Discussion notes by Heba Islam
Professor Bharadwaj presented a paper on the Bhopal gas tragedy, with particular attention to in utero exposure. He kicked off the event by pointing out that the consequences of environmental externalities in poorer countries can’t be absorbed by neoclassical solutions, pointing out the need for a different approach to understand the impact of the disaster. He added that it was important to take the duration of the impact, especially long term impacts, into account. Providing context as to how the disaster occurred, Bharadwaj noted that with the Green Revolution in India came increased chemical production of pesticides, which required storage. When these storage facilities exploded in the case of the Union Carbide pesticide plant, state responses were mixed. After a certain point, data collection on the effects of the tragedy stopped. Bharadwaj questioned many of the assumptions made in the studies of the explosion and its impact, including especially that the area of the gas exposure was determined to be a paltry three kilometers. At the same time, the central government became the sole representative of the victims within this predetermined area, which limited any future claims of liability, whether civil or criminal. Bharadwaj’s study does not use the predetermined area of spread, rather focusing on survey data that takes into account migration to and from the region of Bhopal as well as different age groups. He found that effects such as pregnancy terminations and cancer were found in regions up to 100 kilometers from the site of the explosion.
Usha Ramanathan took a different approach to understanding the impact of the disaster. Impacted communities were invisibilized, she said, and by the time people came to ask questions about the causes and effects of the tragedy, it became extremely difficult to address them. As Union Carbide became involved in managing hospitals and clinics in the region, this invisibilization became more pronounced, as records and documents were taken away, manipulating what data was permissible for collection, and preventing this data’s neutrality. She added that issues such as impact on mental health were left unaddressed altogether. Ramanathan emphasized that this invisibilization meant that contemporary accounts and narrative accounts were essential in determining the validity of data collected on the incident. She questioned how the three kilometer parameter was established, adding that this limited geography affected questions of who received compensation and how much was received. The mobility of many was affected, unable to move in case it would prevent them from receiving compensation. The court case itself came to be a test case for the liability of multinationals in such disasters, and was categorized as an “accident” in a rush to reassure the market. A question was also raised about how unborn children were to be understood before the law in the legal case, since in utero exposure and mutagenic effects were already acknowledged at the time.
Professor Shareen Joshi focused on the difficulties in working with data that addresses chronic and degenerative factors of adult health affected by in utero conditions, as well as other impacts very difficult to pin down – wages, mental health, and education for example. She added that economists are trying to expand on in utero conditions as well as mental health, IQs, test scores and wages, creating an interdisciplinary collaboration of analysis. But the problems with data remain, she added, saying that more data, better data, and more refined protocols were needed – though this was becoming more difficult to gather. Even while more money is spent on collecting data in India, much of it remains unusable. One such problem is “noisy data”, for example in the case of infant mortality in the region – where Union Carbide is not the sole offender and the regular use of pesticides and fertilizer (“smaller disasters”) continue to impact mortality through in utero exposure. Joshi said that data on children’s health, corporate activity, and Indian courts that remain overburdened and under resourced needs to be linked, in order to determine where discrimination and bias emerged in the process of studying 3000 cases. Data, she pointed out, is not currently incentivized for any accountability and asked how we can make data useful for accountability.
In the ensuing discussion, Professor Das asked whether the problem concerned the need for better data, or better data architecture. The idea of evidence itself came under scrutiny, including concerns that the poor don’t have the resources to produce evidence that would convince the court, what the rights of children were once evidence was presented, and as Maya Ratnam added, how to deal with the ‘state of knowledge’, when we know data collected at the time/in the aftermath was manipulated or disappeared. Jishnu Das asked how we can think about the ‘discounting’ of compensation: When children are affected, how much do we compensate today for something that might happen thirty years from now? Economics doesn’t have a distinction between what happens to an individual now, and their child thirty years from now. Joshi answered by saying that the need to address this was exactly why an evidence based approach was necessary. Ali Khan remarked that Economics as a discipline created a discourse of invisibilization of the myriad effects of tragedies like the Bhopal disaster in the first place, thus raising the question: Does data contribute to invisibilization of certain patterns or making them more discernable? Bharadwaj replied that while he would agree that data when mixed with political objectives can be invisibilizing, he would be hesitant to refer to this as a general rule. To make data useful in the case of the law, Joshi added, judges need to be trained to both interpret data and understand its limitations. Veena Das concluded with considering disciplines and the different lives of data (where does the data go?) How can we turn from general critiques of disciplines to specific questions? Data and evidence need to stimulate conversation about how we should use it before we use it.