Between the devil and the deep blue sea: Objectivity and political responsibility in the litigation of the Exxon Valdez oil spill

Pitch and catch

The events in question began with a phone call during the summer of 1996. An engineer from Exxon was on the line for William Freudenburg, a prominent environmental sociologist whose work focused on perceptions of risk. The engineer wanted to discuss the fallout from the Exxon Valdez oil spill in Prince William Sound, Alaska, which occurred seven years earlier, in 1989. The captain of the ship was inebriated at the time. Despite Exxon’s knowledge of his drinking problem, the company allowed him to pilot the 990-foot long tanker, which was transporting 53 million gallons of crude oil from the Trans-Alaska pipeline.² Captain Hazelwood ‘inexplicabl[y]’ quit the bridge just prior to a difficult course manoeuvre required to avoid an underwater reef off Bligh Island, despite being the only officer on board licensed to navigate in those waters (Exxon Shipping Co. v. Baker, 2008: 477, 478). Third mate Joseph Cousins failed to make the necessary turn, running the tanker aground on the reef and tearing open the hull (Exxon Shipping Co. v. Baker, 2008: 478). Eleven million gallons of oil seeped into the fragile sub-Arctic environment, the largest single oil spill in the United States at the time, only exceeded by the BP oil spill in the Gulf of Mexico 21 years later.

Exxon was on the hook for several billion dollars in fines and clean-up costs. About this, there was no dispute. But in 1994, an Alaskan jury concluded that the company owed an additional $5 billion in punitive damages to fishermen and others whose livelihoods were affected by the Exxon Valdez spill, the largest punitive damages judgment in history. Such awards are intended to punish corporations that cause harm (see Benson and Kirsch, 2010). They are meant to prevent companies from internalizing ordinary fines and penalties as regular costs of doing business, helping to deter repetition of the offense. This is accomplished by allowing jurors to assign punitive damages at an unrestricted multiple of actual liabilities in what might be called ‘capital punishment’.

Large punitive damage awards make headlines when announced. Less attention is paid when they are reduced on appeal, which is a regular occurrence (Nader and Smith, 1996; although see Liptak, 2008 in the Exxon Valdez case). Exxon anticipated appealing the punitive damage award all the way to the US Supreme Court. As part of its effort, the company reached out to Freudenburg. Would he be interested in writing an article for a high-ranking social science journal on whether punitive damage awards alter corporate behaviour? Assuming that his research supports their interests, they would like to sponsor his work. The goal in commissioning the article, and others like it, was to promote doubt and uncertainty about the value of punitive damage awards, thereby reducing the corporation’s liabilities.

Freudenburg was intrigued and flattered by the invitation. Much of his work describes how the interactions of people, policies and complex technology combine to produce blind spots responsible for institutional failure, including catastrophic events. One of his major findings was that heuristic biases that prevent members of the public from making accurate assessments also affect the judgment of experts. He wrote or co-authored numerous articles and several books on oil spills, concluding in part that their environmental impacts persist far longer than predicted by contemporaneous assessments (Freudenburg and Gramling, 1994, 2011). Throughout his career, Freudenburg was equally at home interacting with members of affected communities, environmentalists, regulatory bodies, and industry personnel, and he often included anecdotes about his interactions with all four groups in his published work. His brand of scholarship relied heavily on pattern recognition. He was also well known for employing folksy cover terms to summarize his empirical findings (see Rudel, 2013: 48). For example, he explained how the debilitating social and psychological consequences of industrial accidents frequently result in ‘corrosive communities’ in which social relations become divisive (Freudenburg, 1997). Punning on the diffusion of responsibility identified by social psychologists, he coined the expression ‘diffraction of responsibility’ to describe the challenges associated with managing complex technology (Freudenburg, 1992: 15–19). Given his expertise in such matters, he was well-positioned to investigate whether punitive damage awards might reduce the likelihood of future environmental catastrophes.

Freudenburg took detailed notes during and immediately after his phone conversations with the engineer from Exxon (Freudenburg, 2005: 11, 17). He mused about what his friends on the right would think if they learned about the project—primarily jealousy that they had not been contacted—and the critical response from his colleagues on the left (Freudenburg, 2005: 18). But in his notes on the initial conversation with Exxon, he also acknowledged his ambivalence about the overall project and his participation:

Part of me is deeply bothered by the fact that this sort of thing is going on—at all—let alone by the fact that I might become part of it. Another part of me—the middle-of-the-road-part—is tapping me on the shoulder, reminding me that I’ve always said I try to be a straight-shooter, I call them as I see them, and whether it’s [the company] or [the company’s] sworn enemies, if they can use my stuff, fine, and if not, that’s their choice. (Freudenburg, 2005: 18; bracketed text inserted by Freudenburg)

Between the devil and the deep blue sea

How should we understand Freudenburg’s account of his experiences? ‘As much in confession as in accusation’ against the influence of corporations on academic research, as he describes (Freudenburg, 2005: 28)? As rationalization or regret for his decision to support efforts by the corporation responsible for the Valdez oil spill to reduce its liability for the environmental disaster? Or perhaps even as behind-the-scenes evidence of wilful blindness, which the editors of this collection define as ‘shutting out uncomfortable information’ (Bovensiepen and Pelkmans, this issue)? The first half of the title of this article, ‘Between the devil and the deep blue sea’, refers to the choice between two undesirable outcomes. It is also intended to invoke Freudenburg’s concern that he might have entered into a ‘Faustian bargain—selling [his] scientific soul, with the role of the devil being played by company lawyers who in the end would tell [him] what to publish’ (Freudenburg, 2008: 144). But Freudenburg seemed less troubled by the possibility that he might end up contributing to the company’s effort to reduce its liability for the shipping accident in Prince William Sound, the other unwelcome result.

In retrospect, however, Freudenburg acknowledges ‘being insufficiently self-critical or thoughtful, even though [he] remember[s] seeing [himself] as having been concerned with balance’ while taking notes on his discussion with the company (Freudenburg, 2008: 145). In an observation that would make Marcel Mauss smile, Freudenburg (2008: 146) compares corporate funding of academic research to a reciprocal relationship that creates expectations of an appropriate return: ‘It may seem only fair for us to ask what we can do for them, “without compromising” our objectivity.’ This is in keeping with studies that document the confidence scientists have that they can appropriately manage corporate influence on their decision-making while being sceptical of the ability of their peers to do the same (Kassirer, 2005: 72).

Freudenburg concludes that he had been ‘so focused on avoiding overt pressures to state predetermined conclusions that … [he] failed to recognize the power of more subtle forms of influence’, the central theme of the pair of articles he published on his interactions with Exxon (Freudenburg, 2005, 2008). He concludes that it may be ‘easier to influence scientists’ thinking by praising their objectivity and independence than by seeking to limit it’ (Freudenburg, 2008: 145).

Despite productively identifying the risk posed to researchers by the corporate exercise of soft power, Freudenburg fails to address the other consequences of his participation in the corporate initiative. Instead he emphasizes his responsibility to remain objective while conducting the proposed research. By focusing exclusively on his relationship to his corporate counterpart and the company that engaged them both, he ignores the potential repercussions of his actions for other parties whose interests he might also have taken into consideration. This includes the impact of the legal proceedings on fisherman in Alaska seeking compensation for the loss of their livelihoods, the possibility of reducing future oil spills by modifying corporate behaviour, or even protecting the interests of plaintiffs in other liability cases on the harm caused by corporations. This suggests that the commitment to objectivity may be double-sided; while intended to bracket out external influence on research, it may also discourage scientists from evaluating their political responsibilities in relation to the larger consequences of their work.

The long game

Freudenburg was one of ten scholars contacted by Exxon in advance of its appeal of the original punitive damage award.³ The others were well-known economists, psychologists, and law professors, including an eventual Nobel Prize winner in behavioural economics. While one of the participants accepted funding for his research and travel expenses but made a point of declining payment for his time, Freudenburg billed Exxon $240 an hour, or just under $2,000 per day (Zarembo, 2003). At least nine articles sponsored by Exxon appeared in peer-reviewed journals, many of which were subsequently assembled for publication in the volume Punitive Damages: How Juries Decide (Sunstein et al., 2002). Exxon’s expenses on the project have been estimated at more than $1 million (Zarembo, 2003), including support for large-scale simulations of jury verdicts that were subsequently criticized on other grounds (Eisenberg, 2001; Lempert, 1999).⁴ Although the contributors to the book asserted that they ‘maintained intellectual control [over] their research … they [acknowledged] that officials from Exxon commented on their drafts and coordinated meetings of the authors’ (Zarembo, 2003).

The articles funded by Exxon were subsequently cited in submissions to the US Supreme Court in State Farm v. Campbell (2003), which resulted in an influential ruling on punitive damage claims. In that case, which considered an insurance company’s failure to adequately represent the financial interests of its client, the court concluded that a single-digit ratio of punitive damages to compensation, that is, of less than 10:1, was appropriate in all but the most exceptional cases (State Farm v. Campbell, 2003: 425). However, in cases where compensatory damages were already substantial, such as the Exxon Valdez case, ‘A lesser ratio perhaps only equal to compensatory damages’, in other words, a ratio of 1:1, ‘might be warranted’ (State Farm v. Campbell, 2003: 425). Relying in part on the scholarship supported by Exxon, the, 2003 ruling in the State Farm case prompted the 9th Circuit Court to remand the punitive damages award in the Exxon Valdez case back to the Alaskan courts for reconsideration. The State Farm decision also began to influence jurisprudence on punitive damages in other cases.

Exxon spent 15 years appealing the original punitive damage award in the Valdez case (Freudenburg, 2005: 24). Its opposition to the $5 billion award from an Alaskan jury was ultimately quite successful, including an initial reduction to $4 billion. After the State Farm decision, the punitive damage award was decreased again, this time to $2.5 billion (Exxon Shipping Co. v. Baker, 2008: 481). Exxon continued to challenge the reduced award, arguing that it:

exceeds the bounds justified by the punitive damages goal of deterring reckless (or worse) behavior and consequently heightened threat of harm. The claim goes to our understanding of the place of punishment in modern civil law and reasonable standards of process in administering punitive law. (Summarized in Exxon Shipping Co. v. Baker, 2008: 490)

In its 2008 judgment on Exxon’s appeal of the punitive damage award, the Supreme Court reviewed relevant case law in the US and foreign jurisdictions, as well as the history of punitive damage claims:

The prevailing American rule limits punitive damages to cases of ‘enormity’, in which a defendant’s conduct is outrageous, owing to gross negligence, willful, wanton, and reckless indifference for others’ rights, or even more deplorable behavior. The consensus today is that punitive damages are aimed at retribution and deterring harmful conduct. (Exxon Shipping Co. v. Baker, 2008: 472; citations omitted, emphasis added)

In a 5–3 ruling, the award against Exxon was reduced to $500 million by the US Supreme Court, 10% of the original figure, saving the company a cool $4.5 billion. The punitive damage award was reduced from the initial ratio of 10:1 to a much more conservative ratio of 1:1, with each plaintiff receiving an average payment of $15,000. The court addressed its rationale in decreasing the penalty, noting: ‘American punitive damages have been the target of audible criticism in recent decades, but the most recent studies tend to undercut much of it’ (Exxon Shipping Co. v. Baker, 2008: 497). ‘The real problem, it seems, is the stark unpredictability of punitive awards … [given that] courts of law are concerned with fairness as consistency’ (Exxon Shipping Co. v. Baker, 1998: 472), an opinion commensurate with the findings of the researchers supported by Exxon (see Priest, 2002: 2–4).⁵

Judge Souter wrote the majority opinion in favour of reducing the punitive damage award, but in a footnote that subsequently attracted considerable attention, expressed reservations about corporate influence on the research cited by the defendants. The relevant portion of footnote 17 read: ‘Because this research was funded in part by Exxon, we decline to rely on it’ (Exxon Shipping Co. v. Baker, 2008: 501).⁶ Yet despite Souter’s disclaimer, the judgment of the court relied heavily on its prior reasoning in State Farm v. Campbell (2003). In other words, Exxon influenced the ruling in the Valdez case indirectly through materials introduced by a third party in the State Farm case. Thus the corporation successfully shaped jurisprudence on punitive damage awards. Exxon also correctly gauged the return on its investments in social science research, recouping its million-dollar outlay several thousand times over.

Souter’s intervention, which has become known as the ‘footnote 17 standard’, questions the influence of industry-sponsored research on legal proceedings. However, using language anticipating the controversial Citizens United (2010) judgment regarding corporate contributions to political campaigns, conservative scholars criticized the footnote 17 rule for infringing on the free speech of corporations (Aiosa and Majkowski, 2009).

The reveal

Freudenburg accepted Exxon’s offer, drafting an article in which he argued that legal proceedings on technological disasters do not benefit society as a whole given that their adversarial character tends to conceal and distort rather than clarify the causes of the disaster (Freudenburg, 2005: 19). Although his argument supported Exxon’s interest in questioning the value of punitive damage awards, his recommendation that corporations should strive for greater transparency in making the causes of disaster public ran counter to the privacy concerns of corporations and threatened to increase their exposure to liability claims (Freudenburg, 2005: 19). Consequently, the company declined to continue funding his research (Freudenburg, 2005: 10).

Although they discussed other articles Freudenburg might write on behalf of the company, neither party was particularly enthusiastic about the options, in part because Exxon was primarily interested in research supporting its position that ‘punitive damages were irrational and “out of control”’, claims Freudenburg (2008: 145) knew were not supported by the existing literature. He subsequently abandoned the article he had drafted, even though the corporate liaison encouraged him to publish it anyway (Freudenburg, 2005: 20). After accepting payment for his work, Freudenburg felt ‘reassured’ that he ‘had remained true to [his] principles and never did allow the potential for income to tempt [him] into writing something … inconsistent with the available evidence’ (Freudenburg, 2008: 145).

Despite having negotiated permission to make presentations and write about these issues after a moratorium of one year, Freudenburg only began to address his experiences six or seven years later, when Exxon’s lawyers subpoenaed his notes on the case in an independent matter (Freudenburg, 2005: 12). One reason he offers for the delay is that he would have been ‘embarrassed … if [his] conversations [with Exxon] became public’ at the time (Freudenburg, 2005: 18). Despite its value as an account of corporate influence on academic research, Freudenburg’s ‘confessional tale’ (Van Maanen, 1988) describing his interactions with Exxon has significant gaps or blind spots because of his failure to address questions that might otherwise have given him pause.

The case at hand is concerned with the decisions of a scholar whose research led him to conclude that the longevity of environmental impacts from oil spills is regularly underestimated, but nonetheless agreed to contribute to a project designed to limit the ability of jurors to impose substantial punitive damages against the parties responsible for such outcomes. Freudenburg also failed to address the implications of Exxon’s strategy of playing the long game by ‘seed[ing] the content of peer-reviewed scientific journals in ways that would shape legal decisions’ (Freudenburg, 2005: 4). He avoided discussing his willingness to further the economic interests of a corporation known to distort public opinion and national energy policy by promoting doubt and uncertainty about climate change (see Supran and Oreskes, 2017). The link between the two initiatives is significant, as Exxon’s success in funding research that questions the scientific consensus on climate change presumably influenced the company’s decision to commission academic research challenging the validity of punitive damage awards. Freudenburg’s neglect of these issues might be seen as a kind of ‘scientific refusal’ parallel to the way ethnographers may be reluctant to betray the political interests of their informants (Ortner, 1995). Despite his expert knowledge of these issues, Freudenburg asserts that his contribution to the larger corporate project does not contravene his ethical responsibilities as long as his research remains independent and objective.

The proximate mechanism through which Freudenburg excludes these issues from consideration in the two articles he wrote about his experiences was by concealing the name of the company involved and thus the specifics of the case.⁷ But there was another mechanism even more central to Freudenburg’s disavowal of responsibility for participating in a project intended to promote doubt and uncertainty about the legitimacy of punitive damage awards. By taking refuge in the ‘objectivity’ of scientific inquiry, Freudenburg reduced the range of ethical considerations associated with his involvement in the project. Although what Freudenburg meant by his invocation of objectivity was avoiding corporate influence on his research, framing his responsibility in this manner allowed him to ignore his complicity in Exxon’s larger project. Ironically, this kind of omission was precisely where Freudenburg made some of his most valuable contributions to sociology: the blind spots produced by the ‘diffraction of responsibility’ in industrial accidents, or cognitive heuristics that offer convenient shortcuts in everyday life but can result in dangerous oversights in complex situations. By remaining neutral and objective, Freudenburg excluded the larger social and political questions about Exxon’s intervention, and his contribution to it, from his purview.

Although initially concerned about the possibility of coercion by his corporate interlocutors, Freudenburg was mollified by their ‘constant praise for the fact that they considered [him] to be such a principled, credible scientist’ (Freudenburg, 2008: 144). A shared commitment to objectivity became the grounds for bonding between Freudenburg and his corporate handler, lending their interests the appearance of commensurability despite the obvious disjuncture between Freudenburg’s desire to contribute to scholarly debates and the corporation’s goal of reducing its financial liability in relation to the Exxon Valdez oil spill. Assurances from the company regarding its respect for his scholarly integrity were convincing enough for Freudenburg to collaborate with the perpetrator of the world’s largest oil spill.

Blind spot, or wilful blindness; how do we know the difference? Let me return briefly to the US Supreme Court’s definition of recklessness. Central to that discussion was the question of intent, referring to actions undertaken in ‘conscious disregard’ or ‘indifference’ to the risk of harm. The second standard was whether a ‘reasonable man’ can be expected to have ‘such knowledge, or reason to know’ the ‘high degree of risk’ involved. Both definitions of recklessness are applicable to the question of wilful blindness in the case examined here. Despite having ample reason to question Exxon’s agenda, Freudenburg chose to participate in the project. From the outset, Freudenburg’s vision was narrowed; in deciding to consult for the company, he also agreed to wilful blindness, although he called it something else: he called it objectivity.

Conclusion

‘Objectivity’, as Lorraine Daston and Peter Galison (2010: 17) explain, ‘has not always defined science.’ The purification of subjectivity and personal experience from scientific practice began in the mid-19th century as new technologies for communication expanded the community of scientific practitioners beyond the ‘boundaries of nationality, training and skill’ (Daston, 1992: 600). This transition demanded new standards for evaluating the credibility of scientific findings, including the advancement of replicable methods and techniques of quantification, while subordinating claims based on subjective observation. Scientific inquiry thus came to depend on a ‘central regulatory idea of detachment’ (Candea et al., 2015: 4).

Objectivity also has a specific history within the discipline of sociology, including Max Weber’s (1917) classic essay on ‘Science as a vocation’, which was intended to ‘delineate … between the role of the scientist and the role of the political advocate’ (Scott, 2005: 324). By the early 1960s, however, sociologists had begun to question whether Weber’s observations on the relationship between the university and the state, and therefore the public role of the sociologist, were still valid (Gouldner, 1962). Sociologists debated the question of political commitment in response to Howard Becker’s (1967) ‘Whose side are we on?’, which emphasized the need for sociological analysis to represent the perspectives of marginalized members of society rather than the views of those responsible for society’s most powerful institutions. But Becker was criticized for his failure to attend to the structural conditions responsible for oppression or human suffering (Gouldner, 1968). David Harvey (1995) subsequently made a similar point when criticizing a colleague and collaborator who prioritized the interests of a particular group of factory workers over the larger concerns raised by his critique of prevailing political-economic conditions. More recently, Michael Burawoy (2010) distinguished between several varieties of sociological practice, two of which are especially relevant to this discussion. The first is ‘public sociology undertaken in service of a goal defined by a client’ (Burawoy, 2005: 9), in which the ‘sociologist acts as a contractor to the policy-making customer … without necessarily having any moral or political commitment to the substantive political goals of the policy’ (Scott, 2005: 326). In contrast, Burawoy (2005: 10) argues that ‘critical sociology’ is defined, much like Becker’s (1967) earlier argument, in relation to representation and advocacy on behalf of subordinated or excluded social groups (see Scott, 2005: 326). As Burawoy (2005: 16) observes, ‘each type of sociology has its own politics’.

According to these definitions, Freudenburg’s work for Exxon fits into the category of public sociology, yet suffers from the same blind spot for which Gouldner (1968) previously criticized Becker (1967). The problem was not that Freudenburg agreed to provide a service to a client, but that in doing so, he lost track of the broader political responsibilities sociologists have toward the public. This indicates how the wilful blindness that accompanies the practice of objectivity, including the shadowy presence of issues arbitrarily excluded from consideration, helps to explain the decision-making processes of individual researchers who accept corporate funding for projects that promote strategic ignorance.

The case examined here sheds light on scientific contributions to literature that is intended to cast doubt and promote uncertainty about the risks posed by corporations. By drawing on legal documents and examples of life writing, it becomes possible to see how and why individual scientists elect to participate in such projects. In contrast to a simple, materialist explanation, that Freudenburg and the other contributors to the project were primarily motivated by monetary remuneration, or normative arguments about compromising the purity of science, that they were corrupt in undertaking this kind of research for hire, the particulars of this case study reveal a more complex dynamic. They suggest that this kind of work is encouraged or at least enabled by a commitment to objectivity that suspends critical attention to the political responsibilities of scholars and scientists.

In the effort to purify science from subjective influence, objectivity sanctions the exclusion of certain questions and concerns. The case presented and analysed here suggests that the avoidance of political considerations by scientists committed to the practice of objectivity may have detrimental consequences. These interactions reveal an unexpected connection between the wilful blindness of objectivity and the cultivation of strategic ignorance: that neutrality and detachment may inadvertently facilitate the promotion of doubt and uncertainty as part of a widespread corporate strategy. It also results in the following paradox: that objectivity remains central to the practice of contemporary science even when it contributes to strategic ignorance about risk, contradicting the fundamental scientific commitment to the production of knowledge.