Keywords

2.1 Introduction

Organizations that have to responsibly handle high operational risks have long aimed to do so by quantitatively estimating both the risks they are faced with and the effects of measures to mitigate these risks. While this remains the dominant approach in risk and safety management, there is growing concern that there are a number of challenges that cannot be adequately met by this approach [2]. Uncertainties need to be acknowledged that by their very nature cannot be quantified and a mere focus on reducing risks and uncertainties in risk mitigation neglects the necessity to improvise, learn, and innovate in view of these uncertainties.

In this chapter, I address these challenges by proposing a new approach to uncertainty regulation in organizations. I then apply this approach to decisions on rules and operating standards as one of the most fundamental tasks of risk and safety management. Lastly, I discuss possible steps toward adopting this new approach to uncertainty.

2.2 Definitions of Uncertainty

Uncertainty is ubiquitous. Accordingly, debates on how to define and manage uncertainty abound. For the purpose of this chapter and the proposed new approach to uncertainty, I refer to two distinctions to position uncertainty:

  • Aleatory uncertainty is related to the randomness of events and cannot be directly influenced, but captured in probabilities at best. Epistemic uncertainty refers to lack of knowledge and can be influenced by seeking knowledge in a given situation or entering new situations with new unknowns. Uncertainty regulation only includes epistemic uncertainty, that is in the most basic terms uncertainty stemming from “not knowing for sure” [9, 17].

  • Exogenous uncertainty is the uncertainty present in an actor’s environment. Endogenous uncertainty is uncertainty as perceived by the actor, which is affected by exogenous uncertainty, but also by the actor’s capabilities, attitudes, and motives related to recognizing and appraising uncertainty. Actors can only influence endogenous uncertainty, which may indirectly impact exogenous uncertainty [7].

2.3 Uncertainty Regulation in Organizations

The proposed approach builds on uncertainty regulation theory as suggested by Griffin and Grote [7]. Regulation in this theory refers to processes internal to an actor, which help the actor to manage internal and external demands during goal striving. The theory was developed from the perspective of individuals who accomplish tasks in a work setting. It posits that work performance relies on two intertwined self-regulatory processes: (1) Endogenous uncertainty related to the task to be performed is addressed by behaviors that enable different amounts of flexibility in response to that uncertainty, called opening and closing behaviors, opening behaviors such as suggesting changes to existing procedures or raising concerns about a current course of action momentarily even increase uncertainty, whereas closing behaviors such as seeking confirmatory feedback or following a routine process reduce uncertainty. (2) Independent of what the specific task requirements are, endogenous uncertainty is increased or decreased in line with an individual’s uncertainty preferences, e.g., a curious person may use many opening behaviors that increase uncertainty, such as trying out new tasks or changing jobs. If individual uncertainty preferences and the uncertainty to be handled in a task align, then the two processes interact in an effective way leading to successful performance. However, the interaction between the two processes can also entail that an individual who is averse to uncertainty, responds with routine behavior to a task that would require innovation, leading to suboptimal performance. The reverse is also conceivable: an individual who loves to learn and innovate tries out new behaviors in a task that would require sticking to existing rules and routines, again resulting in suboptimal performance.

This theory can be transferred to the functioning of organizations in two ways. First, one may assume that the same kinds of individual processes apply to key decision-makers in organizations, such as members of the top management team, which brings the suggested processes to the organizational level because these individuals’ behavior impacts the whole organization. For instance, a new CEO may even have been hired because of their known willingness to increase uncertainty for themselves and the organization, leading to many opening behaviors that—depending on what the actual organizational challenges are—may affect organizational performance positively or negatively. Second, one can turn to processes of organizational governance and seek to identify preferences for opening and closing behavior at the institutional level and their effects on organizational performance. A young company may mostly rely on opening behaviors, such as trying out new products and markets in search for the best opportunities to grow, whereas more established firms often have built a stock of organizational routines for core processes and mostly rely on those. The effects on performance again depend on the particular circumstances, where possibly young organizations embrace uncertainty to a point where they cannot handle the ensuing operational risks anymore, or established organizations realize too late that their routines do not match emerging environmental changes. Lastly, one has to consider that individual and institutional processes depend on each other. It is through the action of key decision-makers that new institutional processes are introduced and old ones abandoned.

2.4 Flexible Rules

2.4.1 Basic Problems with Rules: Uncertainty and Autonomy

Rules, defined as any prescriptions that guide behavior and coordination among actors in a social system, are a cornerstone of risk and safety management [13]. Rules are to reduce variation in behavior and with that ensure the correct—that is also, safe—way of acting in any given situation. Rules thus are considered the silver bullet for ensuring safety by many. However, anyone involved in writing and implementing rules has a different story to tell. It is impossible to come up with rules that control all conceivable behavior in all conceivable situations, not to speak of the non-conceivable behaviors and situations [5]. Even if rules can capture the necessary behavior, rule compliance is notoriously difficult to monitor and enforce.

Two issues lie at the heart of these difficulties: uncertainty and autonomy. Rules are derived from knowledge about the correct course of action in a given situation. They therefore require that uncertainty about what the correct course of action is has been sufficiently reduced to be confident in the chosen action being the right one. Obviously, this cannot always be achieved, and the resulting rules may lead actors to do the wrong thing, unless they decide to violate the rule as they realize that it does not fit the circumstances they find themselves in. Many accidents have been caused because people followed the prescribed, but inadequate course of action, just as other accidents resulted from people not following rules, potentially with all good intentions because they felt that the prescribed behaviors did not match the situational demands. The opposite is obviously also true: People follow rules in useful ways and also sometimes deviate from rules in useful ways. However, because recovery from difficult situations is usually not systematically tracked, organizations often lack knowledge on successful rule violations. Therefore, there is generally too little awareness of the uncertainty actors are faced with when trying to follow rules [18].

The second issue, autonomy, has two important facets. One facet is directly linked to the previous discussion about uncertainty. To successfully face uncertainty, actors need to have freedom of action, i.e., autonomy, to adapt their behavior as new knowledge emerges and to proactively choose new courses of action to explore and improvise in the face of highly unexpected circumstances. If actors violate rules because the rules turn out to not be helpful, they exercise this autonomy. The second facet of autonomy is the expression of agency through autonomous action which is a major motivating force for human behavior. Individuals are motivated to engage in activities of their own accord (= autonomous or intrinsic motivation) when these activities are inherently interesting and/or meaningful [6]. People in general resent attempts to curtail their autonomy and are found to be less motivated and satisfied at work or in other settings when they feel externally controlled. Rules are a very direct form of external control, therefore, it is hard to create autonomous motivation for rule compliance. Punishment for non-compliance has been found to be an insufficient or even counterproductive driver of behavior, frequently leading people—and even whole organizations—to cover up inadequate behavior. Rewards for compliance may work better, but in this case the motivation for the correct behavior nevertheless stems from external factors rather than from an internal desire.

Following from these considerations, there are two questions to be answered: How can rules successfully guide behavior even if substantial amounts of uncertainty about the right course of action in any given situation remain? How can rules instigate autonomous motivation for rule compliance?

2.4.2 Rules as Part of Uncertainty Regulation

With respect to the first question, understanding how rules operate through the lens of uncertainty regulation helps to redefine and broaden the function of rules in guiding behavior even in situations with substantial uncertainty. For this to happen, organizations should adopt a process for systematically managing the introduction and adaptation of rules along the following five steps [12]:

  1. i.

    Any situation for which a rule is to be devised should be analyzed in terms of the uncertainty the situation entails.

  2. ii.

    Depending on the amount of uncertainty, different rule types should be employed that include different degrees of flexibility in line with the uncertainty that needs to be handled by the actors [14]. Such “flexible rules” may also momentarily increase uncertainty, for instance when they require the actor to consider more choices for possible behaviors in a given situation.

    • If there is little or no uncertainty as to what the correct course of action in a situation is, then action rules should be used, that is rules that specify in much detail what that correct course of action is.

    • If the uncertainty is fairly limited, then the right choice are action rules with systematic inclusion of degrees of freedoms to allow for adaptations in behavior in view of that uncertainty, by using terms such as “if needed”, “in light of specific conditions”, or “depending on the availability of further information”.

    • If there is uncertainty about which of a number of possible behaviors is the correct one in a given situation, and that decision has to be made by the actor based on emerging information in the situation, process rules should be used, that is rules that guide the actor in collecting information, coordinating with others, etc. in the process of deciding on the right course of action.

    • If there is much uncertainty and there are no means to reduce that uncertainty for the actor beforehand, the only way to guide action is to provide goal rules, that is rules that help to set priorities and keep the overall direction of action within certain bounds.

  3. iii.

    The use of different rule types should also be considered in light of the assumed uncertainty preferences of the actors for whom the rule is devised. Obviously, differences between single individuals cannot be accounted for when rules are written, but differences between professional groups, hierarchical levels, or other broader categories should be examined. For instance, differences in knowledge, ability, and prior experience as well as in professional codes of conduct or cultural norms are likely to shape the willingness to cope with existing uncertainty and the desire to further reduce uncertainty or on the contrary increase uncertainty as a means for learning and competence development.

  4. iv.

    The experience with existing rules needs to be monitored, and feedback by all actors encouraged as part of a continuous learning process for which knowledge of “successful” rule violations is particularly important. In this process, it is also important to consider whether there are outdated rules, which need to be abolished, and whether all actors are always fully up to date on the currently valid rules.

  5. v.

    The processes involved in rule making should be embedded in an overarching concept of how an organization approaches uncertainty. Following the Griffin and Grote [7] theory and earlier work by Grote [10], organizations should consider costs and benefits of reducing and increasing uncertainty in light of performance requirements in different situations. We return to this point in the final section when we discuss which steps organizations might take to adopt a new approach to uncertainty.

2.4.3 Rules as Constraints and Enablers for Behavior

To answer the second question of how autonomy can be preserved when for safety reasons there must be strict bounds on permissible behavior, a closer look at who decides on rules is helpful. Mostly, rules are made for others as a means of controlling their behavior, which then creates the motivational problems already discussed. If, however, people make rules for themselves, they are given a different kind of autonomy, that is the autonomy to make decisions on restricting one’s own autonomy. Such “higher-order autonomy” [8] can keep autonomous motivation alive as restrictions of one’s freedom to act are taken voluntarily and with full understanding for why they are necessary. Participation in rule making, assessment of existing rules, and continuous feedback on and adaptation of rules thus is key to promoting autonomous motivation for rule compliance [3, 16].

Moreover, if the restrictions are chosen in line with the uncertainty considerations discussed in the previous section, then it becomes clear that rules do not only constrain but may also enable behavior [1]. When actors have to handle difficult situations with high amounts of uncertainty, goal and process rules provide some guidance, but also leave sufficient flexibility to adapt behaviors as needed. Thus, actors receive help in managing the uncertainty, but are also empowered to make their own decisions. This autonomy needs to be coupled with adequate education and training to ensure that actors have the required knowledge and skills to make adequate use of the offered degrees of freedom. Additionally, actors’ uncertainty preferences have to be taken into account. Flexible rules demand a willingness by actors to accept or even increase uncertainty, which can be fostered by competence building and by creating a culture in the organization that encourages empowerment and learning [12].

2.5 Steps Toward Adopting a New Approach to Uncertainty

For organizations to adopt flexible rules as a way to more effectively regulate uncertainty, predominant perspectives on risk and safety management need to be reflected upon. At their core, these perspectives concern deep-rooted assumptions about the relationship between safety and autonomy [11] and the necessity to create stability by reducing uncertainty [10]. Despite the increasing emphasis on organizational resilience as the answer to managing the unexpected [15, 21], many organizations still aim to minimize uncertainty in the hope to maintain stability and control as prerequisites for safety. Autonomy that would allow flexible action and resilient adaptations to external changes is considered unnecessary or even dangerous as it threatens routine operations. Accordingly, a first step is for organizations to reevaluate the balance between stability and flexibility that is required for different operations. Flexibility demands mostly stem from having to cope with uncertainty, while stability demands result from internal and external requirements for predictability and control. These demands may vary between different parts of the organization and also between different points in time and have to be monitored and reassessed frequently [4].

Once it is clear that uncertainty regulation is about more than just trying to minimize uncertainty, the relationship between safety and autonomy needs to be scrutinized as a second step. If safety requires the full range of opening and closing behaviors because situations frequently change and behavioral routines have to be complemented by learning and exploration, actors have to be empowered to autonomously choose the right behaviors. The different kinds of flexible rules discussed earlier can help them to make good choices.

As a third step, organizations should reflect on basic assumptions and values shared by their members which promote or hamper a fuller understanding of this new approach to uncertainty regulation [19, 20]. Thereby, organizational culture is highlighted as a driving force behind organizations’ practices in risk and safety management. Only if there is a shared commitment to broadening the perspective on uncertainty and empowering actors to respond to and possibly even increase uncertainty, can measures such as flexible rules be successfully implemented.