Krishnan on Bhopal in the Federal Courts

Jayanth K. Krishnan, Indiana University Maurer School of Law, has posted Bhopal in the Federal Courts: How Indian Victims Failed to Get Justice in the United States, which is forthcoming in the  Rutgers University Law Review (2020):

35 years ago, the city of Bhopal, India, witnessed a horrific gas leak that originated from a facility operated by Union Carbide India Limited (UCIL), which had as its parent company, the American-based Union Carbide Corporation (UCC). Thousands were killed, with many more injured. 145 cases were filed throughout various U.S. federal district courts on behalf of the victims asserting that UCIL and UCC were liable. Eventually, these cases were consolidated through the Multi-District Litigation (MDL) process and placed onto the docket of federal Judge John Keenan. In 1986, Judge Keenan issued his famous forum non conveniens opinion, which stated that the Indian courts – and not the U.S. federal judiciary – were the proper venue for hearing these claims.

Between 1986 and 1993, Judge Keenan dismissed all of the other MDL-Bhopal cases he heard. Then, between 2000 and 2014 a set of distinct, non-MDL Bhopal matters appeared in front of Judge Keenan. In all of these too, he issued dismissals. Indeed, the original MDL-process – coupled with the existence of internal federal courthouse rules – created a type of path dependence, allowing for all of the Bhopal-Union Carbide matters to come before Judge Keenan.

The thesis here is that following the MDL-consolidation, Judge Keenan became only more deeply wedded to the position he staked-out back in 1986. Subsequent, non-MDL Bhopal plaintiffs, seeking an independent assessment of their claims, found themselves tethered to the initial MDL-decision from years past. The broader lesson – beyond just this case study – is that in order for deserving plaintiffs to receive a fresh review in federal court, there needs to be an alternative imagination for how to deal with later cases that, although seemingly connected, are nevertheless distinct from the earlier MDL-process.

 --Dan Ernst