A new paradigm for nuclear liability


back to issue

AMERICAN and Indian corporates upset by the provisions of India’s civil nuclear liability law should take a deep breath and think of two words: Bhopal and BP. Framed by these two totemic references to corporate liability for death and environmental destruction on a colossal scale, the Civil Liability for Nuclear Damages Act passed by Parliament in 2010 is one of the toughest pieces of legislation on the subject anywhere in the world. The only country to have a stricter law is Austria. But then Austria has no nuclear industry and its legislature has actually banned nuclear power as a source of energy.

Thanks to Bhopal and BP, the corporate-political nexus which might otherwise have gotten away with a toothless law was forced to adopt liability legislation with at least potential bite. By a peculiar coincidence, the trial court’s verdict in the 1984 gas disaster case and the oil leak in the Gulf of Mexico came just around the time the liability law was being drafted and entering public discourse. The lack of criminal justice for Bhopal, coming as it did in the wake of inadequate compensation for the victims, made ordinary citizens and, by extension, the political class, aware of a lacuna in the Indian legal system that needed urgent filling. And the manner in which the U.S. government announced it was prepared to throw the book at BP for the oil spill drew attention to the double standard involved in its insistence that American nuclear vendors could not be held liable for an accident caused by their products in India.

To be sure, the Indian law still suffers from various weaknesses from the standpoint of victim protection and compensation and even civil and criminal liability in the event of a nuclear accident. And yet, it marks a positive step away from the standard template for liability legislation that has been in vogue across the world since the 1950s, when the international nuclear liability regime first took shape under the watchful eye of the American nuclear industry and U.S. government.


Broadly speaking, this international regime has five main features. First, all liability for nuclear accidents is channelled onto the operator of a nuclear facility rather than the manufacturer of the reactor. Second, the operator is given an extremely limited ‘right of recourse’ against suppliers in the event of an accident. Third, as a result of bearing the burden for this channelled liability, an upper cap is fixed on the operator’s liability which is typically much less than the true damage a major nuclear accident would cause. Fourth, ordinary tort law is usually set aside, thereby disallowing fault-based claims by victims against the operator or its supplier. Supplementing these is a fifth element: legal jurisdiction belongs to the courts of the country where the accident takes place and victims are barred from approaching or filing a claim in a forum elsewhere such as a court in the country of the nuclear supplier.

This fifth element is especially important for American companies since Bhopal, where Indian victims approached a U.S. court, is the ghost that looms large. ‘While ultimately the court declined to take jurisdiction,’ Ben McCrae, legal counsel for the U.S. Department of Energy, notes, ‘this was not because it doubted its capacity to do so: it basically waited to ensure that there was an adequate remedy available in India.’

Even though these principles were clearly designed to benefit the supplier, they were pushed globally by the U.S. in the name of lowering insurance premia (by avoiding the duplication of risk bearing) and speedy compensation for victims. This is because the quid pro quo for channelling was the rule of ‘strict liability’ under which the operator is liable even if he is not at fault. Victims benefit from this rule since there is no ambiguity about who must pay. They do not have to establish, for example, that a particular accident was caused by negligence on the part of the operator or anyone else in order to be able to legally press their claim for compensation.

But as Tom Vanden Borre – one of the leading environmental law scholars to challenge the dominant liability regime – has argued, channelling was ‘not introduced to protect the victims of nuclear accidents, nor to reduce the insurance costs, but to protect the American nuclear industry.’ The irony is that even as it has pushed the regime of legal channelling on the rest of the world, the U.S. system of economic channelling of liability allows tort claims as well as an unrestricted right of recourse for the operator. That is how, for example, Metropolitan Edison, the operator of the Three Mile Island reactor, sued its supplier, Babcock & Wilcox, after the 1979 accident.


Though civil liability as an issue never figured in any of the discussions India and the United States had during their negotiations over the 18 July 2005 nuclear deal, the American nuclear industry quickly flagged the absence of a suitable Indian liability law as a matter of key concern. ‘Addressing nuclear international liability issues must be a key component of U.S. policy on nuclear power and non-proliferation,’ Omer Brown III wrote to US Energy Secretary Samuel Bodman on 1 March 2006 on behalf of the Contractors International Group on Nuclear Liability (CIGNL). ‘The U.S. Government should urge other States without comprehensive nuclear liability regimes (such as India) to adopt nuclear liability laws containing the international standards for such legislation. In particular, these States should be urged to join the International Atomic Energy Agency’s 1997 Convention on Supplementary Compensation for Nuclear Damage (CSC).’

CIGNL is an ad hoc group of major nuclear suppliers and contractors, including Babcock & Wilcox Company, Bechtel Power, Fluor Corporation, General Electric and the Westinghouse Electric Company. Brown’s letter also warned the American administration that the absence of appropriate liability laws abroad meant privately owned U.S. companies would not have a level playing field against state-owned competitors from elsewhere. ‘Some European Union States (whose nuclear suppliers are government-owned) are less concerned about liability, and actually oppose improving nuclear liability protection via the CSC as a way to discourage U.S. suppliers and contractors from working in Eastern Europe and elsewhere, such as India.’


Thanks to pressure from the CIGNL, Washington started pushing India to adhere to the CSC and adopt the conforming domestic liability legislation that this would involve. India waited till the successful completion of the ‘123’ negotiations and the adoption of a waiver by the Nuclear Suppliers Group before giving the U.S. a written commitment on the matter. Finally, on 8 September 2008, Shiv Shankar Menon, who was foreign secretary of the time, sent his American counterpart a letter of intent promising, inter alia, that India would take all the steps necessary to adhere to the CSC.

That India was able to make such a commitment owed as much to American pressure as to its own internal assessment on the attractiveness of the CSC. Long before the Indo-U.S. nuclear deal of 2005, the Department of Atomic Energy had conducted a review of various international liability practices and conventions under the chairmanship of Prof. V.B. Coutinho and S. Rajagopal. The impetus for this study was the work which had just started in Tamil Nadu for the construction of two large Russian reactors at Koodankoolam. One of the questions which arose at that time, according to DAE officials, was the obligations for India in the event of an accident which might affect victims in Sri Lanka. The report, which was submitted to the government in November 2001, is still classified. But Prof. Coutinho has confirmed that he and Dr. Rajagopal recommended that India sign the CSC. They also drafted a model domestic liability legislation.


Given the context, the choice of CSC, which was framed by the IAEA in 1997 in the aftermath of the Chernobyl disaster, was hardly surprising since it was the only nuclear liability convention to deal properly with accidents having transboundary implications. Even in the absence of such a convention, the legal obligations of a state where an accident occurs are clear. Ever since the Trail Smelter arbitration settlement between Canada and the U.S. in the 1940s, it has been a settled principle of international environmental law that a polluter must compensate his victims wherever they may be. But the attractiveness of the CSC was its provision for international contributory funding in the event of an accident causing damages beyond 300 million Special Drawing Rights.

The CSC specifies in its annexe a draft law that countries seeking to accede must model their domestic nuclear liability legislation on. The Indian law broadly follows the structure of the annexe but makes two improvements on the CSC annexe. First, it amplifies, in Section 17(b), the grounds on which a nuclear operator can exercise a right of recourse against his supplier. Where the CSC allows this only if there is a prior contract to this effect or there has been wilful damage by an individual, the Indian law allows the operator to sue his supplier for recovery of any damages he is obliged to pay if an accident has resulted ‘as a consequence of an act of supplier or his employee, which includes supply of equipment of material with patent or latent defects or sub-standard services.’

The second improvement is Section 46, which explicitly states that nothing in the civil liability law will prevent the operation of other laws in force in the country. According to the Indian government, this section makes it clear that criminal liability in the event of an accident remains, as indeed do tort claims.


The American government says these two provisions make the Indian law incompatible with the CSC. India, which has since signed the CSC, disagrees. Indian officials say the convention cannot stay the operation of ordinary Indian legal processes. Torts and criminal liability, therefore, are not excluded by the CSC. All the Indian legislation has done, they say, is to explicitly state in Sections 17(b) and 46 something that would implicitly apply anyway.

Though the American objections have been framed with a view to the immediate problem of minimizing the exposure of U.S. companies intending to sell reactors to India, Washington’s overriding concern is that the new Indian liability law might become a template for other countries embarking upon an expansion of their civil nuclear energy programmes. The existing international liability regime might have had some justification when nuclear energy was in its infancy and some element of public subsidy – which is what the capping and channelling of liability means – may have been required. But four decades on, the case for a new paradigm is compelling.

‘The operator’s incentives to prevent nuclear accidents may be affected by the partial internalisation resulting from the nuclear subsidy,’ argue Michael G. Faure and Karine Fiore. ‘Indeed, since the operator does not take into account all of the risks he generates, his behaviour might be inadequate to prevent accidents in an optimal way. Rationally, he will adopt the level of prevention corresponding to the risks he generates. Therefore, if he takes, as a reference, an underestimated level of risks, his preventive actions are necessarily maladjusted and, thus, insufficient to impede an accident. Indeed, the optimality of his level of prevention is determined by the optimality of the level of the considered risks. As a result, a sub-optimal estimation of risks leads to a sub-optimal level of prevention.’1


If the inefficiency of liability channelling and capping is well established in the environmental economics literature, Evelyne Ameye, in a recent article makes a safety-cum-engineering argument in favour of suppliers remaining liable for accidents their products may cause.2 This can be done in two ways. Liability for an accident can still be channelled on to the operator but his right of recourse in the event of supplier negligence is left unrestricted. The Russian Federal Act on Atomic Energy, for example, does not impose a limit upon the operator’s right of recourse.3 South Korea’s liability legislation also allows operators to recover damages from suppliers in the event of negligence.

A second way would be to allow victims to sue suppliers for fault-liability under tort law so as to win damages over and above what the operator pays through strict liability. Thus Germany, a party to the Paris Convention on nuclear liability, entered a reservation stressing its right, under national law, to hold persons other than the operator liable for nuclear damage. Besides, several conventions on environmental damage – such as the 2003 Kiev Protocol on industrial accidents in transboundary waters – now explicitly provide for strict as well as fault-liability to run side by side.


Ameye argues that channelling can no longer be justified on the grounds of nuclear power being an infant industry. Nor is it healthy to exclude suppliers from the liability chain when nuclear technology is rapidly evolving. ‘Given the increasingly complex designs of the new generations of nuclear power plants, it is… both legally and realistically incorrect to maintain the heavy burden of legal channelling upon the nuclear plant’s operator … To the extent that design knowledge becomes more hermetic, it will be hard to sustain the operator’s liability for risks he is not aware of or, even worse, for risks he cannot perceive.’ This is especially so when all major nuclear accidents in the past – Windscale, Three Mile Island and Chernobyl – have occurred, in part, because of design flaws.


The Indian law, in a sense, incorporates both of Ameye’s suggestions. The operator has an expanded right of recourse. And the right of victims to file tort claims seems to be intact, providing a further incentive for all those whose action or inaction could contribute to an accident to take due precaution. Ameye and other scholars like Faure and Fiore have also argued why better stringent regulation – though desirable in and of itself – is no substitute for stricter liability rules:

‘The role of the nuclear safety authorities may fill this gap… However, [their] rules are regulatory and so, do not provide incentives, in a strict sense… Even though the purpose is identical (the prevention of accidents), regulation and the liability rules do not employ the same means to reach it. Regulation has a compulsory and external dimension. It is implemented and controlled by an outside and superior authority. The operator is only required to apply and respect the standards it imposes. Therefore, the operator’s behaviour changes with the changes of the norms.

On the contrary, with a liability rule, the operator is led to modify, on his own, his behaviour in accordance with the risks he generates. Liability rules are thus more dynamic than safety regulation (which is more difficult to change and, as a result, more static). Especially in such a complex industry, it is often argued that liability rules are more efficient, from the prevention of accidents viewpoint. Because the operator has more information about his activity, and more information about its risks, as compared to anyone else (e.g. the regulators), he would be in a better situation to evaluate his risks, and thus, his corresponding level of prevention’ (emphasis added).

Without any reference to the latest international scholarly thinking on the subject, Indian legislators, driven mostly by the democratic pressure of accountability, pushed for a liability law that breaks with the existing international regime in two vital ways. The supplier of nuclear equipment has still not been made fully or directly responsible for an accident. But the partial opening of a window is path-breaking. Nuclear suppliers can choose to react to the Indian law in one of four ways.


First, they could decide to keep off the Indian market altogether, an unlikely outcome given the multi-billion dollar order book India has in its hands. Second, they could use their political clout in America and India to try and amend the Indian law to protect their interests. Such an effort may succeed if the correlation of political forces in India shifts dramatically but is unlikely. Third, they could simply jack up their prices to reflect a higher insurance premia. But this response would raise the cost of nuclear power compared to other sources of energy and a vigilant and cost-conscious polity may then be forced to scale back its ambitious nuclear plans. Finally, they could accept the inevitable and enter into a partnership with the Indian operator to ensure their designs and systems are the safest and most secure possible.

The purpose of a strict liability law is to raise the safety bar. Let us hope the Indian law does that for India. And for the rest of the world too, once it becomes clear that the new liability paradigm India has established is precisely what every country going down the nuclear path needs for itself.



1. ‘An Economic Analysis of the Nuclear Liability Subsidy’, Pace Environmental Law Review, 2009.

2. ‘Channelling of Nuclear Third Party Liability Towards the Operator’, European Energy and Environmental Law Review, 2010.

3. Alexander Matveev, ‘The Russian Approach to Nuclear Liability’, International Journal of Nuclear Law, 2006.