The Ecological Liability of Corporate Groups:
Comparing US and European Trends
Professor da University of Zurich
Visiting Scholar89 (HLS)
"Bhopal", "Amoco Cadiz" and "Seveso" have become buzzwords in the general debate about the ecological hazards of modern industry. In addition, they have a very specific common legal denominator: all three cases involved the question of parent liability vis-à-vis creditors of a subsidiary. Of course, the issue of corporate group or parent liability is not limited to ecological catastrophes. It can arise in connection with all tort and contract claims against a subsidiary. Yet, the problem is particularly acute in the field of ecological liability, as damages resulting from ecological disasters can run very high, thereby easily exceeding the assets of the primarily responsible subsidiary.
In this chapter I try to outline and evaluate some evolving trends in the USA and Europe with regard to parent liability for ecological damages. I will first describe recent developments in US law, mainly in connection with the Comprehensive Environmental Response, Compensation and Liability Act (CERCLA). Second, I will focus on concepts of parent liability in Europe. Third, I will deal with the need for an adequate paradigm to assess parent liability regimes. I will make reference to the economic analysis of law and discuss whether systems theory can provide additional insights. This epistemological excursion will serve as a basic for exploring the idea of self-organization in connection with parent liability law. Finally, I will conclude with some general comments.
TRENDS IN US LAW
Piercing the veil in corporate law
The overarching legal mechanism in US parent liability law is that of "piercing the corporate veil" (Clark, 1986: 71ff.; Blumberg, 1987: 681-683; Hofstetter, 1990b: 591-593). Its uncertain and metaphoric name is partly reflected in its application: US courts refer to it in very different situations, and it is not easy to detect a coherent theory behind the uses of the concept. US piercing law requires basically that a parent company either controlled a subsidiary for illegitimate purposes (e.g. fraud) or controlled it in such a complete manner that the control seems illegitimate1. The courts use words like "sham", "shell", "agent" or "instrument" to describe the liability-triggering dependency of subsidiaries (Blumberg, 1987: 107). They look at a myriad of factor which may together indicate the critical level of control. Such factors are the percentage of stock ownership, the existence of interlocking directorships, the commingling of assets between the parent and the subsidiary, the total financial dependency of the subsidiary, its inadequate capitalization or the non-observance of its formal legal requirements (Blumberg, 1987: 185ff.).
In the famous case of Lowendahl v. Baltimore & ORR2 a New York court described the required level of control in the following words:
Control, not merely majority or complete stock control, but complete domination, not only of finances, but of policy and business practices in respect to the transactions attacked so that the corporate entity had at the time no separate mind, will or existence of its own.
Blumberg notes though, that there is a greater willingness on the part of US courts to pierce the corporate veil in tort cases than in others (Blumberg, 1987: 110-111). A recent judgment handed down in the case involving the "Amoco Cadiz" oil spill off the coast of France in 1978 further corroborates this view3. Even though the court for the Northern District of Illinois did not make specific reference to the piercing doctrine, it applied a similar instrumentality test and, moreover, seemed to gear it to the tort considerations in question. The court noted:
43. As an integrated multinational corporation which is engaged through a system of subsidiaries in the exploration, production, refining, transportation and sale of petroleum products throughout the world, Standard is responsible for the tortious acts of its wholly owned subsidiaries and instrumentalities, AIOC and Transport.
44. Standard exercised such control over its subsidiaries AIOC and Transport that those entities would be considered to be mere instrumentalities of Standard. Furthermore, Standard itself was initially involved in and controlled the design, construction, operation and management of the Amoco Cadiz and treated that vessel as if it were its own.
45. Standard therefore is liable for its own negligence and the negligence of AIOC and Transport with respect to the design, operation, maintenance, repair and crew training of the Amoco Cadiz.
46. Standard therefore is liable to the French claimants for damages resulting from the grounding of Amoco Cadiz.
A comparable cause of action, referred to as "multinational enterprise liability", was introduced by the government of India in its Bhopal complaint in the USA (see Westbrook, 1985: 321ff.). The action was dismissed on the basis of a ruling of forum non conveniens4.
Parent liability concepts under environmental law statutes
CERCLA (42 USC § 9601-9657 (1982)) provides a regulatory programme controlling hazardous waste disposal on US soil. It also established a "Superfund" to the finance environmental clean-up at hazardous waste sites (Superfund Amendments and Re-authorization Act (SARA) of 1986). Section 107 (a) of the Act imposes liability on:
(1) the owner or operator of ... a facility,
(2) any person who at the time of disposal of any hazardous substance owned or operated any facility at which such hazardous substances were disposed of, ...
(3) any person who ... arranged for disposal or treatment, or ... transport ... of hazardous substances owned or possessed by such person ...
From the beginning, the courts have emphasized that the statutory purpose of CERCLA requires a liberal construction of its liability provisions5. This tenet also influenced their approach to the question of parent liability under the Act. The courts have chosen two main routes to parent liability: either the statutory construction of the parent as an "owner or operator" or the piercing of the corporate veil6.
In the first alternative US courts used the language and the purpose of CERCLA to reach beyond the constraints of piercing the corporate veil law. In State of Idaho v. Bunker Hill Co.7 the court concluded that the parent and its subsidiary were "so intertwined" and the parent "so controlled the management and operations" of the subsidiary that the parent constituted an "owner or operator" for purposes of CERCLA. The court also found the parent to be an "owner or operator" because it was "intimately familiar" with hazardous waste disposal at the subsidiarys facility and had the "capacity, if not total reserved authority, to make decisions and implement actions and mechanisms to prevent the damage". Similar rationales were enunciated in subsequent decisions8. The District Court of Rhode Island in US v. Kayser Roth Corp.9 first stated that a parent corporation cannot ordinarily be deemed and operator solely on the basis of its status as a shareholder. But the court here considered as critical the fact that the parent had exercised "pervasive control" ("practical total control") over its subsidiary through, among other things:
(a) total financial control, including collection of accounts payable;
(b) restrictions on the subsidiarys budget;
(c) its directive that subsidiary-government contacts, including in environmental matters, be funnelled directly through the parent;
(d) the requirement that the subsidiarys leasing, buying or selling of property first be approved by the parent;
(e) its policy that the parent approve any capital transfer or expenditures greater than US $5000.00;
(f) placement of parent personnel in almost all of the subsidiarys directorships or officerships, as a means of ensuring that the parents corporate policy was exactly implemented and precisely carried out.
The court also considered "illustrative" the fact that the parent controlled all actions with regard to environmental matters, such as the release of hazardous substances and related investments10. In addition, the court emphasized that actual knowledge of the release is not required for establishing liability under CERCLA11.
Even though the control test under CERCLA resembles that in traditional piercing law, it, by contrast, does not seem to require indicia of illegitimate conduct. A very firm grip on the subsidiary, particularly if extending to environmental matters, seems to be sufficient in order to render the parent liable as an "owner or operator".
The ruling in US v. Fleet Factors12 further supports the view that de facto control over the environmental operations of a company can be determinative for extending CERCLA liability. The Court of Appeals of the 11th Circuit held that even:
a secured creditor will be liable if its involvement with the management of the facility is sufficiently broad to support the inference that it could effect hazardous waste disposal decisions if it so chose.
Hence, the court observed that liability could even attach where secured creditors had the power and discretion to affect hazardous waste decisions of a company. This means that the mere existence or potential of control, not its exercise, may be dispositive.
Still, there seem to be limits to the scope of owner operator liability under CERCLA. In Edwards Hines Lumber Co. v. Vulcan Materials Co.13, for example, Judge Easterbrook held that an independent contractor or joint venturer who designed the manufacturing facility and trained the facilitys workers did not exercise control over the facilitys operations and, therefore, was not an operator under CERCLA14.
Before applying the second alternative, that is the piercing concept under CERCLA, the courts first have had to address the question of whether to apply state law or whether, instead, the development of federal common law was justified (choice of law). Citing the decision in Re Accushnet River & New Bedford Harbor Proceedings15, the court in US v. Kayser-Roth Corp.16 held that a uniform federal rule was warranted in view of the "overriding federal interest" in promoting environmental protection. It went on to state that piercing the corporate veil was an alternative route to rendering the parent company at hand liable as an "owner". The courts ruling was two-pronged. It first relied on the basic principle that under federal law, "a corporate entity may be disregarded in the interest of public convenience, fairness and equity". In addition, it considered determinative the same indicia of control already cited in connection with the qualification of the parent as an "owner or operator".
CERCLA is, of course, just one of many environmental liability statutes in US federal or state law. Similar concepts of parent liability seem to emerge in the context of other statutes. A leading case involving the issue of group liability was State Department of Environmental Protection v. Ventron Corp.17, which arose under the New Jersey Spill Compensation and Control Act. The New Jersey Supreme Court held that:
Given the extended liability for the Spill Act, we conclude that the legislator intended that the privilege of incorporation should not, under the circumstances that obtain here, become a device for avoiding statutory responsibility. A contrary result would permit corporations, merely by creating wholly owned subsidiaries, to pollute for profit under circumstances when the legislator intended liability to be imposed.
The newly enacted Federal Oil Pollution Act of 1990 establishes, inter alia, strict joint and several liability of ship owners and operators for oil spills in US waters (Anderson & Wethmar, 1991: 1001ff.). Yet it does not contain a parent liability provision. Given the trends under CERCLA, and assuming that the Amoco Cadiz judgment might have some effect as a precedent, it can be expected that the Oil Pollution Act, too, will be subjected to a liberal standard of parent liability (Anderson & Wethmar, 1991: 1005-1007).
In summary, then, we can say that in the field of general corporate law the US courts still stick to a traditional piercing the corporate veil test which requires indicia of control and some form of illegitimate conduct. In tort law and under environmental law statutes, however, courts seem to have gone out of their way to be more liberal in the construction of parent liability. In this, they are able to adjust their parent liability concepts more specifically and flexible to the policies underlying tort and environmental law.
TRENDS IN EUROPEAN LAW
Corporate law concepts
Unlike in the USA, where parent liability has had a relatively dynamic evolution in the context of specific liability laws and statutes, European trends are concentrated around corporate law, and as a result piercing the corporate veil is also well known in European jurisdictions. But corporate group law has moved beyond this concept (Hofstetter, 1990b: 576ff.), and in this move a leading role has been played by Germany.
The pioneering role of Germany in the area of corporate group law is undisputed. Even though the codification of the German Konzern-recht in 1965 turned out to be ineffective in various respects, most notably as regards the provisions for de facto concerns18, this legislation seems to have inspired the German legal culture. In the area of parent liability, the German Federal Court has enunciated special rules for so-called "qualified de facto concerns". In Autokran19 it held that where a parent company was "permanently and extensively" involved in the management of a subsequently bankrupt subsidiary, a rebuttable presumption exists that the parent did not show sufficient consideration for the subsidiary. Hence, unless the parent is able to defend itself successfully, it is directly liable to creditors for the subsidiarys obligations. In this decision the German Federal designed a way around the notorious enforcement problems in connection with parent liability provisions for highly integrated de facto concerns. It did this by resorting to two legal techniques: the shifting of the burden of proof to the parent company and the granting of a direct claim against the parent. In its subsequent Tiefbau decision20, the court restricted the parents defence to situations in which the subsidiarys "losses were caused by circumstances outside the parents managerial control"21.
French bankruptcy law, Swiss corporate law and the draft for a Ninth Directive in the European Community (Hofstetter, 1990b: 585-590) work with "de facto director" notions that put parent companies under the same fiduciary duties as corporate directors and officers. Parent liability requires that the parent acted like a director, that is to say, involved itself in the day-to-day operations of the subsidiary. Furthermore, creditors have to prove that in so doing the parent violated its fiduciary duties and thereby caused damage to the subsidiary. This approach to parent liability, even though promising, faces significant enforcement problems due to the heavy burden of proof placed on creditors. In highly complex and integrated modern corporate groups it can become almost impossible for outsiders to trace and prove past violations of the subsidiarys interests by the parent (Hofstetter, 1990b: 593ff.). Shifting the burden of proof to the parent company in analogy to the German qualified concern doctrine might, thus, be the only way to rebalance the situation22.
Developments in the area of environmental liability law
The general emphasis on corporate law notwithstanding, European courts and scholar have approached parent liability from other angles as well. French labour law23, or the Australian Eumig decision24, are examples outside ecological liability. With regard to environmental statutes there have, as far as I can see, been no specific developments (see Bender & Sparwasser, 1990: N 52, 86; Schmid, 1990: 1ff.; Tercier, 1990: 73ff., 161-162; Krämer, 1991: 85ff., 335ff.). Scholarship, however, has emphasized the potential of using general civil law constructions in order to render parents liable for their subsidiaries obligations. This could also have implications for ecological liability. Eckard Rehbinder (1969: 85ff., 133ff.), for example, has elaborated a myriad of civil law devices that could lead to the liability of parent companies for specific claims against their subsidiaries. Particular plausibility goes to the concept of vicarious liability (Rehbinder, 1969: 529ff.) which could also play an important role in the field of ecological liability. Under this notion, which is also advocated by Swiss scholar (Hofstetter, 1990b: 590-591), parent liability could attach for tort claims against its subsidiary. It requires proof that the parent had management control over at least those activities of the subsidiary which caused the tort (Hofstetter, 1990b: 595-596).
An instructive case touching upon the parent liability issue developed in the aftermath of the Seveso catastrophe in 1976. The community of Seveso brought an action against Givaudan, the immediate parent of the primarily responsible Italian subsidiary Icmesa, but also against the latters ultimate parent, Hoffmann-La Roche, in Switzerland25. The complaint against Hoffmann-La Roche contained several causes of action. One was that Hoffmann-La Roche had assumed the obligations of Icmesa by publicly declaring after the accident that Hoffmann-La Roche always backed its subsidiaries in need. Another cause of action was vicarious liability. The plaintiff maintained that Icmesa had been an agent or subordinate to Hoffmann-La Roche and that the latter was therefore vicariously liable for Icmesas negligence (Art. 55 of the Swiss Code of Obligations). An alternative legal construction brought forward by the plaintiff was based on Art. 55 of the Swiss Civil Code. It provides that companies are responsible for the torts of their directors and officers. The plaintiff maintained that the directors and officers of Givaudan, who were in charge of production safety at Icmesa, were also constructive officers of Hoffmann-La Roche, because they acted directly in the interest of the ultimate parent. The was settled, however, before the Swiss courts could render an opinion on it.
In general, though, it can be said that, in contrast to the USA, European jurisdiction focus primarily on corporate law based parent liability concepts. They are thereby moving beyond traditional piercing notions. They show sensitivity to the specific risks for creditors in connection with the parents management influence over the subsidiaries. Devices such as de facto director liability or the German "qualified concern" doctrine seem particularly well suited to deal with the general organizational risks faced by creditors in connection with corporate group integrations. As to ecological liability of parent corporations, the vicarious liability concept, as asserted in the Seveso action for example, exemplifies a second though relatively weaker trend in European law attempting to tackle parent liability by way of general civil law constructions.
IN SEARCH OF A PARADIGM FOR ECOLOGICAL PARENT LIABILITY
Economic analysis of parent liability law
In order to evaluate parent liability in environmental law and elsewhere, one needs a reference model. Traditional legal thinking does not suffice. It tends to be formalistic and is of little value when it comes to policy questions of optimal social engineering (see Adams, 1985: 8ff.). The rise of law and economics, therefore, was a major step forward in the debate on liability law (Adams, 1985: 17ff.; Shavell, 1987: 5ff.; Landes Posner, 1987; Hofstetter, 1990a: 301ff.). Its perception of limited liability as a means of promoting efficiency has created a much more differentiated view of the pros and cons of parent liability (see Hofstetter, 1990a: 306ff.; 1994: Ch. III.3). Parent liability, in this view, appears as a function of three factors:
(a) the relative monitoring cost of alternative risk bearers;
(b) the diversification (relative risk attitudes) of alternative risk bearers; and
(c) the legal (opportunity) cost of developing and operating alternative liability regimes.
Starting from this, I have tried elsewhere to show that the limited liability of subsidiary corporations is above all an efficient solution to the problem of internalizing political risks (Hofstetter, 1994: Ch. III.3; 1990a: 309-12). Such risks are best borne by the political community at large, which is in the better position to prevent those risks from materializing. However, I developed three efficient exceptions to limited liability: first, when a parent abused a subsidiary; second, when the parent diluted or wasted a subsidiarys assets through its interference; and third, where the parent caused direct and separable harm to subsidiary creditors (Hofstetter, 1990a: 315-19).
The traditional remedy of piercing the corporate veil is, thus, a basically efficient instrument within the first category. German corporate group law for "qualified concerns" or French, Swiss and EC de facto director laws would belong in the second category. Tort law constructions, like vicarious liability, might have particular efficiency plausibility as instruments of the third category. More specifically, vicarious liability appears to be an efficient tool because it remedies the particular risk-externalization potential in connection with tort creditors of a subsidiary who are not able to renegotiate given corporate group structures or receive compensation for bearing limited liability risks. In addition, if combined with adequate prerequisites and defences for the parent (taking due account of the involvement of the parent at the subsidiary and its relative monitoring cost), the vehicle can be fine-tuned flexibly and restricted to specific situations. Finally, by not being a "broad sweep" rule and by basically covering insurable risks, vicarious liability does not undermine the limited liability of subsidiary corporations in a way that would render them useless as a means of hedging against political risks.
The US courts parent liability constructions under CERCLA would also seem to have efficiency potential within the above-mentioned third category (see Note, 1989: 986ff.). It is, of course, questionable from an efficiency point of view whether current owners of pieces of land who had no influence on pollution should be jointly and severally liable with the former owners and polluters of the land26. The political community at large would possibly be a more efficient bearer of such risks than the parent company of the current owner, no matter how tightly controlled the subsidiary was27. Where a parent company controlled a polluter-subsidiary and directed the latters hazardous waste release and environmental investments, however, a strong efficiency case can be made in favour of parent liability, the major reason being that the monitoring cost for preventing the pollution would basically be lower for the parent than for any part of the political community.
Beyond economic analysis of law
From a broader social perspective, the economic analysis of law appears reductivist and, hence, arguably inadequate (see Hofstetter, 1994: Ch. III.4). The debate about ecological liability in particular gives rise to doubts about the appropriateness of mere economic thinking, since the ecological problem is in itself an expression of the inability of the economic system to adequately "internalize the environment" (see Luhmann, 1989: 17ff.). In addition, the microeconomic equilibrium model commonly applied by the economic analysis of law has indisputable deficiencies from an epistemological point of view. Features such as methodological individualism, reducing human actors to strategies and opportunistic utility maximizers (see Hofstetter, 1994: Ch. III.4), or the perhaps inherent tendency of economic equilibrium analysis to undervalue the role of innovation and new knowledge (see von Hayek, 1989/91: vol. 1, 10ff., 72ff., vol. 2, 34ff., vol. 3, 97ff.; Ladeur, 1988: 332; Huber, 1988: 155ff.) are sufficient to warrant that the economic analysis of law should not monopolize any normative legal discussions. Let there be no doubt: economic models greatly help to integrate economic implications into legal thinking. In other words, the economic analysis of law is most appropriated vehicle for gearing law to the parameters of the economic system. And that is, if no more, at least an economically productive achievement. Yet, where the economic system is part of the problem itself, as is arguably the case with the ecological challenge, it seems at best doubtful whether this challenge can be overcome by merely resorting to economic rationales. The whole "internalization" debate prompted and led by economic thinking in connection with environmental law and liability (e.g. Wagner, 1989: 359ff.; Petitpierre-Sauvain, 1989: 466ff.) could, thus, lead into a vicious circle28. For a better ecological future more dramatic changes might be required than just the reallocation of cost within given (assumed) economic conditions.
Is there a superior theoretical framework for assessing and devising ecological liability law? The work of Teubner (1990a: 761ff.; 1991: 204ff.) for example, suggests that systems theory might serve as an appropriate reference. Indeed, systems theory seems particularly well suited to put the ecological problem into a broad evolutionary perspective. There, ecological liability appears as a problem of communication between the legal system and the economic system (Teubner, 1989: 123ff.) where the legal system has taken on the task of channelling signals from the environment into the economy ("risk dialogue"). "Sustainable development", not social efficiency, thus emerges as the ultimate paradigm (Becker, 1991: 19; see also Schmidheiny, 1992: passim). Does this make a practical difference? Assuming a natural incongruence between the legal and economic systems, communication between law and the economy is bound to fail if carried out in a one-way linear and hierarchical fashion. Accordingly, communication by legal rules has to be carried out on a two-way basis in that the legal system gears its messages to the parameters of the economy. Yet, this is where law and economics, at least operationally, comes in again. It might, for example, suggest guidelines for promoting the proper application of existing environmental technologies within a given framework of preferences and knowledge.
Stimulating the development of new attitudes, preferences and technology, however, transcends the borders of simple microeconomic equilibrium theory (see Ladeur, 1988: 312ff.). It leads communication between law and the economy to a more subtle level. Under these conditions, law can do no more than work its way intuitively towards the economy and private enterprise (Ladeur: 314ff.; Teubner, 1991: 205; 1989: 81ff.). The task is comparable to one of management trying to stimulate innovation such as in Japanese "knowledge-creating companies". These seem intentionally to refrain from efficiency-oriented, explicit or rational messages. Instead, they subscribe to tacit communication based on symbolism (cryptic (open), simple comments) and redundancy (overlapping messages on various levels and in various forms) (see Nonaka, 1991: 96, 101-102).
Thus, looking beyond simple economic analysis, an appropriate paradigm for ecological liability does have to integrate totally unknown variables into its calculus. It does this at the price of losing conceptual certainty, however. As a consequence, liability law itself becomes a "discovery procedure" (see Hofstetter, 1994: Ch. III.4.; Kramer, 1990: 252; von Hayek, 1969: 249ff.).
The revealed experimental character of liability law suggests openness and perhaps also caution (see Mastronardi, 1991: 449ff.). It calls, among other things, for a diversified approach to ecological problems involving a variety of (soft29) instruments within and outside liability law (see Brüggemeier, 1989: 230; Ladeur, 1988: 319ff.). "Decentralized" social arrangements based on such diverse measures as specifically tailored liability laws, liability caps, mandatory insurance, administrative controls, criminal law, ethical norms and education must thereby be given particular plausibility (see Schmid, 1990: 16ff.; see also the concept of the German Gesetz über die Umwelthaftung of 1991).
Parent liability law as a means to stimulate self-organization
The open paradigm of sustainable development underscores the autonomy of the economic system and its enterprises vis-à-vis environmental demands and liability law. It reveals the necessary lack of an overarching concept that would allow for an all-encompassing internalization of ecological interests into the economic process. What does this mean for parent liability law?
If the optimal ecological conduct by economic actors cannot be deduced from a general theory and translated into unequivocal legal commands, there still remains the possibility of getting closer to such conduct by way of trial and error (that is through falsification, Popper, 1984: 15). The corollary of this is an operational legal concept that is based on piecemeal (Popper, 1969: 157ff.) as opposed to utopian social engineering30. In other words, law assumes the more modest role of being "one player amongst many" instead of being the "paramount dictator" overlooking society. This leads, almost by necessity, to some form of labour-sharing between the legal system and private enterprise. Law and the economy observe each other and interact as equals.
This respect to law for the economy and its enterprises calls for legal self-restraint. The task of environmental legal norms is thereby limited to pushing private econoic activity in ecologically responsible directions. Laws job, in other words, is to inject environmental messages into the economic system hoping that they will build "structural couplings" (see Teubner, 1990b: 532ff.) towards some form of sustainable development. The targets of such legal norms are accordingly those economic facts (like costs, technology, attitudes or preferences) which can presumably be influenced by being tipped from the outside.
This approach respects the autonomy of economic actors, but it is by no means congruent with laissez faire. Law, including liability law, maintains a crucial function in communicating ecological pressures to private enterprise. Yet it does this primarily with the hope of triggering self-organization processes in the economic system31. It issues signal, the consequences of which it can neither foresee nor calculate. However, it can experiment with them, observe their impact upon the economic process, learn from that process and eliminate measures if they fail. An adequate ecological concept, therefore, employs liability law as a tentative measure to generate ecologically responsible self-organization (see Teubner, Chapter 2). Within this strategy, parent liability rules assume both offensive and defensive functions.
On the offensive side parent liability rules amplify the messages of liability law in general. By extending the liability of subsidiaries to parent corporations, parent liability can help stimulate those self-organizing processes that are already aimed at by general liability rules. In other words, parent liability can be a means of carrying environmental liability beyond the basic barriers of corporate law32. Yet parent liability can also be a stimulus to self-organization in its own right, that is, it can promote ecologically adequate corporate group structures. In this second variant parent liability does not aim at carrying out the specific purpose of tort-based environmental law, but those of corporate law as such33. It focuses on management control and the delegation of responsibilities as between a parent and its subsidiary - those issues which are usually considered to be within the domain of corporate law.
On the defensive side parent liability rules fulfil the task of limiting the liability exposure of corporate groups so as to motivate their continued investment and risk-taking. Parent liability rules, therefore, have a second function of protection autonomous parts within private firms against the potentially demotivating uncertainties of liability law. Their job is the establishment of a motivational setup that is conductive to further investment. Its time horizon should ideally be congruent with relevant planning and investment cycles (see Schmidheiny, 1992: 14). The defensive function of parent liability rules accordingly calls for only selective exception to the principle of limited liability.
Looking at the results of empirical research the notion of self-organization turns out to be a highly adequate macro-approach to liability law34. Eads and Reuter (1983: 72ff.) report that the effect of liability judgements on the conduct of firms is at best ambiguous. The same liability rulings can result in defensive, demotivated and, thus, counter-productive responses as much as they can lead firms to make the desired investments35. The outcome seems to hinge primarily on the existence of corporate cultures that are amenable to the productive tackling of the envisaged problems. Yet corporate culture, by its nature, cannot be finetuned by law, though it can be influenced by legal signals (see Schmidheiny, 1992: 49ff.). Empirical research, thus, seems to confirm that a very important aspect of liability law is its signalling effects, which influence corporate culture and thus promote self-organization processes within firms. It is this potential for "scratching enterprises just below the surface" and thereby triggering semi-voluntary changes of attitudes, preferences and motivations which ecological liability can above all hope to exploit36.
Against this background limited liability emerges in a new light. It is to be perceived as an instrument offering certainty within uncertainty to private enterprise, thereby motivating corporate actors to invest. If exceptions to limited liability remain conceptually narrow, the positive symbolic effects of limited liability may continue. This in turn supports well-tailored parent liability rules. Hence, a multilayered approach consisting of the principle of limited liability combined with various flexible, yet selective exceptions would seem to deserve preference over broadbased simple parent liability rules (Teubner, 1991: 207-11). It might make sense in this connection to differentiate between organizational risks of corporate group integration and production or transactional risks, whereby the former would be governed by corporate law concepts and the latter by various statutory or civil (common) law constructions37.
Additional practical conclusions may be drawn from the concept of self-organization. If there is reason to assume that enterprises are in a relatively better position than the legal system to generate new ecological production standards, then there exists an argument for shifting burdens of proof onto companies. This puts them in a position of having to work towards meeting their perceived standards of proof. Law, in turn, could observe this process and "free ride" on it, adjusting the standards of proof as the economys know-how and attitudes develop. As a consequence it seems important to keep such standards of proof open, in the sense of conceptualizing them procedurally. In other words, if the greater potential for generating sustainable production patterns lies in the economy, then legal standards of proof should focus above all on procedure and organization, and much less on the imposition of substantive targets38.
Thus, parent liability law could start with a presumption that a certain level of integration with the subsidiary renders the parent prima facie liable. But it would then be crucial that the parent be given the right to show the application of appropriate standards of group organization39. These standards could be set higher the more industry developed them. Such a process would allow the economy to feed its legally inspired creation of new ecological measures, techniques and attitudes back into the legal system. The legal system, in turn, could screen them and resignal them into the economy40. This self-perpetuating mechanism would seem to be one way of maintaining the differences between an autonomous economy and its dynamic environment at a sustainable minimum (see Luhmann, 1988: 324ff., 348).
There seem to be conceptual differences between the USA and Europe with regard to ecological liability of parent corporations. Traditional piercing the corporate veil is still paramount in US corporate law, but statutory constructions like those under CERCLA are partially making up for the lack of development in the corporate law field. The European "action", on the other hand, appears to be concentrated in the area of corporate and corporate group law, even though civil law constructions such as in tort law (vicarious liability of parent corporations) do have some appeal to the European legal community.
Holding these developments in the USA and Europe against the background of economic analysis of law, we find that there is much to be said in favour of both, as long as the respective liability regimes keep focusing on the relative risk bearing advantages of parent corporations (monitoring cost, diversification) and legal cost. An attempt to broaden the perspective beyond efficiency analysis into the realm of systems theory does not change this assessment dramatically. To be sure, it puts significant emphasis on evolution and openness and, thus, on the experimental character of liability law. As a consequence it favours a multilayered, flexible, diversified and basically cautious approach to ecological parent liability. It also supports the notion of self-organization calls for selective parent liability rules that might be based on shifting burdens of proof to the parent. It would however seem to be crucial that the parent be offered procedural defences allowing it to escape liability by showing the application of proper standards of group organization.
1 Walkovsky v. Carlton 223 NE 2d 6 (NY 1966); van Dorn Co. v. Future Chem. & Oil Corp. 753 F 2d 565 (7th Cir. 1985).
2 247 AD 144, 157, 287 NY 62, 76 (1st Dep.) affd., 272 NY 360, 6 NE 2d 56 (1936).
3 In Re Oil Spill by the "Amoco Cadiz" off the Coast of France on March 10, 1978 MDL Docket No. 3 76 ND III. 1984, American Maritime Cases, 2123-2199.
4 In Re Union Carbide Gas Plant Disaster at Bhopal, India, in December 1984 634 F Supp. 842 (SDNY 1986), affd & mod. 809 F 2d 195 (2nd Dist. 1987).
5 US v. Reilly Tar & Chemical Corp. 546 F Supp. 1100 (D Minn. 1982).
6 A third alternative is tort law (respondeat superior) (Blumberg, 1987: 621).
7 635 F Supp. 665, 670-672 (D Idaho 1986).
8 See the discussion of recent decisions in Hein, 1991: 637-639.
9 724 F Supp. 15 (DRI 1989). The judgment was confirmed by the Court of Appeals of the 1st Cir. (see Hein, 1991: 637, N. 10).
10 724 F Supp., 1991: 22-23.
11 724 F Supp., 1991: 25, n. 5.
12 901 F 2d 1550 (11th Cir. 1990), rehg gen. en banc, 911 F 2d 7421 (11th Cir. 1990), cert. den. 111 S Ct. 752 (1991).
13 861 F 2d 155 (7th Cir. 1988).
14 See also Joslyn MfG. Co. v. T. L. James & Co. 893 F 2d 80 (5th Cir. 1990), where the court denied the existence of a more liberal parent liability standard under CERCLA and espoused traditional piercing law.
15 675 F Supp. 22 (D Mass. 1987).
16 724 F Supp. 15 at 19 (DRI 1989).
17 94 NJ 473, 463 A 2d 893 (1983).
18 For details, see Hofstetter, 1990b: 578ff.
19 95 BGHZ 330 (1985).
20 107 BGHZ 7 (1989).
21 See also the "Video" decision, Betriebsberater 1991, 2173ff, and the recent "TBB" decision, Der Betreib 1991, 825ff.
22 This is being proposed under Swiss Law by Hofstetter, 1994: Ch. V.3.
23 If a parent has, for example, directed its subsidiary in its hiring decision or has directly addressed subsidiary employees by giving them work instructions, it automatically assumes joint and several liability to the subsidiarys employees as their "real employer" (see Hofstetter, 1990b: 586).
24 A bank has first refinanced a corporation and acquired its shares before it let it go bankrupt. The bank was declared liable for negligence (see Der Gesellschafter 1987: 46-50).
25 For details see Hofstetter, 1994: Ch. I.4.4.
26 There are defences available, for example the "third party defence" and the "innocent landowner defence", which are, however, difficult to meet (see Hein, 1991: 641).
27 If the implementation of CERCLA with its arguably inefficient, that is, too broad liability concept is itself to be seen as an opportunistic political act vis-à-vis private enterprise, limited liability could be perceived as a shield against that political risk too.
28 Analogous to the production-consumption spiral, in that the influence of the economic and social environment on attitudes and preferences of the economic actors are neglected.
29 That is not laws which could "kill" an enterprise, like exorbitant punitive damage awards: these could be "wrong" and, thus, irreversible like the "death penalty"; as a consequence, their effect could be overly demotivating and, hence, counterproductive in their effects on private enterprise.
30 Legal engineering along a mechanistic economic efficiency paradigm would arguably belong in this category (see Huber, 1988: 6, 231).
31 See the idea of legally guided self-regulation that is considered as one important road to ecologically responsible conduct by industry itself (Schmidheiny, 1992: 50ff.).
32 See the third efficient parent liability category described above.
33 See the first and second efficient parent liability categories described above.
34 On the operational (micro) level the economic analysis of law may keep its role as an instrument to deduce concrete, yet hypothetical, guidelines for optimal liability rules.
35 See also the critical report on the impact of modern US tort law in Huber, 1988: passim and 155ff.
36 It thereby takes advantage of its potential to influence the rationality of the enterprise, something conventional economic analysis of law would usually ignore; on the "bounded rationality" of organizations see Simon, 1981: 114ff.
37 See the attempt made in this direction by Teubner (1990b: 49ff). See also the Report by the Swiss Tort Law Commission of 1991 which suggests a division of strict liability for corporations into transactional liability (Gefährdungshaftung) and the organizational liability (Organisationshaftung). See also Brüggemeier, Chapter 4.
38 See in this connection the reference in Eads & Reuter, 1983: 79ff., to a study conducted by Weinstein et al.
39 See also Brüggemeiers concept of "proving faultlessness through documentation", Chapter 4.
40 See Blechers discussion of group wide quality systems in Chapter 11.
Adams, Michael (1985) Ökonomische Analyse der Gefährdungs - und Verschuldens-haftung. Heidelberg, R. von Deckers: C.F. Müller.
Anderson, Charles B. & Wethmar, Robert P. (1991) US-Umweltrecht: Der Oil Pollution Act 1990. Recht der Internationalen Wirtschaft, 1001.
Becker, Thomas A. (1991) Risikogesellschaft Schweiz? Politische Implikationen einer segmentierten Arbeitswelt. Neue Zürcher Zeitung, No. 186 (14 August) 19.
Bender, Bernd & Sparwasser, Reinhard (1990) Umweltrecht, Grundzüge des öffentlichen Umweltschutzrechts. Heidelberg: Recht und Wirtschaft.
Blumberg, Phillip (1987) The Law of Corporate Groups, Substantive Law. Boston/Toronto: Little Brown & Co.
Brüggemeier, Gert (1989) Umwelthaftungsrecht, Ein Beitrag zum Recht der Risikogesellschaft. Kritische Justiz, 22, 209-230.
Clark, Robert (1986) Corporate Law. Boston/Toronto: Little Brown & Co.
Eads, George & Reuter, Peter (1983) Designing Safer Products, Corporate Responses to Product Liability Law and Regulation. Santa Monica: Rand.
Hayek, Friedrich A. von (1969) Der Wettbewerb als Entdeckungsverfahren. In Hayek F.A. von. (Ed.) Freiburger Studien, Gesammelte Aufsätze. Tübingen: Mohr, 249.
Hayek, Friedrich A. von (1989/81) Recht, Gesetzebung und Freiheit. Landsberg a. Lech.
Hein, Peter (1991) Umweltrechtliche Aspekte gesellschaftsrechtlicher Transaktionen in den USA. Recht der Internationalen Wirtschaft, 636.
Hofstetter, Karl (1990a) Multinational Enterprise Parent Liability: Efficient Legal Regimes in a World Market Environment. N.C.J. Intl L. Com. Reg, 299.
Hofstetter, Karl (1990b) Parent Responsibility for Subsidiary Corporations: Evaluating European Trends. International and Comparative Law Quarterly, 576.
Hofstetter, Karl (1994) Sachgerechte Haftungsregeln für Multinationale Konzerne. Zur zivilrechtlichen Verantwortlichkeit von Muttergesellschaften im Kontext internationaler Märkte.(Forthcoming.)
Huber, Peter W. (1988) Liability. The Legal Revolution and Its Consequences. New York: Basic Books.
Kramer, Ernst A. (1990) Entwicklungstendenzen des Wirtschaftsrechts im ausgehenden Jahrhundert. Schweizerische Zeitschrift für Wirtschaftsrecht, 62, 249.
Krämer, Ludwig (1991) Umweltrecht der EG. Baden-Baden: Nomos.
Ladeur, Karl-Heinz (1988) Umweltrecht und technologische Innovation. Jahrbuch des Umwelt- und Technikrechts, 305-333.
Landes, William M. & Posner, Richard A. (1987) The Economic Structure of Tort Law. Cambridge, Mass: Harvard University Press.
Luhmann, Niklas (1988) Die Wirtschaft der Gesellschaft. Frankfurt/M.: Suhrkamp.
Luhmann, Niklas (1989) Ökologie und Kommunikation. In Criblez/Genon, (Eds). Ist Ökologie lehrbar. Bern: Zytglogge.
Mastronardi, Philippe (1991) Experimentelle Rechtsetzung im Bund. Zeitschrift für Schweizerisches Recht, NF 110 I, 449.
Nonaka, Ikujiro (1991) The Knowledge-Creating Company. Harvard Business Review, Nov/Dec, 96.
Note (1989) Liability of Parent Corporations For Hazardous Waste Cleanup and Damages. Harvard Law Review, 99.
Petitpierre-Sauvain, Anne (1989) Le principe pollueur-payeur. Zeitschrift für Schwei-zerisches Recht, NF 108, II/4, 433.
Popper, Karl R. (1969) The Open Society and its Enemies. London: Routledge/Kegan Paul.
Popper, Karl R. (1984) Logik der Forschung. Tubingen: Mohr.
Rehbinder, Eckard (1969) Konzernaussenrecht und allgemeines Privatrecht. Eine rechtsvergleichende Untersuchung nach deutschem und amenkanischem Recht. Bad Homburg: Gehlen.
Schmid, Gerhard (1990) Rechtsfragen bei Grossrisiken. Zeitschrift für Schweizerisches Recht, NF 109, II, 1.
Schmidheiny, Stephan (1992) Kurswechsel, Globale unternehmerische Perspektiven für Entwicklung und Umwelt. München: Beck.
Shavell, Steven (1987) Economic Analysis of Accident Law. Cambridge, Mass.: Harvard University Press.
Simon, Herbert A. (1981) Entscheidungsverhalten in Organisationen. Berlin: Moderne Industrie.
Tercier, Pierre (1990) Lindemnisation des préjudices causés par des catastrophes en droit suisse. Zeitschrift für Schweizerisches Recht, NF 109, II, 73.
Teubner, Gunther (1989) Recht als autopoietisches System, Frankfurt/M.: Suhrkamp.
Teubner, Gunther (1990a) Die Politik des Gesetzes im Recht der Konzernhaftung. In Festschrift Steindorff, 261.
Teubner, Gunther (1990b) Steuerung durch plurales Recht. Oder: Wie die Politik den normativen Mehrwert der Geldzirkulation abschöpft. In Verhandlungen de 25. Deutschen Soziologentages in Frankfurt a.M. Frankfurt/New York, p. 528.
Teubner, Gunther (1991) Unitas Multiplex - Das Konzernrecht in der neuen Dezentralität der Unternehmensgruppen. Zeitschrift für Unternehmens- und Gesellschaftsrecht,198.
Wagner, Beatrice (1989) Das Verursacherprinzip im Schweizerischen Umweltschutz-recht. Zeitschrift für Schweizerisches Recht, NF 108, II/3, 331.
Westbrook, Jay Lawrence (1985) Theories of Parent Company Liability and the Prospects for an International Settlement. Texas International Law Journal, 20, 321-33.