Reisman considers three major forms of international arbitration: in the International Court; under the auspices of the World Bank; and under the New York Convention of 1958. He discusses the unique structures of control in each situation as well as the stresses they have sustained. Drawing on extensive research and his own experience as a participant in the resolution of some of the disputes discussed, Reisman analyzes recent key decisions, including: Australia and New Zealand's attempt to stop France's nuclear testing in Muroroa; AMCO vs. Republic of Indonesia, concerning the construction of a large tourist hotel in Asia; and numerous others.
Reisman explores the implications of the breakdown of control systems and recommends methods of repair and reconstruction for each mode of arbitration. As a crucial perspective and an invaluable guide, this work will benefit both scholars and practitioners of international dispute resolution.
|Publisher:||Duke University Press|
|Sold by:||Barnes & Noble|
|File size:||539 KB|
About the Author
W. Michael Reisman, Wesley Newcomb Hohfeld Professor of Jurisprudence at Yale Law School, is the author of numerous books, including Folded Lies: Bribery, Crusades and Reforms and Nullity and Revision: The Review and Enforcement of International Judgments and Awards.
Read an Excerpt
Systems of Control in International Adjudication and Arbitration
Breakdown and Repair
By W. Michael Reisman
Duke University PressCopyright © 1992 Duke University Press
All rights reserved.
* * *
Arbitration is a delegated and restricted power to make certain types of decisions in certain prescribed ways. Any restricted delegation of power must have some system of control. Controls are techniques or mechanisms in engineered artifacts, whether physical or social, whose function is to ensure that an artifact works the way it was designed to work. In social and legal arrangements in which a limited power is delegated, control systems are essential; without them the putative restrictions disappear and the limited power may become absolute. The impulse to establish control systems, one might add, is animated neither by cynicism nor by fear of initiative. It rests on a healthy grasp of reality, indeed, one on which the United States constitutional experiment is based. Nor should controls be conceived in a negative sense. Controls are necessary not only for efficient operation. Effective controls are the only assurance of limited government. In this sense controls are a sine qua non of liberty.
In sociolegal arrangements, controls may be internalized or externalized or combinations of both techniques may be used. Internal legal systems of control have ranged from devices such as reliance on supernatural intervention (for example, by the use of oaths or rituals), on the inculcation and demonstrable application of craft skills, on collegiate decision structures in which tasks and roles are distributed among a number of participants, on insistence on manifest reasoning, and on peer pressure.
External controls have included such devices as checks and balances exercised by coequal or coordinate decision entities or on complex hierarchical arrangements, each successive level providing some supervision and potential review and rectification of the work of the preceding. Hierarchical configurations of control are animated, in part, by the desire to control the preceding levels. Controls effected by balances may dispense with hierarchies; they depend, for their effectiveness, on power parities. For several centuries the international political system used this mechanism of "balance of power" as its control device. Although the language of mechanics and hydraulics is often used in these discussions, all legal controls, one should remember, must be effected by people.
Some control systems are genetic, in the sense that they are designed as part of the original conception of the process to whose operation they contribute. Others are added on, as experience reveals that the process is encountering factors or obstacles or operating in environments which had not been considered at the time of its creation. Hence new control arrangements—new fuses, new governors, another set of emergency brakes, an additional comptroller—are grafted on. The biological evolution of life on our planet has followed this trial-and-error approach. In social and legal arrangements, efforts are always made to plan controls, but subsequent corrections or additions are often required. In our constitutional scheme, for example, some control systems are built into the genetic constitution. Others, such as judicial review, were added on, by the initiative of some participants, when their appraisal persuaded them that additional controls were required. Still others are fashioned after calamity generates the belief that recurrences can be avoided by additional controls.
When physical control systems break down—the brakes fail, the thermostat cuts out, or the governor pops—the system may simply stop functioning, go out of control, or destruct. When social and legal control arrangements break down, the decision process does not necessarily fail. But it certainly changes, as more power shifts to the now comparatively less-controlled decision-maker. This is particularly the case in arbitration. With controls it remains a delegated and restricted power. Without controls it becomes absolute. Thereafter, the arbitrator, like the Roman emperor, may be tempted to say "quod voluit arbiter habet vigor em legis": "whatever the arbitrator wants is the law." Hence the linkage between controls, limited power, and liberty.
Controls are a virtual prerequisite for international third-party decision. When we deal with optional international decision processes, that is those in which, for the most part, participation is voluntary and in which the expectation of the operation of the control mechanism may be an important factor in decisions opting to use a particular process, an actual or anticipated control breakdown such as this is likely to induce many actors henceforth to refrain from using the process. Thus controls in international adjudication and arbitration are not simply conditions of efficient operation. They are conditions of operation. Much as lawyers cannot practice law without clients, international tribunals cannot decide disputes without litigants. Litigants come on an entirely voluntary basis and have no reason to come to an uncontrolled process.
Yet breakdowns in legal controls in general are not necessarily calamitous. Depending on the values of the appraiser, some breakdowns may even be viewed as positive. If oligarchical controls break down, for example, those who have been freed from them may characterize the change as democratizing and liberating. When the breakdown concerns controls designed to restrain capital from creating concentrations which might reduce competition, the wealth elites who have been released may, similarly, characterize the change as liberating.
In international law one of the traditional controls on the exercise of power by states over investments there by nationals of other states was the insistence that applications of local law by the host state be subject to international review under a variety of criteria. When many of the developing states, in a movement called the New International Economic Order, resolved that, henceforth, they could expropriate without international controls and control themselves, as it were, in their own courts, they characterized it as a liberation. But the very foreign investors who were necessary both for their development and for the application of their domestic jurisdiction over them began to withdraw. When control systems break down and the result is a reduction in the overall effectiveness of the decision process of which they are a part, for example, by a reduction in participation, the breakdown must be characterized as a pathology. The point bears emphasis. While controls, from the perspective of some actors, may be viewed as unpleasant, unfair, or even intolerable restraints, the ultimate evaluation of their utility, if not indispensability, is their contribution to the effective operation of the system of which they are a part.
Because control systems affect the allocation of costs and benefits of the processes in which they operate, they are always highly political, often controversial, and frequently the subject of explicit policy consideration. When control systems cause changes which appear to have negative effects on the system to whose operation they are supposed to contribute, it becomes appropriate, if not urgent, to inquire about better alternatives.
Judicial Control Systems
In some domestic adjudication, a hierarchical judicial bureaucracy operates as the control system for state-sponsored dispute resolution. Successive levels of the bureaucracy appraise and reappraise the general workings of the system as well as the accuracy, consistency, and justice (the calculus of rectitude of the system and its environing culture) of particular applications, making adjustments where they seem appropriate. These levels are comprised of a permanent staff which is recruited to the bureaucracy, inculcated with organizational values and itself subjected to effective policing and disciplinary mechanisms. A distinct legislative bureaucracy provides an additional, contingent system of control, which can intervene when its own appraisals indicate that the performance of the judicial bureaucracy, for some of the tasks it has been assigned, is unsatisfactory.
"An ounce of prevention is worth a pound of cure." In terms of economy, the preferred control system is always internalized and prospective rather than externalized and retrospective. Judicial control is better accomplished, for example, within the court system than by the legislature, and better accomplished by the legislature than in extralegislative agitation. It is better that the system anticipate and obviate problems rather than repair them and whatever peripheral damage they will have caused after they erupt. Where control is vital for mission accomplishment and/or system survival and it is probable, following "Murphy's law," that some initial efforts at control will, as a matter of statistical probability, not succeed, back-up control systems are devised. Hence the common phenomenon of redundancy, multiple, sequential control levels, found in nature and human artifacts, is not necessarily a product of faulty planning or oversight.
All control systems involve costs. Judicial control systems have costs in terms of the funds needed to establish and maintain them and in terms of the time which must be expended in the successive levels of appeal before a dispute is finally resolved. It is arguable that the more hierarchical layers of protection a bureaucracy has, the more likely it is to get things right. But here, as elsewhere, there is no free lunch. The expenses of establishing and maintaining the vast control superstructure are "passed through" to users, raising the costs of justice, in terms of money and time, and, for some marginal actors, pricing them out of the very system which is supposed to offer them the opportunity to protect their rights. At the same time, other actors, calibrating the rising nuisance value provided by the sequential appeal options, may be encouraged to start dubious actions with the not implausible expectation that settlement, at an early stage, will appear more cost-effective to those they are harassing than a perforce lengthy and expensive vindication of their legal rights.
Added layers of control also increase time costs: the length of the interval between claim and disposition. The old adage "justice delayed is justice denied" is often true in the sense that a delayed victory may deprive the winner of substantial economic value. The costs generated by control systems are imposed on the parties who must pay, at many levels, to defend their positions and whose treasure may be immobilized pending final decision and on the community which must finance the control system and may, itself, be deprived, pending the final decision, of many of the benefits of the values frozen in dispute.
Considerations such as these generate a tension between two control system policies: justice and finality. The old Roman maxim said interesse rei publicae ut sit finis litium, "the public interest requires that there be an end to disputes," which imports a clear cutoff point, an arbitrary "enough-is-enough" point. Abraham Lincoln's statement that "nothing is final until it's right" rejects any arbitrary cutoff point. That statement may have been based as much on political realism as on moral conviction; for, right or wrong, diffused popular indignation can acquire a common vector and become political power. New arrangements to regain social stability may then be required. The interest in finality means that an arbitrary limit for control systems must be established. The interest in justice means simply that justice must be done, no matter what the cost nor how long it takes. Control systems perforce strike compromises between these interests. The compromises themselves will often reflect larger political configurations.
There is no such thing as an "ideal" control system, for each system is, like all legislation, an experiment in social control in which legal ingenuity and sometimes creativity try to adapt or invent institutional arrangements within the limits of available resources and political tolerances and in response to the special features of the context in which it operates. Because control systems are creatures of their contexts, it is no surprise that there are such striking differences between national and international control arrangements. In brief national systems are, for the most part, hierarchical and multilevel, while international control systems are, for the most part, non-hierarchical systemic adaptations.
Control Systems in International Arbitration
For a new and rapidly growing category of events which we often call, for lack of a better term, "transnational," international commercial arbitration, broadly understood, now performs many of the functions of domestic adjudication. But international arbitration lacks a set of bureaucratic institutions comparable to its domestic counterparts which might perform its control functions, and the international political environment has not been conducive to their creation. Hence international arbitration has approached the control problem in an entirely different way.
In international law the basic theory of arbitration is simple and rather elegant. Arbitral jurisdiction is entirely consensual. As in Roman law and the systems influenced by it, arbitration is a creature of contract. The arbitrator's powers are derived from the contract. Hence an arbitrator is not entitled to do anything not authorized by the parties: in its classical formula in the Digest of Justinian, arbiter nihil extra compromissum facere potest. An arbitral award rendered within the framework of the common agreement of the parties is itself part of the contract and hence binding on them. But an award which is produced in ways inconsistent with the shared contractual expectations of the parties is something to which they had not agreed. The arbitrator has exceeded his power or committed what French law and international doctrine calls an excès de pouvoir. If an allegation of nullity can be sustained, the putative award is null and may be ignored by the "losing" party.
A theory like this is quite advantageous to parties, for it gives them an additional contractual option for resolving disputes without engaging the community's courts. It is also advantageous to the community. Private disputes are essentially diversions from productive activity. An arbitral control mechanism facilitates their economical resolution without more general disruption and without direct cost to the community. The doctrine of excès de pouvoir is supposed to function as a control mechanism in this theory. Without it, whatever an arbitrator did, no matter how inconsistent it might have been with his instructions, would have produced a binding award. The arbitrator would become an absolute decision-maker and arbitration would lose its character of restrictive delegation. Excès de pouvoir is the conceptual foundation of control for arbitration. It is not, however, without certain problems.
This kind of control mechanism can work well in an organized political-legal system, with a hierarchical control system equipped with an effective compulsory jurisdiction to review allegations of excessive jurisdiction and to decide impartially the alleged nullity of the award. But it is susceptible to abuse in a system, like the international one, in which there is no such permanent and effective hierarchical structure. In the absence of a reviewing authority, a party alleging that an arbitrator did something not authorized by the agreement to arbitrate is simultaneously prosecutor, judge, and jury in sua causa. The potential for abuse is obvious. Ironically, the very theory of nullity which serves, in domestic contexts, to police the arbitrator and thus encourages arbitration tends, in the international setting, to undermine it. But the alternative to the risk of unilateral abuse, no control, is also unacceptable.
As long as arbitration was infrequently used and parties could discount losing beforehand, this system, while hardly optimal, worked. But as modern transnational arbitration increased as a function of the expansion of transnational activity, the inadequacy of this control mechanism in this new context became more apparent. Arbitration was in danger of being undermined.
The apparent solutions were unsatisfactory. Efforts to make awards binding without regard to their possible excès de pouvoir—in effect, abandoning all control systems—may well have discouraged resort to arbitration. Legal decision without control runs the danger of reducing predictability, and rational actors are unlikely to submit matters that are important to them to a voluntary dispute system that ranges from uncertain to capricious. Efforts to allow claims of nullity and to try to control them by making them precise have been countered by fears of abuse, for it has been assumed (and not unreasonably) that the more precise the grounds for nullifying, the more claims for nullification there will actually be. A leitmotiv of modern international arbitration has been the search for a way of breaking free of the unsatisfactory consequences of each of these alternatives by devising some sort of new institutional device that could provide responsible and predictable control, while minimizing the potential for abuse of claims of nullity.
The possibilities of solving this problem were quite limited in the absence of some sort of permanent or "standing" international institution or set of institutions. If such institutions could be created, even if they were not necessarily hierarchical, they would at least be there when the need arose. They could be authorized, preferably in advance of the dispute, to decide a claim that an international award was null for some excès de pouvoir. An unratified treaty between the United States and the United Kingdom signed on January 11, 1897, indicates the limits of a control scheme which depended, for its effectuation, on an institution that did not yet exist: if members of the tribunal were not unanimous with regard to claims exceeding £100,000, "either party may demand a review.... In such a case the matter in controversy shall be submitted to an Arbitral Tribunal, consisting of five jurists of repute." But the establishment of this second tribunal required the agreement and cooperation of the parties, precisely the ingredient that would be absent at the stage in which it was most necessary.
Excerpted from Systems of Control in International Adjudication and Arbitration by W. Michael Reisman. Copyright © 1992 Duke University Press. Excerpted by permission of Duke University Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents
2 The International Court: The Atrophy of Informal Control Mechanisms,
3 The ICSID Experiment: The Breakdown of International Institutional Control Mechanisms,
4 Non-Institutional International Commercial Arbitration: The Breakdown of the Control System of the New York Convention,
5 The System and Self: Changing, Repairing, and Reconstructing,
Table of Cases: Arbitration,