Keywords

1.1 Introduction

This chapter suggests several dimensions which are meant to help frame a complex topic representing a very rich diversity of situations across industries, countries, and epochs. The idea is to sensitize readers to several aspects associated with the topic of rules and autonomy in the domain of safety, and of this book. It implies several simplifications. Such choices strongly limit nuances, but the aim is to emphasize the importance of contexts when it comes to (safety) rules. Contexts refer to organizations, to industries, to risks, to histories, to practices, to situations, and to countries. Nuances would require far more space but would also divert from the general idea of this text which is to provide elements of analysis and perspective to situate several issues associated with the topic of this book.

The result is a relatively short but dense text which relies on a mix of conceptual and empirical research in a diversity of traditions in the field of safety and beyond. In this introductory chapter, three sections develop the importance of context: the advent of safety rules as an established narrative (1), there is more than rules in safety (2), and historical trends … a bureaucratization of safety (3). The last section presents the chapters of this book, in three categories, finding or losing the balance; the role, position, and influence of middle-managers and top management, finally; when autonomy, initiative, and resilience take the lead.

1.2 Rules in Context (1): The Advent of (Safety) Rules as an Established Narrative

The principle that rules exist to protect workers from unsafe conditions in a capitalist era has a long history. The horrific working conditions depicted in the industrial revolution of the nineteenth century remind us of how much progress has been made over the past two centuries in many countries (starting with child labor). The notion of safety rules is thus embedded in the idea, and a long history, of making sure to protect workers from harms and hazards, limiting the risks, for instance, of falls, anoxia, electric shocks or crushed limbs when respectively working from height, in confined space, close to live electric cables or close to moving mechanical parts. Ideally, system design should prevent people from exposure to hazards, yet rules understood as practices (not “behaviors”) are also necessary to make sure that workers limit their exposition to hazards when safer designs are not possible.

The law has progressively evolved to make sure that such rules are in place and enforced to prevent harming workers. The idea that the expectations should be set by states is also particularly ingrained in the struggle to improve workers’ safety (the same applies to product safety, road safety, or fire safety). So, the idea that the absence of safety rules or the absence of their implementation (when existing) is a problem because it exposes workers to hazards, is a strong legacy of the past, and of its successes. History shows us that countries which tighten, through laws and state enforcement of rules their expectations in terms of safety, obtain results (a broad comparison between countries across the world makes it clear).

One reason is that the forces of capitalism are not compatible with safety if not counteracted by strong requirements imposed, from within or from outside, to organizations of all sorts. These requirements are translated in rules. To work safely, hazards need to be identified (which is itself a rule), then safety rules designed to prevent and to mitigate these hazards, then complied with in practice. This is potentially constraining for businesses because such rules can slow down work, can make products or services more expensive. Rules in terms of design can cost extra money. Therefore, and seen from this perspective, an accident in the workplace is very often associated with the lack of sufficient attention granted by employers to the safety of their employees, and the design, implementation, and enforcement of adequate rules. It is a discourse which makes sense considering many industries and countries’ experiences in history.

In other domains than workers’ safety, with the advent of new modes of transportation during the industrial revolution for instance, starting with trains and steamships, safety has also been linked to the importance of rules. The first severe train crash in France in Meudon in 1842, caused a fire which killed fifty-five passengers. The doors were locked by the controller after departure, and many could not escape their carriages, increasing the death toll. The rule was then changed. The doors were no longer locked after this event. The Titanic sank in 1912 and didn’t have sufficient lifeboats to evacuate its passengers. It had 20 lifeboats, half of what was needed to save everybody on board. The design rule changed afterward to make sure enough lifeboats were available for all passengers of steamships.

Because of the safety risks to workers and passengers, rules became central to the achievement of safe operations in such transport systems, not to mention design improvements much later in road safety, from safety belts to airbags and other features implemented in cars’ designs to secure driving, including of course design of infrastructures (roads, tunnels, bridges) and rules to regulate traffic and drivers’ practices. There would be a list of many other cases in which rules (of design or of practices) evolved to improve the safety of systems as they matured (mining, explosives’ production, or transport of hazardous material for instance). But the advent of more complex and larger technical systems in the second half of the twentieth century is another good example to show the importance of rules (of design and of practice) to reach high levels of safety.

Nuclear power plants, ballistic missiles, space exploration, nuclear weapons or submarines, civil aviation, chemical plants, pipelines, and refineries are many other examples in which safety had to be translated in rules (of design, of practices) to maintain operations and activities under control. It is difficult to imagine obtaining such levels of reliability and safety without the existence of rules, standards, procedures and of compliance with them. And regulatory systems also played a strong role in many of these cases. For some of these systems, they are protected from the forces of markets and capitalism (mostly in the military domains), but many are not. This view of industries in general (e.g., construction, factories, warehouses) and safety-critical systems (e.g., nuclear, oil and gas, aviation, chemical, railways) therefore constitutes a well-established discourse. From this macro perspective of the history of safety, rules matter.

It is a narrative which makes a lot of sense considering what we know and what to recommend for improvements. However, this discourse also needs a certain degree of refinement. First, there is always room for improvement. For instance, in France, a country in which safety in the workplace has been incorporated in the law for decades, two people died at work every week in 2021 (more than 700 deaths) in industries and services (this figure excludes workers killed when traveling to work or employees committing suicides). This situation is obviously unsatisfactory and reflects profound issues. And, in highly regulated safety-critical systems, accidents and disasters still occur from time to time, also indicating the limits of safety. In France, if one considers as an indication the number of casualties, the last major event in the chemical industry was in 2001, in Toulouse; the last major aircraft crash happened in 2009 between Rio and Paris, and the last major derailments were in 2013 in Brétigny and 2015 in Eckwersheim.

Second, a point to be developed next, rules need to be strongly contextualized to understand how safety is really produced, and this tends to bring another discourse to the one sketched above, not replacing it, but refining it. Safety can only be achieved through workers’ expertise for instance, but also when the proper working conditions are created for rules to be implemented, which is, in both cases, an organizational outcome. Rules also need to be further contextualized from a historical point of view, to incorporate trends in work, labor relations, organizations, businesses, markets, and states over several decades. To contextualize rules, the contribution of many different disciplines or research traditions is needed. Law, sociology, psychology, social psychology, cognitive engineering and ergonomics, political science, and management have much to bring to such contextualizing. Indeed, our understanding of rules in safety has been tremendously improved over the past decades thanks to these traditions, and this book is an addition to this.

1.3 Rules in Context (2): There Is More Than Rules in Safety

The key problem with rules is simply that work, organizations, and regulations are far more complex than what rules can actually reveal about daily activities. There is more to work, organizations, and regulations than rules in general. This is the result of decades of research. Studying work shows how workers must adapt to circumstances in ways that requires their expertise and judgment. Realities of work often create imperfections not accounted for, including imperfection of rules themselves. Dilemmas and surprises also challenge the application of (safety) rules. A certain degree of autonomy is therefore expected. French ergonomists have long and ethnographically described the difference between what is defined as “prescribed work” and “real work”, translated in English with the notions of “work as done” and “work as imagined”.

Studying organizations shows that bureaucracies, epitomizing the ideal of the legal-rational principles of conduct through formal structures (e.g., standardization, division of work, lines of authority), also reveal informal realities beyond rules. The perfect, clockwork, ideal of organizations based on rules must be complemented by what psychologists and sociologists have shown for decades to be a far more complex mix of cognitive, informal, relational, social, and political dimensions beyond the structure of formal organizations. In this respect, safety risks are not given but collectively constructed instead. Rules (of design, of practices) are constructed. And regulation, one important role of the state consisting in producing laws and their enforcement through administrative bodies (e.g., inspectorate), is no different.

These two paragraphs shortly frame broadly some of the key analytical elements of several decades of findings in different research traditions in the field of safety, such as cognitive system engineering and high-reliability organizing for instance. The outcomes of many empirical studies in these traditions show the importance of understanding safety beyond a simplistic approach to rules. Ethnographic studies of cognition and organizations in safety-critical contexts show the ability to cope with complexity (i.e., imperfections of work and organizations, unexpected events), and therefore the combination of rules and of a certain degree of autonomy.

These different research traditions, beyond their intellectual background and the way they frame it, using different vocabularies, address the ability to find ways to combine anticipation and resilience, centralization and decentralization, compliance and initiative, or standards and expertise. The choice of vocabulary can convey differences in conceptualization, but the problem is shared by researchers in safety irrespectively of their disciplinary backgrounds (e.g., sociology, psychology, ergonomics, management, political science). However, as already discussed in the introduction when contrasting highly regulated organizations (e.g., aviation, nuclear) with other industries (e.g., factories, warehouses), but also countries, there are many differences in the level of resources and strategies of companies, industries, and states involved in managing this combination.

Centralizing in certain safety-critical systems (e.g., aircraft, nuclear submarines, space exploration, nuclear power plants) implies great expenses in training, engineering, and design to make sure that rules and standards are created with adequate support for them to be properly implemented, and to be relevant to the level of risks of different tasks and activities. Decentralizing means that a certain degree of autonomy is made possible thanks to the adequate level of expertise of the workforce (including management) obtained through training, socializing, and professionalization.

In other words, the quality of the balance between rules and autonomy or centralization and decentralization to manage safety risks is thoroughly an organizational, or system one. One cannot understand what happens locally in daily frontline practices, in terms of rules and autonomy, independently of the organizational contexts and resources involved. Thus, the ability to reflect and to find solutions to problematic situations met in the field in terms of the balance between rules and autonomy is an organizational matter, not a frontline one. Safety-critical systems must manage this daily.

To do so can imply the involvement (preferably before a major event) of an awful lot of roles and functions depending on the extent of the problem, from workers, supervisors, middle-managers, engineers, and perhaps sometimes also, top managers and regulators. What happens at the front line, in the field, depends indeed on a wide range of vertical and horizontal coordination, cooperation, negotiation, and communication in all kinds of situations in terms of resources, delay, and objectives. The ability to calibrate organizations to meet their safety expectations translated in rules is a challenge which varies greatly according to domains.

For instance, organizations outside of the category of safety-critical systems, with different levels of resources, expectations, and requirements, do not have the same options. This, of course, does not imply that major events in safety-critical systems do not happen. It does not also imply that improvements are not possible in these highly regulated, reliable, and safe systems, and that they do not face complex problems daily, and risks of degradation in their accomplishments (many cases illustrate this, the Boeing 737 Max comes to mind). It does not imply either that other industries cannot be successful in safety, and many are. However, the point made is that the construction industry, for instance, because of its very distinct (business) model is very different from aviation in terms of the combination of rules and autonomy, or centralization and decentralization.

It is therefore important for the topic of rules and autonomy in safety to be strongly contextualized when discussed. This issue cannot be approached in the same way in the construction industry or in the aviation industry. Different systems, different approaches. So, this contextualization of (safety) rules matters, but that is not all. Another type of contextualization is needed. It refers to the historical trends in work, labor relations, organizations, businesses, markets, and states over the past decades which have been now widely commented in the field of safety across research traditions, industries, and countries.

1.4 Rules in Context (3): Historical Trends … a Bureaucratization of Safety?

While the importance of rules (of design, of practices) in safety are variously established across domains, requirements and expectations are not fixed and have also considerably evolved over time too. Three important sources of evolution can be identified and briefly discussed. One is the increase of standards. As the economy evolved toward extended global markets in the 1980s and 1990s, management standards to ensure quality of services and products across contracting companies, certified through audits by external parties, proliferated. It created new demands on organizations to show to auditors how compliant they were with management standards, requiring an important level of paperwork (or increasingly digitalized version). Safety followed that trend, to also become a domain framed by management standards, certified by auditors, expecting traceability for auditing then certifying, or not, companies.

This transformation implied the recruitment and training of people dedicated to such activities of certification inside companies, including safety professionals. It also created a new business for certification and consulting companies. This increase in standardization was concomitant with evolutions in states’ strategies in terms of regulatory approach. A move from the prescriptive, command-and-control philosophy was replaced by tools with various degrees of self-regulation. The genesis of these changes in regulatory regimes led to an increased reliance on the production of internal rules, adapted to their safety risks, to be inspected by authorities. The change of regulatory philosophy was not at all the end of rules, but a new way of prescribing rules internally for companies, and the need to show how such rules were in place for inspectorate to check on them externally.

Another source of important changes connected to the production of rules in safety in the past decades is the outsourcing of many activities in companies which implies the production of contracts, and a codified, legal mode of interacting between organizations, including the topic of safety. For multinationals operating across the world, these evolutions (management systems, certification, and audits; self-regulation, internal standards, and inspections; externalization and contractual relationships) have meant an important investment in the production of safety management systems and of a range of categories of rules. This was coined, when it first appeared, the development of an audit society, then of a bureaucratization of safety.

While multinationals are indeed concerned across domains, other organizations, smaller ones working as subcontractors for these bigger ones for instance, must cope with this increase too. This description is not to say that we discover the issue of red tape (the excessive, slow bureaucratic process of submitting decisions to multiple authorizations and formalities) but that the bureaucratization of safety addresses the problem of a potentially overly bloated, disproportionate level of requirements in relation to the safety risks involved. Another dimension of this expansion of rules is the lack of consideration, in the context of inflated bureaucratic processes, of the complexity that it entails for workers and managers. Moreover, it has become subsequently easier (although of course far from new), after an event, to find a rule which has been breached. Workers’ autonomy, expertise, and professionalism are employed to find ways to cope with a plethora of rules that have become a risk for them to manage.

1.5 A Complex Empirical Question, a Sensitive One

Diversity and variation in epochs, countries, industries, organizations, practice, situations, and systems…the aim of this very sketchy picture, intently so, is to signify the need for a very careful, nuanced, and empirical approach to the question of rules and autonomy in safety and the importance of contexts. Dedicated studies on safety and rules over the years show how researchers have considered contexts and this idea, for instance, that there is more to rules, from cognition to organization [3, 5–9] but then paying more attention to historical contexts in terms of bureaucratic and regulatory trends of the late twentieth and early twenty-first centuries [1, 2, 4, 10, 11].

In this respect, the perspectives offered in this book by multiple traditions (sociology, psychology, management, criminology, cognitive engineering) are a blessing. Bringing multiple angles, methods, concepts, and vocabularies, these chapters avoid the excessive restrictions of a single view and contribute to sensitizing to contexts. Associated with the picture introduced in this chapter, they reveal the importance of thinking rules and autonomy, centralization and decentralization, or anticipation and resilience from an organizational or system angle. They can be categorized into three types of contributions.

  • Finding or losing the balance

  • Uncertainty Regulation in High-Risk Organizations: Harnessing the Benefits of Flexible Rules by Gudela Grote

  • Producing Compliance: The Work of Interpreting, Adapting, and Narrating by Ruthanne Huising

  • Untangling Safety Management: From Reasonable Regulation to Bullshit Tasks by Kristine Størkensen

  • The situations, roles, and influence of middle-managers and top management

  • Ambiguity, Uncoupling, and Autonomy: The Criminology of Organizational Middle-Management by Paul Almond

  • The Effect of Top Managers’ Organizational Reliability Orientation by Rangaraj Ramanujam

  • When autonomy, initiative, and resilience take the lead

  • Interlocking Surprises: Their Nature, Implications, and Potential Responses by Moshe Farjoun

  • Resolving the Command—Adapt Paradox: Guided Adaptability to Cope with Complexity by David D. Woods.

1.6 Finding or Losing the Balance

The issue of “finding or losing the balance regarding safety rules” is addressed in three of the chapters of this book. Gudela Grote advocates, from a psychological angle, in Uncertainty Regulation in High-Risk Organizations: Harnessing the Benefits of Flexible Rules, a practical way of finding a balance. It consists in articulating in a coherent manner dimensions such as uncertainty, autonomy, flexibility, and (safety) rules to the diversity of situations met in work contexts. Rules can be of different types, action rules (describing precisely what to do), process rules (providing guidance about what to do), and goal rules (helping set priorities about what to do), and these types should be adapted to the level of uncertainty of situations. An attention to the right level of expertise is therefore needed when flexibility is expected.

In Producing Compliance: The Work of Interpreting, Adapting, and Narrating, Ruthanne Huising follows a descriptive, ethnographic perspective, showing how compliance with the law and standards by producing and following rules is concretely made possible daily in practice, in organizations. It requires the work of dedicated people, compliance officers (technicians, officers, and managers) involved, pragmatically, in what is defined as an accountability infrastructures made of offices, technico-legal experts, programs, operating procedures, technologies, and tools. This pragmatic translation, to find a balanced strategy between (safety) rules and work, follows a pattern of interpretation, adaptation, and narration involving an array of actors, and their relationships, which produce contexts’ relevant outcomes.

When such a balance is lost, Kristine Størkensen in Untangling Safety Management: From Reasonable Regulation to Bullshit Tasks argues that the expression of bullshit jobs becomes a good characterization of what an excess of rules entails for many people affected by this excess. When workers, managers, or engineers start doubting about the relevance of their activities and the value of what their paperwork generates, it is a likely indication that rules might have become a hindrance without much purpose rather than a support to daily operations. Specifying the different levels of sociotechnical systems which contribute to a widespread case of bureaucratization across domains, Størkensen offers possibilities of improvement through strategies of relabeling, moving, and ultimately, removing such unwanted and problematic tasks.

1.7 The Situations, Roles, and Influence of Middle-Managers and Top Management

The topic of “situations, roles, and influence of middle-managers and top management” is covered by two chapters. In Ambiguity, Uncoupling, and Autonomy: The Criminology of Organizational Middle-Management, Paul Almond develops the insights from criminology on the topic of this book. Describing the role of middle-managers in their active contribution regarding compliance with rules, he situates them at the heart of complex organizations, in their difficult roles halfway between top management and frontline realities. Their responsibility and autonomy in this respect take place in the midst of potential ambiguities of meaning (regarding the extent of complying—or not—with rules in tough business contexts), structural uncoupling (existing differentiated units within organizations with a certain degree of freedom, identities, and cultures), and autonomy deficits (managing difficult situations—getting the job done—on behalf of organizations without their support in case of compliance breach, retrospectively).

In The Effect of Top Managers’ Organizational Reliability Orientation, Rangaraj Ramanujam proposes a management analysis, based on high-reliability research, contrasting what he describes as a modular orientation and a systemic orientation. A modular orientation represents a view of work and organizations as stable, unambiguous, formal, and proactively managed, while a systemic one represents an understanding made of instability, ambiguity, informal reality, and improvisation. With this conceptualization, he addresses the likely role and influence of top management team on practices in organizations when it comes to rules, reliability, safety, and events. Because top management teams affect organizational design, performance measurements, incentives, and accident investigations, such orientations, capturing two opposite mindsets, might indeed play a role in how they shape practices. Ramanujam argues that the need for a systemic orientation is ripe, considering multiple contemporary trends (including a post-Covid world) exposed in his chapter.

1.8 When Autonomy, Initiative, and Resilience Take the Lead

The idea that such current trends should be included in a discussion about (safety) rules prompts some researchers to emphasize the importance of resilience over anticipation. This situation “when autonomy, initiative, and resilience take the lead” is explored in two chapters. With Interlocking Surprises: Their Nature, Implications, and Potential Responses, Moshe Farjoun develops an argument about the likelihood of events which seriously defy our past habits of coping with the unexpected and require new practices of resilience. The reason for this evolving situation is the possibilities of interlocking surprises across space, time, and scale which, in his own words, are “evolving, recursive, multiplex, cumulative, and nested”. An increase of interconnectedness is certainly one source of such novelties, and the COVID-19 has played role in increasing this awareness. Farjoun advises organizations to follow diverse strategies when the time comes, from requisite variety to breaking affinities, through robust response.

To Dave Woods, the relation between anticipation—resilience is a paradox. The paradox is that there is a definite need for plans to specify action but people must adapt such plans at the sharp end because of the uncertainties, constraints, complexities of real-life situations. So here is the paradox of safety management. In Resolving the Command–Adapt Paradox: Guided Adaptability to Cope with Complexity, Woods argues that the two work together through the principle of guided adaptability. In this chapter, he relies on several decades of studies to delineate a theoretical yet practical answer to this paradox. Situating his reflection in complexity science, at the interfaces of natural, artificial, cognitive, and social sciences, the notions of brittleness, collapse, boundaries, and envelopes frame this issue as an adaptive one for which the principle of “plan and revise” constitutes the engine.