This article was first distributed in Professional Safety in June 2010.
To see the full article with images, charts and figures, click here .
Click here to view infographic
Who would have thought that with general injury rates declining the rate of fatal workplace injuries would not decline at a similar or faster pace? (BLS, 2008; Hamalainen, Takala & Saarela, 2006; HSE, 2008; Takala, 2005). And, who could have predicted that locations with superior safety history would suddenly experience fatalities, life-altering injuries or catastrophic events? Many organizations have been caught off guard because they were relying on injury rate performance to predict future success (Manuele, 2008). As occupational injury data show, most of these events do not result from unknown or unpredictable circumstances, or from weird occurrences. The major causes of life-altering injuries and fatalities continue to include the basics: falls, failures of key permitting systems such as deenergizing equipment, confined space entry, line breaking and mishaps related to mobile equipment (NIOSH, 2006; BLS, 2008).
In the experience of the authors and their colleagues, these life-altering events occur when companies accept deviation as normal, fail to manage control systems and tolerate substandard processes. In other words, these incidents were not inevitable, nor were the resultant pain and suffering. The exposures were known and the root causes shared common threads. Had organizations followed a basic prevention mechanism and ensured that prevention methods were robustly applied, most of these events could have been avoided. Above all, the prevention mechanism for stopping life-altering injuries and catastrophic events requires a rigorous level of oversight and participation from senior leadership.
This article provides guidelines that senior leaders can use to maintain a sense of vulnerability even when no disastrous event has occurred recently and especially when the lagging indicators are “favorable.” Additionally, it suggests actions that senior leaders should take if a fatality or catastrophic event occurs.
The Connection Between Leadership & Workplace Safety
The relative infrequency of fatalities and other serious events can cause them to seem random, beyond any reasonable degree of anticipation and prevention. In fact, most of these events result from high-energy potential exposures that are identifiable, measurable and manageable. The lessons of prominent incidents such as the space shuttle Columbia, Oxy’s Piper Alpha, Esso Longford and BP Texas City, as well as lessons from single fatality events, are that alongside the proximate causes of each incident is an underlying fabric of systems, mechanisms and culture that allowed risk in the workplace to persist and often to become acceptable (CAIB, 2001; Cullen, 1990; Baker, 2007; Hopkins, 2000).
All components of an organization’s safety fabric lend themselves to senior leaders’ influence and intervention.
The decisions leaders make, the things they say, the systems they implement and oversee, and the value they place on safety with respect to other objectives affect:
work practices and sustained behaviors that increase or reduce hazards;
the level to which the culture supports safety objectives and activities;
workers’ own interest in safety and safety activities.
Creating an organization that eliminates fatalities and life-altering injuries cannot be delegated. It requires the integrated involvement of the entire organization, from the CEO to each worker.
Culture, Leadership & Safety
Among the strongest indicators of safety performance are workplace culture and leadership quality (Hidley, 1998). An extensive body of research identifies nine measurable cultural characteristics that, in addition to predicting safety outcomes (such as level of safe behavior, injury rates and event reporting) (Hofmann, 1999; Bell,O’Connell, Reeder, et al., 2008), have been shown to predict variables indirectly related to safety, such as turnover (Ferris, 1985), citizenship behavior (Coyle-Shapiro & Conway, 2005; Konovsky & Pugh, 1994), trust in the organization (Kickul, Gundry & Posig, 2004), and trust of employees, innovation and creativity (Ruppel & Harrington, 2000). These characteristics are:
Procedural justice: Fairness and transparency of supervisor’s decision-making process.
Leader-member exchange: Level of mutual trust and respect between employee and supervisor. Employees treated with dignity.
Management credibility: Management actions consistent with words.
Perceived organizational support: Employees perceive that the organization values them.
Work group relations: Level of mutual trust and respect among coworkers. Teamwork: Ability of the work group to effectively get things done.
Organizational value for safety: Extent to which employees perceive that the organization is serious about safety performance.
Upward communication: Extent to which safety concerns, suggestions and ideas flow upward through the organization.
Approaching others: Extent to which workers are comfortable speaking to one another about safety.
How employees perceive these nine dimensions has been shown to correlate to injury rate. Figure 1 illustrates the results of a study of 94 organizations (representing eight countries and 18 industries) that assessed these dimensions (Bell, et al., 2008). Results show that organizations where employees rate these dimensions consistently more positively across all scales had significantly lower occupational injury rates compared to those that scored consistently more negatively. Organizations in the middle had injury rates between the high and low groups.
The difference between the three groups is statistically significant: (df(94), -.331, p < .01). A similar study of low-injury-rate companies (those with an occupational injury rate less than 3.0) shows the same relationship holds true.
Not surprisingly, leadership has been shown to influence culture. Figure 2 illustrates the results of a study that examined the relationships between how the top site-level leader is perceived by direct reports on safety leadership best practices, defined as vision, credibility, action orientation, collaboration, communication, recognition and feedback, and accountability (Krause & Weekley, 2005), and site-level scores on the nine dimensions described. The study showed strong positive correlations between subordinate ratings of each best practice and overall ratings of each dimension of organizational culture. The leadership overall score (the aggregate of the seven best practices) predicted culture overall (Bell, et al., 2008).
In addition to culture, leaders must consider the climate of the organization with respect to catastrophic events. Climate refers to what is most important in the short term. It is influenced by a host of factors, but most often the biggest factor is what leadership is focusing on and discussing. Safety performance can suffer when the focus is on short-term gains and not long-term sustainability. When organizations are under cost-cutting pressures or are being pushed in other performance areas, such as production velocity, then leadership typically focuses less on safety.
During these times, basic, underlying safety systems can be undermined or eliminated. While in some instances the effect can be seen immediately through the relaxation of safety standards, other changes may not be seen until much later. These insidious, in-the-moment changes can lead to significant events years later. The leaders who made these changes may be long gone and may not understand the ultimate impact of their decisions.
Eight Key Questions
While no one can guarantee that an organization will never have a serious incident, senior leaders can employ behaviors and practices to set the tone and climate around how vigorously the organization creates and sustains a focus on the prevention of exposure to serious hazards (Erickson, 1997; Stricoff, 2007). During their combined experience of nearly 100 years in the safety field, the authors have noted a common pattern of behaviors and practices among leaders of organizations with outstanding safety performance.
Through this work, and through examination of their colleagues’ work with more than 7,000 leaders among 500 leadership teams, the authors have noted a distinct difference in the leadership practices of organizations that struggle with serious and fatal events compared to those that do not.
For example, in organizations where serious injuries were an issue, senior leaders seldom, if ever, knew the company’s history of serious incidents, while leaders in higher-performing organizations tended to know the names of injured employees and expressed concern for them as individuals.
By comparing this work to independent reports of serious workplace events, such as Piper Alpha, Esso Longford and Columbia, the authors identified eight qualities that tend to exist among leaders of organizations that maintain a sense of vulnerability and that have sustainable systems to recognize risk before an event happens.
These qualities are framed as questions that senior leaders can ask themselves to assess whether they and their organizations are doing what is necessary to prevent serious incidents. All members of the senior leadership team, including all functions, should be able to answer these questions. Having continual command of the answers to these questions strongly suggests that a leader is doing the right things to sponsor serious incident prevention.
QUESTION 1: When was the last fatality, life altering injury or catastrophic event in this organization and what were the victims’ names?
Fundamentally, prevention takes an emotional commitment. In the authors’ experience, the primary motive driving leaders to improve safety is empathy and compassion. Leaders with this commitment are more effective at ensuring sustainability in fatality prevention (Spigener, 2007).
Talking only about metrics such as frequency rates depersonalizes the information. When leaders think or talk mostly about injury classification (e.g., lost time, restricted), they can quickly forget about the event’s impact on the employee, his/her family and coworkers. It also makes leaders less likely to have the emotional connection and commitment necessary to sustain their focus. Life-altering injuries profoundly affect a person’s ability to lead the type of life s/he did before the event. A worker disfigured from a burn is not worried about injury classification; s/he is a person whose life has just changed.
One factor that helps determine success in safety is top management’s demonstrated concern (Griffiths, 1985; Petersen 2000). At the same time, catastrophic events, fatalities and life-altering injuries are relatively rare. As time passes, the emotional impact associated with the event diminishes, and the urgency of the rigorous application of prevention mechanisms wanes. Additionally, as new leaders join the organization, they may be unable to relate personally to past serious events and to the emotions associated with them. Keeping this personal information in front of the leaders decreases the likelihood that they will become detached from the prevention process and helps ensure sufficient motivation to lead safety performance.
QUESTION 2: Do any of my behaviors as a leader suggest, even unintentionally, that fatalities are acceptable and a part of doing business? Do any of our systems suggest the same?
This question may seem odd in comparison with the first. How could anyone suggest that a responsible member of senior leadership would send messages that fatalities are acceptable? And who would allow systems that imply fatalities are acceptable or that the factors, which might contribute to these events, would be tolerated?
These questions are not about intent. No reasonable person would intentionally or explicitly send this message. However, actions can be misinterpreted or misread, and systems can indirectly reward undesired behaviors. Senior leaders who open themselves up to the scrutiny of this question position themselves to uncover unintentional drivers of undesired behaviors. Consider what messages might be received in the following examples:
Measurement of safety performance is based on a single indicator of performance (e.g., Baker, 2007), such as the OSHA incident rate. In such a system, a fatality and a small cut requiring a stitch would be counted equally. Which has a greater consequence to the location manager: having one fatality or having three “minor” OSHA recordable injuries?
Use of a compensation system that allows leaders to receive a safety bonus when they meet their case rate number, even if a fatality has occurred.
Implementation of an employee incentive program that bases pay on injury rates (e.g., Krause & McCorquodale, 1996).
Senior leaders have retrospective discussions with their direct reports when an injury occurs, yet rarely discuss and review proactive prevention activities.
Leadership sets up a system that routinely forgives people for making fatal risk errors.
Management safety practices are among the best predictors of accident rates and compliance with safety behaviors (Hayes, Perander, Smecko, et al., 1998). What leaders emphasize, intentionally or not, affects safety outcomes (Diaz & Cabrera, 1997). Not surprisingly, among organizations with which the authors have worked, leaders who are aware of how their decisions and actions affect the organization tend to be more sensitive to safety issues even when the subject at hand is not specific to safety.
QUESTION 3: What is the difference between process safety and personal safety? Do our facilities require compliance with the process safety standards?
While senior leaders are not expected to be the SH&E professionals in their organization, leaders must understand several fundamental concepts. Most critical is the difference between process safety and personal safety.
Process safety refers to the prevention of catastrophic events associated with the storage, handling, production and use of hazardous chemicals.
QUESTION 4: What are the major sources of exposure that have caused or could cause major events (fatalities, life-altering injuries, fires and explosions)?
In addition to understanding the difference between process safety and personal safety and the typical elements that make up each, senior leadership must understand the exposures associated with the serious mishaps that have occurred and the exposures with the greatest potential for major events.
This knowledge should be combined with an understanding of what would most likely cause these exposures to persist. When armed with this information, senior leaders are in a position to routinely review the indicators of whether these exposures are being managed. Additionally, knowing the major sources of exposure positions leaders to ask about exposure, for example, when visiting locations or when considering budget requests.
When looking for major exposures, ask whether the organization has any safety rules that can result in termination if violated. Many organizations call these cardinal safety rules, crucial safety decisions, life critical safety procedures or similar names. These rules point to exposures that are known to cause fatalities or catastrophic events (e.g., fall protection, hot work permits, lockout/tagout, confined space entry). They are considered so safety critical that the penalty for violation is the harshest.
Having the rule and threatening dismissal are not enough. These exposures must be routinely monitored and tracked to ensure behavioral reliability. They are the focus of a fatality prevention program. Similar to the importance of understanding the distinction between process and personal safety, knowledge of the organization’s safety landscape allows leaders to shape safety functioning, which is associated with worker compliance and positive safety outcomes (Simard & Marchand, 1997).
QUESTION 5: What leading metrics do we track to ensure that our fatality prevention mechanisms are robust?
Many leadership teams excel at reviewing lagging indicators that provide a sense of how well the organization has performed. However, what can be learned from this information is limited to how the organization ranks against others and whether performance has stayed the same, declined or improved. Lagging indicators do not address the status of prevention activities (e.g., whether they have eroded and increased exposure).
One paradigm change that needs to be understood is that the “safety triangle” which has been relied on for decades is the wrong model for thinking about fatality prevention. This triangle suggests that stopping small injuries will produce a corresponding reduction in more serious events (Heinrich, 1959). This concept is not absolutely true for fatalities and life-altering events. These events may not be preceded by a series of more minor mishaps; instead, the probability of a serious event is much higher from these exposures. Stopping eye injuries by getting employees to wear safety glasses with side shields will not reduce fatal risk exposure.
Leading indicators vary by organization (e.g., Edkins & Pollack, 1996; Petersen, 2000), but it is increasingly common to see successful organizations using the following types:
Number of near-miss, high-energy/high-potential events. These are events in which a fatality could have occurred, but due to the misalignment of one factor, the organization suffered a near-miss.
Corporate audit results associated with the fatality and catastrophic event elements. Results from corporate SH&E audits provide an understanding of the status of closure on deficiencies in these elements.
The status of compliance with single-layer prevention barriers (Reason, 1997). A single-layer barrier is a safety measure that supplies protection based on a single system, often at one of the lowest levels of the hierarchy of controls. For example, suppose an organization relies on employees to deenergize equipment routinely, reliably and correctly each time, without having a system to measure the level of variation. Such single-layer systems have a high likelihood of failure, unless another layer is added on top of this barrier. High-quality verification audits would provide a second layer of protection and likely would reduce variation.
Managing safety on par with other business functions requires that leaders have valid performance measures by which to assess progress and drive strategy (Arezes & Miguel, 2003). In addition, leaders need measures of root causal factors of serious events (Manuele, 2008; Reason, 1997) by which they can monitor the conditions, systems and practices for variation that increase exposure to these events.
QUESTION 6: How do we know whether we are building strong safety leadership at all levels and creating a culture of commitment?
In addition to asking the right questions and having the right systems and activities, leaders must assess whether they are creating a culture of commitment to the organization’s value for safety. A culture of commitment is defined as an environment in which employees at all levels will do what is right for themselves, their boss and the organization, even when they would personally gain from noncompliance, because they have bought into and connected to the organization and leaders’ vision. They do so even when no one is around to encourage compliance.
Senior leaders must lead the way and sponsor such a culture. For a senior leadership team to successfully manage fatality prevention, it must actively help create the type of culture desired and be emotionally committed to achieving it. Workers do not create such a culture; it is developed and sustained by the organization’s leadership (Krause, 2005).
One way for leaders to help create this culture is to walk through the operation and ask workers and their front-line supervisors questions about what they observe:
Have you experienced or heard recently of any significant close calls, where an inch or a foot this way or that, or if not for the heroic action of a person, someone could have been injured seriously? What happened? What did we do about it?
Does your supervisor or operations manager have difficulty deciding to shut down operations when s/he receives a report about a potential for serious injury?
Senior leadership should also ask what the local management team is doing to enhance safety leadership and how it knows that its leaders are improving in this area. Additionally, leadership should ask about instruments that provide a true measure of the culture and what leadership is doing to improve the results of the measures.
When senior leaders are willing to ask these questions and are ready to listen to the responses, they demonstrate an emotional commitment to improving culture. When they are further prepared to understand what people are saying, to evaluate the influence of observed behaviors on culture, and to influence a change in that culture, leaders are leading the way to a culture that places a high value on safety.
Such practices are consistent with a transformational, versus a transactional, leadership style. Transformational leadership, also known as relationship-oriented or inspirational leadership, is characterized by behaviors that engage and motivate followers to act beyond mere self-interest. This style of leadership has been shown to predict enhanced safety performance (Barling, et al., 2002). In addition, direct reports’ ratings of a leader’s influencing style predict that leader’s best practices, which aggregate across leaders to predict the characteristics of organizational culture and safety climate (Bell, et al., 2008).
QUESTION 7: When we look at safety-related events, are we influenced by attribution bias?
Cognitive biases are mental shortcuts used to make judgments about uncertainties (Kahneman, Slovic & Tversky, 1982;Hammond, Keeney & Raiffa, 1998). These biases can skew perceptions about workplace exposures and adversely affect safety related decisions (Krause, 2005).
One such bias is attribution bias, which describes a tendency to misinterpret cause and effect. When something bad happens to someone who is not a peer, people may tend toward an internal bias. That is, they see the bad outcomes as something caused by the person, that s/he had a bad attitude or intentionally did something wrong to produce that outcome.
When an individual personally experiences something bad, s/he tends to have an external bias. That is, the person will likely point to and focus on systems issues or factors outside of his/her control as the reason for the event. In the authors’ experience, managers tend to have an internal bias when they hear about an injury or serious event, while workers are inclined to have an external bias.
Given this natural bias, leaders must support a thorough investigation to understand fully both immediate and root-cause factors. Attribution bias can creep in when the immediate causes of an incident are first reported. Commonly, the immediate causes of an incident are a misalignment in the working interface resulting from improper or inadequate equipment, inadequate processes or employee action or inaction. When leaders hear that an employee performed in an at-risk manner that contributed to the event, they may conclude that the employee is at fault. This is attribution bias.
During the investigation process, senior leaders must ask, What is the likelihood that this is the first and only time these immediate causes have existed? How probable is it that no other employee has violated this rule or not followed that procedure? In most cases, the answer is near zero.
Digging deeper allows the investigation to move quickly past the immediate cause to the root causes and, finally, to an understanding of why and how nonconformance became acceptable in the organization. Leaders must ask the right questions to show that they seek full disclosure of the chain of contributing causes and confidence that a similar event will not recur. These questions include:
What were the immediate and root causes to this event?
Do the recommendations address both immediate and root causes?
Who is responsible for tracking the recommendations and reporting to me on progress?
How will we know that the action plan we developed will result in the changes we need and be sustainable?
Are all parties in agreement with the investigation and recommendations?
These questions are another way for management to ask, What can we learn about our exposures, and how can we better manage them? In many cases, the actions organizations and their leaders take to manage safety arise more from attributions than from actual causes (DeJoy, 1994).
Not surprisingly, in organizations where leaders ask questions of this nature, and show a true desire to address root causes (including weak systems, leadership practices and culture), employees have a better understanding of safe operating procedures and are less likely to be injured or involved in a near-miss incident (Michael, Guo, Wiedenbeck, et al., 2006). Countering attribution bias promises a more powerful, positive and lasting impact on the safety climate in an organization than the placement of blame for action or inaction.
QUESTION 8: Are we maintaining a sense of vulnerability?
One of the most dangerous developments in an organization is the leaders’ loss of their sense of vulnerability to catastrophic events (CAIB, 2001; Baker, 2007). In some ways, leaders are more at risk of losing this sense than other employees. The sheer scope of their job means they are continually monitoring and managing a wide range of threats to the organization. It is easy to lose a sense of urgency for safety when the severity or frequency of accidents is low. However, the absence of injuries does not indicate an absence of exposure (Hale, 2001;Manuele, 2004). Nor does it mean that exposure levels are trending downward. In fact, the opposite may well be the case.
Leaders maintain their sense of vulnerability by continually gathering and receiving information about the true state of workplace hazards, safety prevention mechanisms and practices, and organizational culture. Leaders maintain confidence by listening to the discussions about injury rates and trends. They listen to determine whether the discussion focuses more on injury classification than on the event and the prevention plan. They assess whether any systems discourage full disclosure.
Consider a situation in which a new senior leader is hired or is new to his/her executive position. How much information does staff share regarding serious events? For this leader to understand the organization’s history and why certain systems and climates exist, then s/he must know defining moments in the organization’s history.
This leader must hear about these events and why they are important, for example:
names of deceased employees and the dates of fatal events;
copies of the investigation reports;
history of corrective and preventive actions for prior serious events;
information about how the organization has ensured that all of its other sites have addressed similar exposures.
When an organization is open to sharing this information with a new leader, it becomes clear that it expects leaders to place a high personal and organizational value on safety.
Fundamentally, a sense of vulnerability helps leaders make better decisions with respect to safety. Leaders must often make operational or strategic decisions where the outcome is unclear. In these situations, leaders are particularly susceptible to cognitive biases such as overconfidence, recency effect and status quo bias, especially if they are acting within the context of few recent incidents or from an incomplete set of indicators.
Lacking a sense of vulnerability, leaders are at risk of making wrong decisions (Hammond, et al., 1998). On the other hand, leaders who maintain a defensive posture with respect to serious events can help others evaluate threats from a new perspective (Johnston, 2004), and make better safety-related decisions (Krause, 2005).
Making Serious Incident Prevention a Reality
For many SH&E professionals, the first challenge to addressing life-altering injuries and fatalities is knowing how to bring the topic into discussion with senior leadership. Interestingly, some people become superstitious, believing that talking about the topic will somehow bring on an event. For others, the task of getting the organization to incorporate this thinking and these systems into its culture is seen as so daunting that it is perceived as unachievable.
Yet, SH&E professionals have an ethical duty to try to facilitate conversations about these concepts. While they cannot force leaders to do the right thing, they can at least ensure that they have considered the possibilities.
In a high-functioning organization, an article such as this one is shared with members of the leadership team who will be asked to read it and be prepared to discuss it at an upcoming meeting. The topic will be placed on the agenda, the merits of the questions will be discussed and considered, and the team will consider how the organization is approaching these questions. It is the authors’ belief that in this type of organization, the right choices and decisions will be made.
For others, the task will be more challenging. The culture may be less open to ideas or the SH&E professional may not be in a position to bring such a sensitive topic to leadership. In this situation, a healthy discussion among the leadership group is unlikely. More importantly, this professional is worried about the issue being dropped without adequate consideration. If s/he believes strongly that the organization would benefit by taking steps to reduce the probability of life-altering injuries and fatalities, then s/he must find an ally in the leadership group. The SH&E professional would then work with the ally to advance the discussion.
Catastrophic events need not happen. Leaders who review and understand the right metrics, ask the right questions, focus on creating a culture of engagement and create the right tension around vulnerability are doing the things necessary to align their organizations for injury-free performance. By following the guidelines presented, senior leaders can look closely in the mirror and feel confident that they are doing what is necessary to provide proper oversight and sponsorship for prevention.
By Thomas R. Krause, Donald R. Groover and Donald K. Martinhere.
