Courts, practitioners, and scholars have recently focused on discovery costs in civil litigation. This produced recent amendments to the Federal Rules of Civil Procedure emphasizing that discovery requests be “proportional” rather than excessive. But this focus has ignored the fact that the information sought in discovery about the defendant’s liability is often consciously created by the defendant’s compliance measures before any litigation actually occurs (a point I am examining in depth in a current project). More importantly, potential defendants may create such information with an eye to ensuring that uncovering it in litigation or through similar regulatory intervention is costly or impossible. Accordingly, reducing the “costs” of discovery by limiting discovery to “proportional” requests may have the perverse consequence of making it even less likely that such information is ever uncovered, giving defendants greater incentive not to comply with the law in the first place.
This perverse consequence of reducing discovery costs is made vivid in Roy Shapira and Luigi Zingales’s new article examining DuPont’s treatment of C8, a chemical used in the production of Teflon-containing products. As Shapira and Zingales detail, DuPont learned in 1984 (and even earlier) that C8 posed potentially serious health risks because of the chemical’s bio-accumulative and bio-resistant properties. Among other things, C8 may cause birth defects in an infant’s eyes and nostrils, a fact discovered when DuPont examined the children of its employees. Nevertheless, DuPont continued to use C8, and even doubled its C8 emissions, for nearly three decades after learning about these risks, stopping its use in 2013.
Although this history is valuable, Zingales and Shapira go further and examine an interesting, and generally neglected, question. Did DuPont continue to use C8 because of ignorance and bad governance? Or, more problematically, because they considered it the rational thing to do? Here Shapira and Zingales use contemporaneous internal company documents, reasonable assumptions, and economic tools to conclude that DuPont chose to use C8 even though it was not socially optimal. The internal documents show that DuPont was well aware of its liability risk in using C8. But Shapira and Zingales surmise that the use of C8 remained rational for DuPont so long as the chance of being detected (and thus forced to pay heavy sanctions) was only a nineteen percent chance or lower. Continuing use of C8 unabated “was a case of ‘rational wrongdoing’: a decision that maximizes shareholder value ex ante, even though it is socially inefficient.”
What is interesting about this exercise from a civil-procedure perspective is the source of the internal documents Shapira and Zingales use in their analysis. The documents were produced during the discovery in the multiple suits filed against DuPont starting in the late 1990s. Shapira and Zingales argue that the “discovery process … provided a unique glimpse into the inner workings of a large company like DuPont.” Indeed, they later note that “[t]he main investigation of the health effects of C8 was made possible by the … initial litigation and the documents discovered during subsequent litigations.”
Compare this acknowledgement with the authors’ diagnosis of the failure of enforcement—“the main way corporations succeed in reducing their expected liability is by suppressing and distorting information.” The civil litigation and liability regime “as currently designed … creates bad incentives ex ante. Companies face incentives to suppress potentially damaging information, so as to keep the plaintiff lawyers away.” The general counsel of DuPont, for example, led workshops “on what not to document and share in regards to C8.” And those incentives to suppress information allowed DuPont to consciously emit C8 despite the social harm such emissions would cause, expecting this use would not be discovered. It is a small miracle that those internal documents were found in the first place. The current trend toward making discovery less costly for defendants only exacerbates those incentives to make such information very “costly” to discover in the first place.
Shapira and Zingales’s work is a sobering reminder that any discussion of discovery reform must acknowledge the regulatory role tort liability, through civil litigation, plays. They confirm that companies such as DuPont pay a great deal of attention ex ante to how potential litigation will affect their bottom line. Indeed, companies like DuPont may engage in “rational wrongdoing” based on predictions of how well litigation and other regulatory measures will be able to police their actions. Thus, while discovery is costly, courts and rulemakers cannot lose sight of its enormous regulatory benefits. Indeed, without discovery, this story could not have been told.