Russel Sage Foundation

This feature is part of an ongoing RSF blog series, Work in Progress, which highlights some of the research of our current class of Visiting Scholars.

Drawing from ten years of ethnographic research, Visiting Scholar Susan Silbey (MIT) is writing a book that examines the growing tensions between federal law and laboratory science. She is investigating the ways in which federal lab regulations and audits, often implemented in the name of safety, are perceived to threaten the autonomy of scientific practice within both the academy and other specialized industries.

In a new interview with the Foundation, Silbey discussed the factors that have given rise to breaches of regulatory compliance in academic and laboratory settings, including industry-specific hierarchies of labor, as well as larger cultural shifts in attitudes about workplace governance.

Q. In your research you have examined the 2009 UCLA laboratory tragedy that sparked the first criminal prosecution over an accident in an academic lab. What does this event, and others like it, reveal about the difficulties of ensuring regulatory compliance in academic lab settings?

Academic settings are difficult to regulate because they enjoy historic protections for academic freedom and because they are officially hierarchical but practically collegial. This means that although some actors (such as faculty) have ultimate authority, workday responsibilities are usually delegated but without formal processes for seeing that the work is done or how it is done. Academics resist being told what to do, insisting that academic freedom is not simply about keeping political interference at bay, but also about creating spaces for discovery and innovation. Therefore, seeking compliance with safety regulations demands a deft accommodation to local cultures while still ensuring safer and healthier practices.

Although the UCLA accident took place in an academic setting, most (if not all) large organizations contain niches of ungovernability. The privileges of faculty members that protect them from regulatory enforcement—as well the gap between policy and practice that effectively renders much university governance into suggestions rather than enforced job requirement—are not unique characteristics of academic institutions. Although my work developed from a study of university laboratories, many other organizations contain pockets of extraordinary privilege, for example, high-status actors such as executives, high-skilled experts such as surgeons or financial analysts, and high-demand employees valued for their connections with clients, funders, or external decision-makers.

Whether it is the elite status, the insulating cultural shrouds surrounding expert labor, or the organization’s dependence on many workers’ discretionary decisions and interpretations of more explicit instructions, many organizational actors—authorized by their membership and location—may perform their work relatively free from effective surveillance and predictable control. Thus, the story of Sheri Sanji’s death in the UCLA laboratory—and the longer story I will tell of one university’s effort to transform its laboratories into more reliably safer, environmentally sustainable work places—are not just stories about academic science. The implications are much larger and more pervasive. This is simply one, rich example of our systemic inability to control the consequences of long chains of distributed yet coordinated labor.

Q. Laboratories, academic and otherwise, require specific divisions of labor. How do these particular configurations lead to violations of protocols and guidelines that have been established to ensure workers' safety?

Organizational actors vary in their range and depth of expertise, and in terms of the autonomy with which they work and how much supervision they receive. These variations lead to differing interpretations of regulations. Actors with great expertise but less supervision or infrequent interaction with regulators often interpret safety rules as obstacles to efficient and creative production. Often such relatively autonomous and expert actors believe that they know more than the regulators. This is often true of many professionals who have been highly trained to assess specific conditions and determine what rules or principles to apply or ignore.

But other actors work with less expertise and less autonomy and also little direct interaction with regulators. Often such subordinate actors fear the consequences of regulatory enforcement, worried that their own job performance will be negatively assessed or that the organization for which they work will be negatively impacted, either financially or reputationally. These actors often interpret regulations as threats. Lacking expertise to find ways to work safely within the rules, they may try to avoid rather than adapt. Or, to sustain production within the rules, they may become overly rule conscious, fearing violations, and in these instances themselves become impediments to effective work.

There are some actors in organizations that occupy middle ranks, some expertise (middling and more), some autonomy (more distance evaluation and supervision), and often more frequent interaction with regulators and regulations. These actors more often than not can forge relationships with regulators and adapt regulations to local conditions—safe, compliant, and reasonable. We have found across a wide range of organizations that these middle level actors are the key to improved organizational performances (compliance and productivity).

Q. How have larger changes over the last few decades in workplace structures, and attitudes about workplace bureaucracy, helped give rise to some of the gaps in safety and accountability in lab and production settings?

Two trends over the last several decades seem to work in opposite directions, leading to significant gaps in safety and accountability. In the first trend, organizations are increasingly adopting forms of self-regulation in response to multiple, often competing, demands for accountability from consumers, investors, competitors, employees, and the government. Risks are proliferating as expectations for risk reduction also escalate. At the same time, the conventional model of a hierarchically governed organization reliably commanding compliance—which has always been more an aspiration than an empirical description—is also disappearing, even as an aspiration. Across almost all economic sectors, organizations are adopting decentralized, flexible, and lean structures as well as various forms of offshoring, outsourcing, and short-term employment relations. Through the diffusion of these two currents in private firms as well as public agencies, the state is, in effect, delegating responsibility to promote the general welfare to organizations that operationally lack the capacity to do this well. Considering the rapid diffusion of increasingly complex technologically generated risks, it is a recipe for disaster, which is exactly what happened in the UCLA laboratory, as well as other recent incidents such as the 2010 BP spill in the Gulf of Mexico, the 2014 anthrax contamination at the Centers for Disease Control, the 2014 lapses in Presidential security, and the September 2014 death from Ebola virus at a Dallas hospital. Some observers would also add the financial collapse of 2008 as just such a disaster borne of systemic risks bred in contemporary organizational practices.

Contemporary workers, including those formally employed as well as those located further along a distributed supply chain of contractors, subcontractors, and freelance laborers, are often engaged in work that is not fully prescribed beforehand, involving contextual adaptations and demanding localized discretionary decisions. Much production cannot be likened to the normal assembly line, and this is particularly true in scientific laboratories, both in academia and in industry. This situationally varied production is becoming ever more so in economies driven by expert knowledge and technological innovation, dependent upon the craftsmanship of skilled workers who create the scripts they enact as much or more so than follow pre-determined rules and narrowly delegated tasks. This situationally or locally dependent judgment characterizes the widest range of activities, no longer just the traditionally self-regulated professions such as law and medicine but including, for example, those working in obviously dangerous places such scientific laboratories, oil and gas prospecting, energy production, nursing, policing, as well as less immediately hazardous fields such as financial investing and accounting. Compliance with organizational and legal expectations must be achieved by securing willing conformity from often loosely coordinated actors up and down, inside and out, the network that constitutes the contemporary organization.