the West Coast and one on the East Coast, • Three medium-sized transit agencies: two on the West Coast and one on the East Coast, and • Three small transit agencies: one on the West Coast and two in the Midwest. A list of top-level management and union contacts was compiled, and the research team e-mailed or phoned the individuals from the selected transit agencies, and in some cases visited them in person, seeking their participation in the mini–case study research. Most of the contacts responded positively. Where possible, the first interviews for each mini– case study were scheduled with the CEOs/general managers (GMs) and the local union presidents; they, in turn, provided a list of additional contacts. Interviews were recorded when consent was granted. To allow candid opinions to be shared, participants were assured of the confidentiality of the mini– case studies and their statements. Beyond the background materials previously collected on each mini–case study, interviewers requested specific docu- mentation of safety performance records, quarterly or annual safety reports, strategic plans that related to safety, safety com- mittee meeting minutes, safety journals or newsletters, and so forth. As interviews were being completed, research team members analyzed transcripts and documents in preparation for writing the mini–case study reports. By February 2012, a total of 64 individuals had partici- pated in interviews, including 43 management representa- tives and 21 union representatives and frontline workers from

29 tion. Comparing operations across locations, transit agencies demonstrate varying degrees of emphasis and focus on the 15 key elements of safety culture that were identified in the transit stakeholder survey. • Recognizing safety as a top priority and core organi- zational value is a common theme in all these examples of positive transit safety culture. Safety as a core value is ingrained throughout these transit agencies’ organizational cultures, is largely unaffected by changes on the manage- ment or union leadership, and is reflected in practices such as putting safety resource needs at the top of the list of pri- orities and being willing to halt services for safety reasons. • Open and frequent communication and feedback is a widespread practice in all the mini–case studies, except for one transit agency where communication still tends to be top-down. • Hiring and training to promote safety. These transit agencies focus on hiring employees with a safety mind-set and continue to cultivate their competence in safety behav- iors and decision making through training and retraining. Transit agencies A and G excel in safety training provided to employees throughout their organizations. • Labor management cooperation. Transit agencies D, E, G, and I use joint labor–management safety committees as the primary platform for safety problem solving. Accord- ing to most of the management and labor respondents at these mini–case study locations with joint safety commit- tee experience, actively involving rank-and-file employees contributes to identifying day-to-day safety deficiencies and improving safety awareness and safety procedures. • Performance indicators. Analysis of lagging indicators is conducted frequently at all the case study transit agencies and shared among different groups of employees to identify trends and pinpoint problems. Some of these agencies have just started using leading indicators such as internal safety inspections/audits and employee attitude surveys to gauge the level of safety culture and drive proactive changes. Tran- sit agencies A, E, and F lead in their continuous effort to monitor and analyze safety data for problem solving. • Trust. A relatively high level of trust between management and workers is found in five of the nine mini–case study transit agencies. Several of these transit agencies reported varying degrees of trust across different modes, divisions, and organizational functions. • Near-miss reporting. In about half of the case studies, accident investigation is focused on prevention of recur- rences rather than finding the party to blame or discipline. When unsafe conditions are reported, these transit agen- cies take actions visible to employees to rectify the situa- tion or provide full explanation if no action can be taken. Near-miss reporting systems are being implemented at Labor representatives generally rated the current status of their agency’s safety culture somewhat lower than did their management counterparts. Union responses emphasized that safety culture and safety results were sometimes compromised by the pressure for on-time performance, the growing fre- quency of assaults on operators, and the lack of sufficient or effective involvement by employee representatives in problem solving and decision making. Safety Culture Mini–Case Studies Conclusions Themes arising from analysis of the mini–case study mate- rials can be summarized by the eight components or elements of safety culture: • Strong leadership, management, and organizational com- mitment to safety; • Employee/union shared ownership and participation; • Effective safety communication; • Proactive use of safety data, key indicators, and benchmarking; • Organizational learning; • Consistent safety reporting and investigating for prevention; • Employee recognition and rewards; and • High level of organizational trust. Detailed mapping of the 15 key elements taken from the survey against these eight categories is provided in Table 3. The first two components can be considered key underlying perceptions, attitudes, and behaviors that drive safety culture, and, as shown in the parentheticals in the first column of the table, the last six correspond to the five subcultures in Reason’s model. The eight categories also sum up the 15 key elements and additional stakeholder comments from the stakeholder survey, with some categories covering multiple key elements. For example, safety as a core value, management and employee commitment, and adequate resources have been clustered under the “strong leadership, management, and organizational commitment to safety” category. Mini–Case Study Analysis A comparative analysis of operating characteristics identi- fied at the mini–case studies is presented in a series of tables in Appendix B. This analysis, which shows the range of imple- mentation found across these nine transit agencies, is sum- marized in the following. Common Themes Transit safety culture is multifaceted, integrating many elements, and has many available points of entry and initia-

30 study transit agencies. However, some transit managers see the value of these rewards systems as limited in positively affecting day-to-day safety awareness and performance. Cross-Cutting Themes Review of the mini–case studies reveals important cross- cutting themes. For example, in all of the mini–case studies, transit agencies demonstrated most of the 15 key elements from the stakeholder survey listed in Table 4. In fact, with the exception of one location (C), which implemented only four agencies (B, E, G, and I), all of which operate FRA- regulated commuter rail service, but near-miss reporting is fully embraced by managers and employees only in agen- cies E and I. These near-miss reporting systems have pro- duced exceptional results. Individuals interviewed at these agencies mentioned open, frequent, and effective com- munication of safety messages, issues, and data analysis throughout the organization as the backbone supporting their safety culture. • Reward and recognition systems for employees with out- standing safety performance are used at many of the case Key Elements Agency Code A (S) B (M) C (S) D (L) E (L) F (M) G (S) H (L) I (M) Agencies Total*** 1. Safety as a core value that management and employees are committed to (updated using stakeholder survey comments) 9 2. Adequate safety training provided 9 3. Open and effective safety communication 8 4. Adequate resources dedicated to safety 8 5. Management and employees willing to interrupt service for safety 9 6. Organizational safety competence 9 7. Visible action on all reported safety issues + + + 5 8. Significant employee involvement + 8 9. Using safety metrics and leading and lagging indicators to gauge safety performance 9 10. Safety culture stable through leadership transitions + 7 11. Accident focus is preventing recurrence + + 6 12. High level of trust between management and workers + + * + 5 13. Near-miss accident reporting in place (and data collected) ** ** ** ** ** 4 14. Union involvement in safety process n/a + + 6 15. Employee safety performance rewards 9 Total*** 13 13 7 12 15 10 13 13 15 Legend: = key element implemented; = key element emphasized; + = missing element linked to trust; = missing elements not linked to trust; S = small; M = medium; L = large. * High trust only in transit rail division. ** These agencies do not directly operate commuter rail and are thus not regulated by the FRA to have a near-miss reporting system. *** Double checks are counted as one, and row 13 (near-miss accident reporting in place) is not counted because this is implemented mostly in rail divisions that are regulated by the FRA. Table 4. Key elements of safety culture observed in nine mini–case studies.

31 lying objective conditions: visible action being taken on all reported safety issues, accident investigation focused on pre- vention, positive worker and union involvement, open and effective safety communication, and stable leadership over time. While mini–case study interviews show a greater skepti- cism among frontline workers and union representatives, the strongest safety cultures produce a stronger convergence of perceptions between labor and management. How do they do this? The answer can be seen in further analysis of the inter- play among trust and the factors that generate trust—most fundamentally, visible action consistently being taken on all reported safety issues, a steady focus on preventing future accident risk, and union and workforce involvement in the agency safety process. This trust-centered complex accounts not only for differences in labor perception but also for the bulk of the recorded missing elements of safety culture in these nine transit agencies—16 out of 18, or 88%. Conclusions This project’s research revealed consistent results between the literature review, the quantitative survey of transit stake- holders, and the mini–case studies and interviews. The three elements of transit safety culture identified receiving the great- est consensus were: • Management and employee commitment to safety as a core value, • Adequate training, and • Open, effective communication. The stakeholder survey found the greatest divergence of stakeholder rankings on three topics: the importance of acci- dent investigation focusing on preventing recurrence instead of on blame, the role of a near-miss reporting system, and the effectiveness of employee safety reward programs. Survey data indicate that perceptions of safety culture are weaker at smaller agencies. The selection of transit agencies for mini–case studies was based in part on stakeholder survey peer nominations and anecdotal supporting evidence, such as industry recognition for outstanding safety performance. However, transit agencies that did not rank well in the safety statistics from the NTD (annual fatalities minus suicides, accidents minus suicides, and incidents from 2008 to 2010) were eliminated. Mini–case studies that focused on safety culture at nine large, medium, and small transit agencies revealed a number of leading areas for developing a positive safety culture, including: • Leadership and organizational commitment, • Organizational learning, • Open and frequent safety communication, seven of the 15 elements, all the mini–case study locations implemented at least 10 elements. Seven of the elements were found at all nine transit agencies, and another three elements were found at eight of the transit agencies. Just as there are different points of entry for building a transit safety culture (for example, through emphasizing, training, partnerships, or systematically reducing risks), the mini–case studies show that these transit agencies have tied together and integrated safety culture concepts and practices. The key is sustained integration with a distinctive narrative that brings together multiple elements within the organization in a self-reinforcing culture. The specific language can vary, but the coherence of the message with mutually reinforcing practices needs to be sustained over time to generate a deeply embedded safety culture. It is noteworthy that the labor representatives inter- viewed in the mini–case studies generally had a more skep- tical and even at times negative view of transit agency safety culture compared to transit agency managers and execu- tives. The absence of a high level of trust between man- agement and employees was found in three of these nine mini–case studies—B, C, and F—and is associated with the following three related elements: • Visible action not being taken on all reported safety issues (three of three cases); • Accident investigation not focused on preventing recurrence (two of three cases); and • Absence of union involvement in safety processes (two of three cases). Low trust, inconsistent accident follow-up aimed at prevent- ing future recurrences, and low employee and union involve- ment are the most significant negative factors detracting from truly outstanding safety culture in these transit agencies. Numerically, this cluster accounts for 13 of the 18 miss- ing features (72%) out of the total of 135 total matches in Table 4’s matrix of 15 safety culture elements mapped across the nine transit agencies. (The element of near-miss report- ing is excluded from this analysis because it is relatively new, an innovation most frequently implemented today in FRA- regulated commuter rail operations that are found in only a minority of the nine transit agencies). The other five missing features occur across four agencies. They are a lack of accident investigation focused on prevention at two agencies (G and H), visible action not being taken on all reported safety issues at one agency (H), and a lack of adequate resources dedicated to safety and lack of stability in safety culture across leader- ship transition at another agency (D). These patterns indicate that the largest group of missing features at these transit agencies is a complex that appears to link trust between labor and management with under-

32 linked to trust. In the interviews, they expressed the opin- ion that safety is sometimes compromised by the pressure for on-time performance. The degree of labor skepticism is greatly reduced by a complex of practices that generate trust, including, most prominently, taking consistent action on all identified risk factors, focusing on prevention rather than blame, and workforce and union involvement in safety processes. This survey and these mini–case studies help define safety culture in transit. They have also led to the design of a draft self-assessment survey instrument. Finally, this research pro- cess identified a number of innovative best practices for pos- sible future adoption by other transit agencies. • A systematic approach to safety reporting and accident investigation, • Active involvement of union and rank-and-file employees, and • Data-driven safety problem solving. The most important of these features are found at all of the nine transit agencies and provide several distinct points of entry for safety culture initiatives. Union representatives and workers interviewed expressed less sanguine views than others about the general status of safety culture within their workplaces, particularly in agen- cies where safety culture fell short on a complex of elements