Putting Victims at the Centre of Liability Law: A Response to Siddharth Varadarajan

As a follow-up to his first op-ed, Siddharth Varadarajan wrote a second piece in The Hindu. Earlier, I raised some issues in my previous response and post here my thoughts on matters addressed in this later article.

His first point is that “[s]ince the government wants to accede to the IAEA's Convention on Supplementary Compensation (CSC), 300 million SDRs has been chosen as national cap so as to receive compensation from countries that are parties to the CSC beyond that.” My reading of the CSC is different. Art. III(1) of the convention states: “Compensation in respect of nuclear damage per nuclear incident shall be ensured by the following means: (a) the Installation State shall ensure the availability of 300 million SDRs or a greater amount that it may have specified to the Depositary at any time prior to the nuclear incident...” I understand this to mean that 300 million SDR is the minimum amount a state must ensure, not the maximum. The cap amount specified by the GoI in the bill cannot therefore be justified on this ground.

The maximum amount of 300 million SDR available through the CSC (assuming every member of the IAEA joins the convention) probably comes from Ben McRae (Reform of Civil Nuclear Liability, p.176). However, if you use the IAEA calculator and add all IAEA member states, you come up with a figure of ~375 million SDR. I am not sure how he got the figure of 50 million SDR as “the best a member state can hope to receive” but if you add Japan and India to the existing list of member states that have ratified the CSC, you land up with the amount of ~94.2 million SDR. Still it is not a lot which also argues for a higher total cap or no cap at all. I am also unable to see a connection between prohibiting discrimination (between national and cross-border victims of a nuclear accident) in distributing compensation and capping total liability for an incident.

He wants the operator cap to be raised and limited only to public sector operators. Under-provisioning of safety and the prospect of over-optimal share of nuclear power in India’s energy mix are the reasons given in support of this contention. The first has been dealt with before and my explanation in the following paragraph also relates to it. As for the second, the government’s authority to authorize (or deny authorization for) the construction of plants is not in doubt. A low cap as an incentive to build more plants can and surely will be countered by activists highlighting possible hazards and the final outcome will ultimately be guided by political judgment (rather than purely on economic grounds) as to the impact of its pursuit on public support.

He points out that insurance costs only amount to a small percentage of the total cost of the plant and third party liability in any case costs several times less than property insurance which is routinely purchased by nuclear plants around the world without hurting the viability of nuclear power. The issue, as I understand it, has primarily to do with availability. If insurance is not available up to the limit laid down by law, the company would have to make up the difference by setting it aside in bulk which will increase the operating cost. For example, if someone were legally obligated to purchase liability insurance for car accidents for a minimum amount of 1 crore but insurance carriers are only willing to offer coverage up to 10 lakhs, the only way to fulfill the requirement would be by setting aside the sum of 1 crore – 10 lakhs = 90 lakhs of his/her own money which is where it gets expensive (Insurance does not generate assets; it only protects what exists. So, it makes good sense for any business to attempt to keep the expense low). It was primarily this concern of availability that prompted the US Congress to mandate that plants maintain as financial protection “the maximum amount available at reasonable cost and on reasonable terms from private sources” and additionally provide for the state to indemnify them when the Price-Anderson Act was first enacted in 1957.

Insurance carriers are willing to underwrite plant property for amounts much higher than third party liability. Inferring the possibility of raising operators’ cap from the former would therefore be erroneous. The reason for this disparity is not so clear but a general explanation of how the business works would help to understand what factors might contribute to it. Insurance plays a role in a zone which lies between having no knowledge of the probability of an accident and having perfect knowledge of it. In the former instance, any protection offered would entail an unjustifiably high risk; on the contrary, if there was absolute certainty of its occurrence, there would be no point in insuring against it. Operating between these two extremes, insurers utilize various models to estimate the risk involved. These models are usually based on a mix of theoretical and real world data. The more actual data there is, the greater is the confidence with which predictions can be made and vice versa. The maximum amount an insurer is willing to underwrite depends on how much confidence it has in its model. For things like cars and aero planes which crash not infrequently, a fair amount of accident related data is available which gives insurers greater confidence in protecting against such risks and they are willing to cover higher amounts (relatively speaking). Nuclear property damage involves both routine property related risks and those specific to nuclear plants. The former are, likewise, more easily measurable and for the latter, they rely on reports of engineers and experts on the risks associated with particular equipment. Small safety incidents at nuclear plants are not uncommon (often involving non-nuclear damage) and that provides a significant wealth of data enhancing reliability of their predictions. Large nuclear disasters entailing third party liability, on the contrary, have only occurred on a few occasions and a major part of calculations is based on theoretical simulation which leaves them with less confidence about how far such estimates mirror reality. I suppose it is this difference that makes insurers much more wary about underwriting very high amounts for third party liability, a sentiment which would ring stronger in a new market where a priori experience is lacking.

When the Price-Anderson Act was first introduced in the US, private insurers only covered a fairly limited amount of $60 million and the state, in return for a mandatory payment of $30/1000kW from any facility, offered indemnity for an additional $500 million. The amount privately covered grew gradually over time as safety measures ensured that payouts were few and allowed the state to gradually withdraw its own role completely replacing it with a secondary layer of coverage financed entirely by the operators’ themselves, a system that exists to this day (the only differences are in the amount which comes to ~$12.6 billion including both first and second layer of coverage and a third layer was recently added to meet the international obligation under the CSC financed now by suppliers).

Regarding cl.17(b), I have already discussed that the key question is what impact retaining it will have on the willingness of suppliers to participate in the Indian market. Increasing that limit will also likely have a greater adverse effect on smaller suppliers who may prefer to stay away as they would have much more to lose than larger ones who may be able to internalize costs better. Domestic companies may also lose out to their foreign counterparts who, being less readily accessible to Indian law enforcement agencies, may be willing to risk more. If suppliers are made liable for total damages whilst keeping the limit much lower for operators, that too would lead to a moral hazard with operators (rather than suppliers) having a lower incentive to undertake adequate safety measures. This may be more so for state-owned entities owing to the fact that in the event of a disaster, the state, anxious to both deflect blame from itself and to pursue the prospect of obtaining large scale damages from the supplier, may be predisposed to use its control over the investigative machinery to downplay the role of the operator in the incident.

As he says, if India retains cl.17(b), it would have a choice to join the CSC with a reservation or choose to stay out of it. Comparison of this provision with article 4(1) of the South Korean law is not entirely appropriate. Article 4(1) of the South Korean act provides for the operator’s right to recourse “only insofar as there has been a willful act or gross negligence by the supplier of the materials concerned or by his employees” but this is overridden by article 4(2) which states that “If, in the circumstances described in Paragraph 1 of this Article, a special agreement has been made regarding rights of recourse, such agreement shall govern.” Cl.17(b), on the other hand, is distinct and applicable regardless of the operator’s agreement with the supplier under cl.17(a). Thus, the Indian bill is stricter but also, as a consequence, less flexible. Whether that is a good thing or not will depend on how strong the negotiating position of Indian operators will be vis-à-vis suppliers. For example, if the plant operator badly wants something from a supplier who is unwilling to provide it for fear of liability, the Korean law would allow its acquisition by providing a waiver but not its Indian counterpart.

Varadarajan calls for extending the filing period for compensation to 30 years from the present limit of 10 years. The main issue in other countries has been the unwillingness of insurers to offer coverage beyond 10 years. Mark Tetley cites three reasons to explain this reluctance: unwillingness of shareholders to commit to long term risk and uncertain exposure, wariness of second guessing future societal problems in an increasingly litigious climate and uncertainty of the security and solvency of the concerned insurer over such a long period of time. Extending the period is meaningful provided that a way is found to ensure reliability of payout possibly through state support either for all claims beyond an initial 10 year period or perhaps only as a backup.

Involvement of health and environmental experts is not specifically mentioned but the Claims commissioner can seek their assistance under cl.12(2) while holding an inquiry; a similar provision is missing for the commission which is only entitled to the assistance of officers and employees the Center provides as it may deem fit under cl.30(1). This suggestion needs to be considered though the main role of such experts will be in conducting studies and publishing reports in the aftermath of a disaster.

As mentioned in the last post, I cannot see any justification for allowing victims to file civil tort claims in addition to the recourse available under this Act as it duplicates the effort without offering any additional advantages to any concerned party (other than lawyers). I believe the proper answer is to amend cl.46 to preclude this possibility. Providing for judicial review may also be a pragmatic step to take as discussed in an earlier post.