On the Philosophy of Precedent: Proceedings of the 24th World Congress of the International Association for Philosophy of Law and Social Philosophy, ... (Archiv Fur Rechts: Und Sozialphilosophie) 3515101500, 9783515101509

Citation preview

On the Philosophy of Precedent Proceedings of the 24th World Congress of the International Association for Philosophy of Law and Social Philosophy, Beijing, 2009 Volume III Edited by Thomas Bustamante and Carlos Bernal Pulido

ARSP Beiheft 133 Franz Steiner Verlag

Archiv für Rechts- und Sozialphilosophie

On the Philosophy of Precedent Edited by Thomas Bustamante and Carlos Bernal Pulido

archiv für rechts- und sozialphilosophie archives for philosophy of law and social philosophy archives de philosophie du droit et de philosophie sociale archivo de filosofía jurídica y social Herausgegeben von der Internationalen Vereinigung für Rechts- und Sozialphilosophie (IVR) Redaktion: Dr. Annette Brockmöller, LL. M. Beiheft 133

On the Philosophy of Precedent Proceedings of the 24th World Congress of the International Association for Philosophy of Law and Social Philosophy, Beijing, 2009 Volume III Edited by Thomas Bustamante and Carlos Bernal Pulido

Franz Steiner Verlag

Bibliografische Information der Deutschen Nationalbibliothek: Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über abrufbar. Dieses Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist unzulässig und strafbar. © Franz Steiner Verlag, Stuttgart 2012 Druck: Druckhaus Nomos, Sinzheim Gedruckt auf säurefreiem, alterungsbeständigem Papier. Printed in Germany. Franz Steiner Verlag: ISBN 978-3-515-10150-9 Nomos Verlag: ISBN 978-3-8329-7625-5

TABLE

OF

CONTENTS

Thomas Bustamante / Carlos Bernal Pulido Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7

I ON THE STRUCTURE AND THE JUSTIFICATION OF PRECEDENTS Pierluigi Chiassoni The Philosophy of Precedent: Conceptual Analysis and Rational Reconstruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Marina Gascón Rationality and (Self) Precedent: Brief Considerations Concerning the Grounding and Implications of the Rule of Self Precedent . . . . . . . . . . . . . 35 Carlos Bernal Pulido Precedents and Balancing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Thomas Bustamante Finding Analogies Between Cases: On Robert Alexy's Third Basic Operation in the Application of Law . . . . . . . . 59 II ON THE STRENGTH AND THE USES OF PRECEDENTS Larry Alexander Precedential Constraint, Its Scope and Strength: A Brief Survey of the Possibilities and Their Merits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Patrícia Perrone Campos Mello The Role of Precedents as a Filter for Argumentation . . . . . . . . . . . . . . . . . . . . 83 III PRECEDENTS IN CONTEMPORARY LEGAL CULTURES Victoria Iturralde Precedent as subject of interpretation (a civil law perspective) . . . . . . . . . . . . . 105 Zhang Qi On the Method of Searching for Guiding Cases on the Basis of Trial Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Ewoud Hondius Precedent Revisited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

THOMAS BUSTAMANTE / CARLOS BERNAL PULIDO INTRODUCTION In this volume, which contains the 3rd issue of the Proceedings of the 24th IVR World Congress, held in Beijing in the year of 2009, the reader will find a selection of papers presented at that International Congress on the general theme of “The Philosophy of Precedent.” The foundations, legal nature, structure, strength and uses of the case law were vividly discussed at two Special Workshops especially dedicated to the study of legal precedent, where the papers collected in this issue were debated by a large number of scholars from around the world. The attention that the theme received at the Congress indicates a trend of growing interest on the topic of precedent among legal philosophers, legal theorists and practitioners from all of the legal families and traditions, including those of the socalled Civil Law Systems. Taking into consideration this movement towards the universality of precedent-based reasoning in legal discourses, the essays comprised in this volume attempt to provide an account of legal precedent which explains in a reasonable way the connections between theoretical issues on the nature of precedent and practical queries over its foundations, structure, strength and uses in contemporary legal systems. In effect, under the influence of authors such as Dworkin, Alexy, MacCormick, Marmor, Waldron and many others, a large part of contemporary legal theory is dedicated to some sort of ‘normative jurisprudence’. One of the main features of such type of legal theory is that it is particularly concerned with the justification of legal decisions and with the rationality of legal reasoning. In this context, it becomes crucial to determine the contents and the argumentative uses of a jurisprudential concept like that of precedent. In fact, it can be argued that some basic legal principles like ‘certainty’, ‘coherence’, ‘fairness’ and ‘impartiality’ are endangered if the law is not applied with some sort of adherence to judicial precedent. The relationship between legal philosophy and legal precedent is one of a dual character. On the one pole, to apply a precedent to a novel case we need a reasoning to justify the connection or similarity between the two cases, and this reasoning can be characterized at least in part as a philosophical argument; on the other pole, there can be no rational coordination in a legal system if its officials or the bodies in charge of its application do not follow their own precedents to a certain extent. The former pole shows us that one cannot avoid theoretical/philosophical arguments, both of analytical and normative nature, if one is to apply precedents in a rational way. That is to say, there is no clear distinction between theoretical or philosophical arguments on the one hand and strictly legal arguments on the other. When judges refer to a precedent, they are inevitably asked to adduce practical reasons in support of the application of the precedent to the novel case. There are multiple ways in which this can be done. First, whenever a judge accepts a precedent which is not binding she must commit herself to the reasons used by the previous court to support the decision that she is quoting: “Taking an affirmative position on a validity claim generates this illocutionary obligation which links reasons to

8

Thomas Bustamante / Carlos Bernal Pulido

motives.”1 Second, any application of a legal precedent needs to be supported by an analogy between cases which requires a complex theoretical discourse about the similarities and the differences between cases. And thirdly, to distinguish and compare precedents judges have the burden of providing reasons to carve exceptions in legal rules or to re-classify the facts of the case in order to leave them out of the scope of a precedent prima facie applicable to the case. In all these situations, jurists need practical reasons in order to justify their arguments, claims and decisions. The latter pole, in turn, shows that the obligation to consider precedents (which does not necessarily imply a strict obligation to follow every precedent) is a condition of rationality for any legal system. As Neil MacCormick and Robert Summers put it very clearly, to follow precedents is a requirement of practical human reasoning,2 for there can be no rational justification of a social practice without universal application of the same rules. It is therefore of vital importance to elucidate the philosophical problems generated by the technique of precedent. Jurists need both a method for interpreting and applying legal precedents and a theoretical apparatus to enable them to determine the force and the status of legal precedent in ordinary legal argumentation. What once was regarded as a doctrine valid only within the boundaries of common law is nowadays correctly perceived as a universal problem of legal philosophy and of its most practical branch: the normative theories of legal argumentation. With these considerations in mind, the studies compiled in this volume will outline some of the most important aspects of a philosophical theory of precedent. These essays will explore and explain issues such as the structure of legal precedents, their philosophical foundations, their relevance for legal theory and practice and the methodological problems that jurists are likely to face when recognizing, interpreting and following them. The essays are divided into three sections, being the first on the structure and the foundations of legal precedents; the second on their strength and the practical uses in legal discourse; and the third on the practice of precedents in contemporary legal cultures. The first section, “On the Structure and the Justification of Precedents,” begins with Pierluigi Chiassoni’s paper, which undertakes a rigorous structural analysis of precedents with a view to reconstruct the core concepts related to precedent-identification and application, which have to do with the definition of the ratio decidendi, the interpretation of precedents and their practical relevance. The paper is followed by Marina Gascón’s essay on the notion of self-precedents and the rationality of the legal system. While the traditional approaches to legal precedent consider them as valid due to the authoritative element comprised in the law-making power of the courts, Gascón expounds another element that is equally important to understand the foundations of precedent, which refers to the Kantian principle of universalizability, which is at stake even when there is no authoritative rule determining the obligation to follow precedents, as we can see in the case of self-precedents. Carlos Bernal Pulido and Thomas Bustamante’s papers, in turn, are concerned more spe1 2

Klaus Günther, ‘Communicative Freedom, Communicative Power and Jurisgenesis’ (1996) 17 Cardozo Law Review 1035, at 1041. Neil MacCormick and Robert Summers, ‘Introduction’ in MacCormick; Summers (eds) Interpreting Precedents (Dartmouth, Aldershot 1997), 4.

Introduction

9

cifically with the structure of precedents and of the arguments based on them in the discourses of justification of legal decisions. Bernal’s paper deals with the connection between balancing and precedents in adjudication, particularly in civil law jurisdictions. He observes that in civil law jurisdictions balancing has become one of the most important methods for constitutional reasoning, since the biggest part of the constitutional norms has the structure of principles. None the less, constitutional courts tend to import from the common law the methods of precedent-based legal reasoning, which start from a ratio decidendi that has the typical structure of a rule. This is, according to the author, the “precedent-balancing paradox”, which can be stated thus: “While the doctrine of precedent requires the application of rules, balancing is the way to apply principles.” Yet, by applying Alexy’s theory of fundamental legal rights to solve this dilemma, Bernal shows us not only that it is possible to dismantle this paradox, but also that balancing is required to make reasoning with precedents a rational form of legal argument. Bustamante’s paper, in turn, starts with Robert Alexy’s description of Analogy or Comparison of Cases as a “basic operation in the application of law”, in order to show the connections between rules, principles, and analogical application of the rules derived from legal precedents. The main points of the paper are, first, to defend the claim that analogy needs to be grounded on a balancing of the principles which stand behind the rules embedded in the decision taken as a paradigm, and, second, to show that the better way to explain this connection is to understand analogy not as a “basic operation in the application of law”, as Alexy does in his recent works, but rather as a judicial development of the law on the basis of balancing. The second section, on “The Strength and the Uses of Precedents” is less concentrated on analytical issues and more concerned with the pragmatic aspects of precedent-based reasoning. The opening paper, by Larry Alexander, examines how broadly precedents constrain (their scope) and how strongly they do so (their strength). The analysis is mainly directed towards the issue of overruling precedents and the reasons that are given for the fidelity to precedent or for the departure from it, and leads to the conclusion that overruling precedents turns out to be only an instance of the more general problem of the rationality of rule following. The paper is followed by Patricia Perrone Campos Mello’s essay, which concentrates on the roles and the limits of a theory of precedents, which is depicted as connected to a set of values such as legal certainty, equality, legitimacy and efficiency in the courts system. The key function of precedents, as the author argues, is to “serve as a filter for legal argumentation, guiding litigants and judges on issues to be discussed and considered in the decision of the case.” The third section, at last, on “Precedents in Contemporary Legal Cultures”, attempts to understand how the notion of precedent performs its functions in some of the different legal traditions. The opening paper, by Victoria Iturralde, focuses on civil law systems, where legal theorists still tend to assume that legislation is the “only and exclusive” source of law. Against this expectation, however, precedent may work either as a formal source or a material source of law in these systems, and they do play an active role in the operation of the legal system. Hence, according to the paper, they need a more developed theoretical account on how one is to interpret them in the application of law, which is the object of the inquiry. Zhang Qi’s paper, in turn, focuses on the practice of precedent in contemporary China, whose

10

Thomas Bustamante / Carlos Bernal Pulido

court system was modified quite recently, in 2005, to establish a more strict approach to legal precedent which rules that some “guiding cases” are to be authoritative. The paper intends to answer, therefore, the questions of how guiding cases are to be found and how one is to evaluate the similarity between cases in the Chinese legal practice. Finally, Ewoud Hondius presents his conclusions on a round-table which he organized in 2006, at the Congress of the International Academy of Comparative Law, as well as some new developments on his empirical research on the different approaches to precedent that are found in contemporary legal cultures. His conclusion, in short, is that there are still conflicting tendencies in civil law and common law approaches to legal precedent. The editors would like to thank the authors of the chapters of this volume for the their relevant contributions, as well as Prof. Zenon Bankowski, the IVR Committee and the dedicated members of the Chinese Law Society, who worked very hard to help them select the papers comprised in this volume amongst hundreds of papers received for the publication of the proceedings, and who trusted them the job of editing this volume. But, above all, they wish to thank Lara Pratt, who generously agreed to act as a reviser of the final proofs of the volume.

I ON

THE

STRUCTURE

AND THE JUSTIFICATION OF

PRECEDENTS

PIERLUIGI CHIASSONI THE PHILOSOPHY OF PRECEDENT: CONCEPTUAL ANALYSIS RATIONAL RECONSTRUCTION A. PHILOSOPHY

AND

OF PRECEDENT: TWO BASIC VARIETIES

Organizers kindly directed to submit papers in “the philosophy of precedent”, concerning: (a) “the structure of legal precedents”, (b) their “philosophical foundations”, (c) their “relevance for legal theory”, (d) their “relevance for legal practice”, and, last but not least, (e) the “methodological problems” jurists and judges must cope with in order to “recognize”, “interpret”, and “follow” precedents. A philosophy of precedent – roughly, any critical inquiry upon the afore-mentioned subjects – may be of two, quite different, kinds. To begin with, it may be conceived as a normative theory about judicial precedent: as a piece of normative jurisprudence or, in a wider sense, of the old “science of legislation” of Enlightenment philosophers. A philosophy of precedent of this kind may be characterized as a discourse, regarding a given legal order or set of legal orders, that purports to provide, and usually provides, answers to legal policy questions concerning the axiologically proper ways of (i) understanding precedents, (ii) defining their practical relevance within a legal order (i.e., their status as a source of positive law, if any), (iii) defining the principles that ought to make up the doctrine of precedent, and (iv) dealing with precedents from the methodological standpoints of interpretation, use, and reasoning. The axiological adequacy of any such theory clearly depends, in turn and ultimately, on some background ethical view: usually, on some principles of public morality, concerning the legal institutions of a wellordered society, which the philosopher is committed to. Secondly, and contrariwise, a philosophy of precedent may also be conceived as an analytical theory: a piece of conceptual analysis, the core business of which is dealing with the terminology, the concepts, and, at a meta-theoretical level, the “theories” concerning judicial precedent as regards to a certain legal culture, in order to get to a better terminology, better concepts, and better ways of thinking, from the standpoint of rationality-values like clarity, precision, discrimination, and justification. There are obvious connections between the philosophies of precedent of the two varieties above. On the one hand, a normative philosophy of precedent is doomed to failure – to be dwelling in the rhetorical world of hazy discourses – unless it is enlightened, supported, and, if necessary, cured, by an equipment of clear and distinct concepts, views, and modes of thought. On the other hand, an analytical philosophy of precedent, to do its job properly, needs full consciousness about the fact that judicial precedent is a matter for ideological, value-disputes and allegiances. The present paper purports to outline a few, tiny, pieces of a philosophy of precedent in the analytical mood. Taking up the standpoint of a piecemeal, explanatory, axiologically uncommitted, analysis, I will deal in turn with three issues:

14

Pierluigi Chiassoni

Firstly, an analytically proper definition of the key notions of ratio decidendi and obiter dictum; Secondly, an analytically proper theory of the interpretation of judicial precedents (really, an explanatory meta-theory, as we shall see); Thirdly, and finally, an analytically proper theory of the practical relevance of judicial precedents. B. A

FEW PUZZLES

Before proceeding along the lines above, however, let me say a few words about two subjects concerning judicial precedents, which were mentioned in the organizers’ list at the outset, and I am apparently discarding in my present inquiry: to wit, “the structure of precedents” and their “philosophical foundations”. 1. THE “STRUCTURE” OF PRECEDENTS What is “the structure of precedents”? Such a question, though it may seem a matter of course, is nonetheless apt to be misguiding. More cautious questions come to the forefront. Do precedents have “a structure”? Is there really something that is “the structure” of precedents? What are we looking for, when we ask for “the structure of precedents”? Clearly, we cannot provide any reasonable answer to these questions, unless we have previously established and made clear what exactly do we mean by the slippery phrase “judicial precedent”. No fewer than three different, though related, candidates show up here. To begin with, a judicial precedent may be considered to consist in a judicial decision (i) as a whole – the opinion, including the ratio decidendi or holding, plus the individual determinations for the individual case at hand – , (ii) pronounced in a previous time t-1, (iii) usefully reported (i.e., reported in such a way as to be liable to be known and used in the future by lawyers and judges working in the concerned jurisdiction), and (iv) bearing on the same, or like, (kind of) facts and questions as the facts and questions to be adjudicated at present time t. This I will call precedentjudgment. Further, a judicial precedent may be considered to consist, more narrowly, in the sole ratio decidendi, or holding, of a judgment pronounced in a previous time t-1, usefully reported, and bearing on the same, or like, (kind of) facts and questions as the facts and questions to be adjudicated at present time t. This I will call precedentholding, or precedent-ratio decidendi. Finally, a judicial precedent may be considered to consist, again in a narrow way, in the individual determinations of a judgment pronounced in a previous time t-1, usefully reported, and bearing on the same, or like, (kind of) facts and questions as the facts and questions to be adjudicated at present time t. This I will call precedent-order. “Structure” is a slippery word too, though. The structure of something may be taken to be, in turn:

The Philosophy of Precedent

15

(a) that part of something, without which it simply could not stand (survive, subsist, exist) – think at the structure of a building, its master-walls, its cage of concrete-pillars; (b) the (axiological) core or kernel of something, its essential part, i.e., the feature or property without which it would be worthless from some relevant standpoint; (c) the set of proper relationships, the net of proper connections, between the several parts of something, without which it would be a poor, defective, instance of its kind. Clearly, “structure” is a value-laden word, provided with a (positive) emotive meaning: indeed, saying of some x that is, belongs, or pertains to, the structure of some y usually casts on x a favourable light, it brings to x the consideration due to things that are, for some reason, important. Apparently, we may usefully investigate about the structure of a precedent-judgment, in all of the three meanings above. No piece of judicial discourse would stand as a (duly-pronounced, well-built) precedent-judgment if, provided a normative model of an opinion-plus-order structure obtains, it only contains either the opinion, but no individual order whatsoever, or an individual order, but no opinion whatsoever. No precedent-judgment would be a worthwhile (valid, proper) instance of the kind, if it does not contain any holding, assuming the holding to be its essential part, its axiological core or kernel (and the same may be said concerning the order). Finally, no precedent-judgment would be a worthwhile (valid, proper) instance of the kind if, provided a normative model of an opinion-plus-order structure obtains, it does contain a holding and an individual order, but they are either totally unrelated, or conflicting, among each other. The same cannot be said regarding the structure either of a precedent-holding, or of a precedent-order. What may we be about, when we make an inquiry on the “structure” of a precedent-holding or the “structure” of a ratio decidendi? Assuming the ratio decidendi to be a sentence in a natural language, its structure would be its syntactical arrangement. This, however, does not seem a worthwhile subject of inquiry – but for the grammatical form/logical form issue, of course. So perhaps, by way of a charitable interpretation, when somebody sets to an inquiry about the “structure of precedents”, and by “precedent” she means the precedentholding, maybe what she purports to inquire upon is the proper notion of a holding, the proper notion of a ratio decidendi. If that is the case, however, talking in terms of the “structure of precedents” is a misleading, obscure, form of speech; it would be better giving it up altogether, and talking instead either of the structure of precedent-judgments, or of the (proper) concept of a holding or a ratio decidendi. Obviously, what is the proper concept of a holding/ratio decidendi depends on the inquirer’s standards of conceptual adequacy. As I said before, from the standpoint of an analytical philosophy of precedent, these standards have to do with the rationality-values of clarity, precision, discrimination, etc.

16

Pierluigi Chiassoni

The first of the three issues I will deal with is devoted precisely to this point (see § 3 below). So, in a sense, I did not really discard the first issue proposed by the organizers; though I had to make clear why. 2. THE “PHILOSOPHICAL FOUNDATIONS” OF PRECEDENTS Like “structure”, “foundations” is another beloved word in the philosophers’ lexicon. Indeed it sounds, at once, technical and ripe with precious contents. Like “structure”, however, it is saddled with indeterminacy: perhaps, even more so than the former. Do precedents have “foundations”? Is there really something working as “the philosophical foundations” of precedents? What are we looking for, when we ask for “the philosophical foundations” of precedents? Two possibilities may be singled out, to mention just these two, out of many – which, by the way, point to an even further, broader, way of understanding and using the phrase “judicial precedents”. (1) To begin with, the search for the philosophical foundations of precedents may belong to normative, practical, philosophy. In such a case, it is a search for justification. Obviously, from this standpoint, justification does not concern particular judicial precedents (-judgments, -holdings, -orders) as such. Rather, it concerns some real or imaginary system of precedent; for instance, it may concern: (a) some, contextually described, actual, practice of (establishing, using, interpreting, following, applying, modifying, expanding, disregarding, overruling, etc.) judicial precedents, in a certain, actual, legal order; or, (b) some, contextually described, ideal doctrine (about the establishing, using, interpreting, following, applying, modifying, expanding, disregarding, overruling, etc.) of judicial precedents, for a certain, real or imaginary, legal order. “Foundations”, like its sibling word “grounds”, is clearly metaphorical; “justification” isn’t. The latter should replace the former in careful philosophical discourse. (2) The search for the philosophical foundations of precedents may also belong to analytical philosophy. In such a case, it would be a search for a better understanding of the institution, doctrines, and practices of judicial precedent, as they are in some legal culture. From an analytical standpoint, such a better understanding may be secured in several, mutually related, and conspiring ways, namely: a) by getting rid of misleading forms of speech; b) by working out a more precise conceptual equipment, as the output of a set of rational reconstructions within the conceptual borders set by linguistic uses; c) by considering the contextual presuppositions of the ongoing conceptual and terminological arrangement: how it is rooted in and mirrors, in the ideal scale of philosophical ascent, (i) the nature of legal institutions as they

17

The Philosophy of Precedent

developed historically, (ii) the nature of human institutions and human societies, (iii) the nature of man as the constitutive factor of these all; d) by imagining which parts of the ongoing conceptual and terminological equipment would stand, and which ones would instead be different, assuming momentous differences would hold concerning the nature of legal institutions, the nature of human societies, and the nature of man; e) by imagining a new conceptual and terminological equipment for looking at, and/or working within, the real institutions and practices of judicial precedent, as we know them. The search for practical, normative, justification of systems of precedent I discarded, opting for an analytical approach. Contrariwise, the search for a better understanding is again, appearances notwithstanding, one of the goal I will try to pursue in the present paper; that may indeed be regarded as being concerned, in this sense, with the philosophical foundations of precedents (as we shall see in §§ 3, 4, and 5 below). 3. THE RELEVANCE OF JUDICIAL PRECEDENTS FOR LEGAL THEORY Do legal theorists bother about judicial precedents? Should they do so? Clearly, both questions command positive answers, whether we endorse a normative conception of legal theory (legal theory as normative jurisprudence), or an analytical one (legal theory as analytical jurisprudence). This is enough for preliminary remarks. Let’s move, now, to the three issues I mentioned at the outset. C. “RATIO

DECIDENDI”,

“OBITER

DICTUM”

Apparently, we face here a sad truth: after centuries-long investigations, jurists and legal theorists seem still to be looking for the proper concept of ratio decidendi. Why is that so? Probably, due to the combined, doomed to be persistently successful, efforts of unclear thinking and the dictates of legal policy, legal ideology, and methodological warfare. Be it as it may, even a quite superficial survey of the notions of ratio decidendi, as they are defined in the scholarly civil law and common law literature, would result in a multiple-headed inventory. We may find, for instance, that the ratio decidendi has been characterized, in turn, as follows: (RD1) the element in the opinion that represents the necessary premiss, or the logically necessary step, for the decision of a case; (RD2) the legal principle that, in a judicial decision, is sufficient to decide the case at hand; (RD3) the argumentation necessary or sufficient to decide a case at hand;

18

Pierluigi Chiassoni

(RD4) the norm (rule, principle) that represents, alternatively: (a) the necessary and sufficient condition, (b) the sufficient but not necessary condition, or even (c) any necessary but not sufficient condition of the decision of a case at hand; (RD5) the norm for the facts of a case that, as a textual analysis of the precedentjudgment discloses, the judge who decided the case has actually established and/ or followed, beyond what she may have stated or thought to have stated; (RD6) the norm for the facts of a case that the judge who pronounced the judgment has expressly declared to have established and/or followed, or tacitly believed to have established and/or followed; (RD7) the norm for the facts of a case that the judge has expressly or tacitly treated as being necessary to decide the case; (RD8) the norm for the relevant facts of a case that – according to existing law, facts, and precedents – the judge who decided the case should have established and/or followed, in order to decide the case properly; (RD9) the norm for the relevant facts of a case that, according to the subsequent judge Js, the previous judge Jp, who decided the case, believed to have established and/or followed; (RD10) the norm for the relevant facts of a case that, according to the subsequent judge Js, the previous judge Jp, who decided the case, has in fact established and/or followed, notwithstanding what she may have meant to do; (RD11) the norm for the relevant facts of a case that subsequent judges Jss should have regarded as having being established and/or followed by the previous judge Jp, in the relevant precedent-judgment. All the notions of ratio decidendi reported above are clearly, though in different degrees, indeterminate. There seems to be need, accordingly, for some work of rational reconstruction, aiming at clearer and more precise notions. Before doing so, however, a few comments seem to be in order. 1. The several notions above refer to, or presuppose, a few distinctions between conceptions and kinds of rationes decidendi, concerning: a) their nature; b) their mode of expression; c) the relevance of the intention of the judge who established and/or applied (“followed”) them to decide a case; d) their objective relevance for the decision of a case; e) their systemic standing. 2. As to the nature of the ratio decidendi, there are basically two conceptions at stake, not necessarily at odds.

The Philosophy of Precedent

19

On the one hand, rationes decidendi are sometimes characterized as reasons, premisses, logically necessary steps, necessary or sufficient argumentations of a precedent-judgment (as it is the case, for instance, with the notions RD1 and RD3). This is the argumentative conception of the ratio decidendi. On the other hand, rationes decidendi are also characterized as norms, rules, principles, rulings (see, e.g., RD2, RD4, etc.). This is the normative conception of the ratio decidendi. Both conceptions – by the way – find support in the latin word “ratio”, the semantics of which includes both the idea of a reason and the idea of a rule (a measure, a criterion of judgment). The general norm applied by a judge to decide a case is a reason (the normative reason) that justifies that decision. It is, on the same token, a premiss (the normative premiss) of the reasoning the conclusion of which is the judicial order. Finally, if we look at any judgment as to a piece of reasoning provided with a logical structure, the general norm applied by the judge may also be regarded as a logically necessary step (the normative logical step) in that reasoning. The normative conception of the ratio decidendi is, hence, perfectly in tune with the argumentative conception. It leads, however, to more determinate notions. Accordingly, from an analytical point of view, normative notions should be preferred to argumentative ones. So, we would do better to regard the ratio decidendi as the norm, rule, or principle that plays a fundamental role (let’s by now be content with this) in the justification of a judicial decision. 3. As to the mode of expression of the ratio decidendi, the basic distinction the several notions above point to seems to run between explicit and implicit rationes. A ratio decidendi is explicit, when it has been provided with a discrete formulation: when it has been expressed, stated, or (as it is sometimes said) “announced” by the judge in the opinion of a precedent-judgment. By contrast, a ratio decidendi is implicit, when it is the tacit normative premiss of a judicial reasoning: when it has not been formulated, stated, announced, though – somebody claims – it has been established and applied to decide a case. In due contexts, the (presumed) implicit ratio decidendi of a precedent-judgment may be played against its (presumed) explicit ratio. Indeed, by combining this distinction with the further distinction, concerning the objective relevance of a ratio, between real ratio and apparent ratio, one may argue, for instance, that the explicit ratio of a precedent-judgment, though carefully formulated by the judge, is just a sham; it is only the apparent ratio, which should be turned down, to pay to the real ratio, the tacit, implicit, unformulated, not-announced one, the respect it deserves (it ought to have) in subsequent judicial decisions. From an analytical point of view, the two distinctions – explicit v. implicit ratio; real v. apparent ratio – are worthwhile keeping in mind, above all for their use as tools in the practical art of reasoning with precedents. One caveat should be made, however. To make sense from an analytical perspective, the distinction between explicit and implicit ratio decidendi should be recast in such a way, that a reasonably objective procedure may be followed to identify them, if any, as an outcome of a textual analysis of a precedent-judgment. This suggests

20

Pierluigi Chiassoni

appealing to the idea of the logical structure of judicial decisions, as we shall see in a moment. 4. Sometimes, the notion of ratio decidendi is defined in such a way as to take into account the (presumed) intention of the judge who pronounced the precedent-judgment (see, e.g., RD6, RD7, RD9); sometimes, however, this is not the case (see, e.g., RD1, RD2, RD3, RD4). The notions of the former kind we may call intentionalist, or subjective, notions of ratio decidendi: here, the ratio is the intended ratio, the one the judge meant to establish and apply. The notions of the latter kind, by contrast, we may call de-psychologized, or objective, notions of ratio decidendi: here, the ratio is identified, and dealt with, as a purely unintentional object, as a piece of discourse the meaning, sense and/or value of which are to be identified and appreciated without making any reference, or paying any attention, to the intention (purpose, beliefs, attitudes) of its author. The notions of ratio decidendi defined by civil law jurists are usually objective; contrariwise, the notions defined by their common law colleagues are frequently subjective. The distinction between subjective and objective rationes decidendi is relevant from two related perspectives. On the one hand, it is, at least apparently, at the core of two different sets of models concerning the “interpretation of precedents”: subjective (or rather, mixed) and objective models, respectively (as we shall see at § 4 below). On the other hand, the distinction has an argumentative use, like the ones I considered in the previous point. Indeed, in due contexts, one may play the (presumed) subjective ratio of a precedent-judgment against its (presumed) objective ratio (this way of proceeding is suggested, by the way, by the notions RD5 and RD10 above), claiming, for instance, that the former is just the apparent ratio, while the real ratio is the latter. 5. Sometimes, the notion of a ratio decidendi is defined in such a way, to include a reference to its systemic standing (see, e.g., RD8, RD11). Here, what is being defined is, more precisely, a – rather vague – notion of the correct, proper, or right, ratio decidendi for a case. The basic use of this notion clearly belongs to the criticism, either by appeal judges or by jurists, of some erroneously decided precedent-judgment. As I said before, from the standpoint of an analytical philosophy of precedents, the next step, following to the previous inquiry, should be devoted to the definition – rational reconstruction – of some notion of ratio decidendi, in such a way to provide jurists and judges with a (purportedly) better set of concepts. Taking into account the presence of subjective and objective notions of ratio decidendi in juristic thinking, and having in mind the need to play down indeterminacy in the identification of the ratio within precedent-judgments (for instance by curing the loose talk in terms of “necessary” and/or “sufficient” “conditions”), an analytical reconstruction, to be of some use, should perhaps meet the following requirements:

The Philosophy of Precedent

21

1) it should include three concepts of ratio decidendi: an objective, a subjective, and a mixed concept – it is in fact eventually up to jurists and judges choosing which one to adopt, if any, in their practice, also taking into account their purposes and the history of their legal culture; 2) in defining the objective concept, it should make reference to the logical structure of judicial decisions; such a structure should be conceived in turn, in its minimal form, as a normative modus ponens, where a normative individual categoric conclusion (the judicial order) follows from a normative general conditional premiss (the applied general norm) and a qualifying, or subsumptive, individual categoric premiss; 3) this way of proceeding, that recalls the centuries-old idea of the judicial syllogism, should carefully reject any formalism, or logicism, concerning judicial decisions; the idea that precedent-judgments have a logical structure, indeed, does not by itself imply any commitment to the mistaken view that judicial reasoning is a purely logical reasoning from evident premisses; contrariwise, as masterly shown, e.g., by Jerzy Wróblewski and Robert Alexy, the idea of the logical structure of judicial decision (in Wróblewski’s original term: the idea of the “internal justification” of judicial decisions) is just a useful tool of analysis and criticism of judicial decisions, that is perfectly compatible with the view that the premisses of the internal justification of decisions are chosen by the judge, though they should be provided in turn with some justification (“external justification”). With these remarks in mind, the three concepts of ratio decidendi may now be defined, tentatively, as follows. Ratio decidendi (objective concept) = Df. A general norm (rule, principle, ruling), that is either expressed by a sentence in the text of a precedent-judgment, or implicit in it, is the ratio decidendi of a case in relation to a certain issue to be decided, if, but only if, it is the normative general premiss of the modus ponens inference from which, together with at least another premiss (the individual subsumption premiss), the judicial order may be derived. Notice that this concept does not rule out two situations which may occur in practice, and should be briefly considered. (1) To begin with, it may happen that the same judicial order may be derived, in the text of the judicial decision, from different, but convergent, modus ponens inferences. In this case, there are as many rationes decidendi as are the different, but convergent, modus ponens inferences at stake. In such a situation, there is a plurality of different, but convergent, first-degree modus ponens inferences, and, consequently, a plurality of different, but convergent, first-degree rationes decidendi. (2) Furthermore, it may also happen that the logical structure of the judicial decision is a chain of inter-connected modus ponens inferences, where the normative general premiss of the lower-degree inference is, at the same time, the normative general conclusion of the higher-degree inference. In such case, we may distinguish for instance, within the logical structure of a precedent-judgment, a firstdegree modus ponens inference, a second-degree modus ponens inference, a third degree … and so on, up to the nth-degree, highest or ultimate, modus ponens

22

Pierluigi Chiassoni

inference. In a situation like this, there are, from a logical point of view, several rationes decidendi: from the lowest, and less abstract one, working as the normative premiss of the first-degree modus ponens inference, up to the highest, and more abstract, one, working as the normative premiss of the nth-degree, ultimate, inference. In both cases, which ratio decidendi, among the several identified on the basis of the logical structure of a decision, should be treated as provided with precedential value is not something that the logical structure as such – and this concept of ratio decidendi – can tell. Only a doctrine of precedent – providing some legal-policy-committed criterion – can do that job. In common law countries, by the way, one such criterion seems precisely to be the intention of the judge: this brings into the fore the subjective and mixed concepts of the ratio decidendi. Ratio decidendi (subjective concept) = Df. A general norm (rule, principle, ruling), that is either expressed by a sentence in the text of a precedent-judgment, or implicit in it, is the ratio decidendi of a case in relation to a certain issue to be decided, if, but only if, the judge who decided the case has intended it to be (a) the paramount legal prescription for the correct decision of the case at hand, and (b) provided with precedential value for subsequent decisions. Ratio decidendi (mixed concept) = Df. A general norm (rule, principle, ruling), that is either expressed by a sentence in the text of a precedent-judgment, or implicit in it, is the ratio decidendi of a case in relation to a certain issue to be decided, if, but only if: (a) it is the normative general premiss of a modus ponens inference from which, together with other premisses, the judicial order may be derived; (b) the judge who decided the case intended it to be (b1) the paramount legal prescription for the correct decision of that case, and (b2) provided with precedential value for subsequent decisions. Clearly, from the viewpoint of certainty and controversy over the identification of the ratio decidendi of a case, the subjective and the mixed concepts fare worse than the objective concept. This is so, however, for the objective concept is defined in such a way, to leave the problem of precedential value aside. The other two concepts, contrariwise, take into account also this important problem. Together, the three concepts bring to the light where the borders between logical analysis, on the one side, and legal standards and policy, on the other side, run: they show up to which point logical analysis may help clearing the way, and where some doctrine of precedent has to step in. From the standpoint of an analytical philosophy of precedent, of course, the limited powers of logical analysis are by no means a drawback. On the contrary: everything is given its due (unicuique suum); furthermore, doctrines of precedent become a further, relevant, subject for analytical inquiry and rational reconstruction. Ratio decidendi is usually opposed to obiter dictum: the same piece of a judgment cannot be, at the same time, and from the same conceptual perspective, both ratio and dictum. Having provided a few concepts of the former, the latter may be defined negatively, as anything, in a precedent-judgment, that is not a ratio decidendi according to the objective, or subjective, or mixed concept of ratio decidendi.

The Philosophy of Precedent

D. INTERPRETING

23

PRECEDENTS

What exactly do we interpret, when we “interpret precedents”? How do, may, or ought we to proceed while “interpreting precedents”? I think a few points should be made clear, before undertaking any inquiry on these issues. 1) From the standpoint of the relationships between interpretation and the ratio decidendi of a case, three different kinds of situations may be singled out for analysis. 2) In the first kind of situation, the ratio decidendi is the output of the interpretation of a precedent-judgment. 3) In the second kind of situation, the ratio decidendi is instead, in itself, the subject or matter of a set of operations of (what may be called) textual interpretation. 4) In the third, and final, kind of situation, the ratio decidendi is again, in itself, the subject or matter of a set of operations of (what may be called) meta-textual interpretation. 5) Most of the traditional common-law techniques for dealing with precedents are, in fact, techniques concerning either the textual, or the meta-textual, interpretation of a previously identified ratio decidendi of a case, as we shall see in the next two sections below (§§ 4.1, 4.2). 6) Sometimes, we run into scholars talking about “determining the ratio decidendi of a case”. From an analytical point of view, such a phrase is highly indeterminate. Indeed, according to the context, it may refer, indiscriminately, to the three sets of operations above at once (i.e., textual interpretation of a precedentjudgment, textual interpretation of a ratio decidendi, meta-textual interpretation of a ratio decidendi), or just to some one of them. 1. INTERPRETING THE RATIO DECIDENDI: TEXTUAL TECHNIQUES The techniques of textual interpretation of the ratio decidendi are suitable for re-interpreting (what is assumed to be) the original meaning (and normative scope) of a previously identified ratio decidendi1. To make this point clear, it seems useful to introduce the distinction between declarative and corrective interpretation of a ratio decidendi. Declarative interpretation aims at establishing the correct meaning and normative scope of a ratio decidendi, basically in two different situations: (a) when the original meaning has proved indeterminate; (b) when the original meaning has been (it is alleged) miscontrued by some other judge in some previous decision. Here, common-law jurists use to speak of “construing”, “measuring”, and/or “explaining” a “precedent”. Corrective interpretation, by contrast, aims at coping with a (presumed) discrepancy between the real and the apparent ratio decidendi of a case. Here, the previously 1

For this and the following section, I have found a very useful, though totally unstructured, catalogue of common-law techniques in G. Marshall, ‘Trentatré cose che si possono fare con i precedenti. Un dizionario di common law’ (1996) 6 Ragion Pratica, 29 ff.

24

Pierluigi Chiassoni

identified, apparent, ratio, i.e. the prima facie ratio established by a superficial reading of a precedent-judgment and/or by public opinion, is being re-interpreted, to make it matching with the (presumed) real, and correct, ratio: this means, in turn, that the apparent ratio has usually to be either narrowed or downsized (by way of an interpretatio restrictiva) – common lawyers talk here of “refining”, “reducing”, “confining”, “pruning”, “nu-ancing”, and perhaps, again, “measuring” the ratio decidendi – or rather extended (by way of an interpretatio extensiva) – “extending”, “loosening”. 2. INTERPRETING THE RATIO DECIDENDI: META-TEXTUAL TECHNIQUES The techniques of meta-textual interpretation of the ratio-decidendi, by contrast, fall basically into two groups. On the one hand, there are the techniques that concern the use of an established, precedent-ratio to cope with a case of first impression: this involves, usually, proceeding by analogy from the established ratio to a new ratio for the new case at hand2. On the other hand, there are the techniques that concern the institutional – precedential – value of a previously identified ratio decidendi: they deal with its having, or not having, the force of the right normative premiss as to an individual case at hand, here and now. Some of these techniques are meant to erase, weaken, or, alternatively, strenghten the authority of a precedent-ratio. Here, common lawyers perform such operations as, for instance, (a) dictum-ising (saying that a ratio is not really ratio, but mere dictum), (b) undermining (claiming there are reasons for not taking a ratio as authoritative), (c) per-incuriam-ing (saying that a ratio was pronounced out of mistake), (d) outflanking (saying that a ratio was pronounced by way of a mere experiment, without the serious intention that it should have precedential value), (e) quondam-ing (saying that a ratio was pronounced very long time ago, and, of course, a lot of things have changed since …), (f) garnishing (presenting a ratio in a better light than usual), and (g) napping (presenting a ratio as part of a rooted line of decisions). Other meta-textual interpretation techniques are meant, instead, to get rid of a precedent-ratio, maintaining that, though surely provided with precedential value in itself, it cannot nonetheless be applied to a specific case at hand. Among these techniques of precedent-ratio neutralization, distinguishing plays a central role. There are, perhaps, two different kinds of distinguishing: to wit, internal, or rule-, distinguishing, and external, or fact-, distinguishing. External distinguishing is the technique showing that a precedent-ratio ought not to be applied to a case at hand, because, in the present case, the material facts are different from the ones encompassed by the ratio – though, at a first glance, they 2

Sometimes, it is alleged that the gist of reasoning with precedent is analogical reasoning. This suggestion has, however, to be resisted. Analogy plays in fact a role in seeing whether the facts of the case at hand are of the same, or like, kind as the facts ruled by a precedent-ratio. But this, important as it may be, does not exhaust neither the techniques of common law reasoning, nor captures the axiological foundation of the authority of precedents. On this issue, see F. Schauer, ‘Why Precedent in Law (and Elsewhere) is Not Totally (or Even Substantially) About Analogy’ (2008) 3 Perspectives on Psychological Science, 454–460.

The Philosophy of Precedent

25

may have appeared like the facts decided by the precedent-ratio. Here, the jurist changes the previous description of the facts of the case, claiming roughly that, allthings-considered, they are substantially/materially different from the facts governed by the precedent. Contrariwise, internal distinguishing is the technique showing that a precedentratio ought not to be applied to a case at hand, because, at a careful reading, the material facts actually encompassed by the ratio are different from what they appeared to be at first sight, and, consequently, they are different from the material facts of the present case. Here, accordingly, the material facts of the actual case stand, while the jurist changes the description of the relevant (material, substantive) facts ruled by the precedent. External distinguishing leaves the precedent-ratio unscathed, though it leads to its non-application to a case at hand. Internal distinguishing is, instead, a form of re-interpreting the precedent-ratio. It belongs, consequently, to the techniques of textual interpretation of the ratio decidendi; apparently, it is sometimes conveyed under the label of “confining a precedent”. 3. MODELS FOR DETERMINING THE RATIO DECIDENDI OF A CASE Situations where the ratio decidendi is the output of the interpretation of a judicial decision (i.e., of the text of a judgment as a whole) are considered by models – usually: normative models – vaguely depicted as concerning the “interpretation of precedents”. Two families of such models may be singled out. On the one hand, there are Common law models: like, e.g., the well-known models by Eugene Wambaugh, Arthur Goodhart, and Rupert Cross. They are intentionalist, mixed-models, where the interpretive contribution of the logical structure of the judicial decision either is not openly recognized, or steps in, apparently, as a default-device. On the other hand, there are Civil law or Continental models (like, e.g., the one that may be derived from the writings of Jerzy Wróblewski and Robert Alexy concerning the justification of judicial decisions). They are objective, utterly depsychologized, models, where the logical structure of the judicial decision is paramount. E. EXPLAINING

THE PRACTICAL RELEVANCE OF JUDICIAL PRECEDENTS

As it is well-known, while accounting for the practical relevance of judicial precedents, and, more precisely, of the rationes decidendi of precedent-judgments, such phrases as “persuasive force”, “binding force”, “vertical binding force”, “horizontal binding force”, “de facto persuasive force”, “formal bindingness”, etc., are usually employed. In this final part of my paper, always by way of outlining some pieces of an analytical philosophy of precedents, I will proceed as follows. First, I will deal with a few – to my view, basic – distinctions (§ 5.1).

26

Pierluigi Chiassoni

Secondly, I will briefly report on what seems to be, so far, the most sophisticated attempt at dealing with the practical relevance of judicial precedents: i.e., the one performed by the scholars of the so-called Bielefelder Kreis (§ 5.2). Third, and to conclude, I will provide an alternative account of ideal-typical legal systems, considered from the viewpoint of the de iure relevance of precedents, if any (§ 5.3). 1. A FEW BASIC DISTINCTIONS Concerning the practical relevance of judicial precedents – if by that phrase we mean, in a very broad and uncompromised way, the relevance of judicial precedents for the practice of adjudication and judicial decision-making – a basic distinction to be drawn runs between de iure (or formal) relevance and de facto (or empirical) relevance. De iure, or formal, relevance is a matter of legal norms: it is institutional relevance, constituted and regulated by legal norms (rules, standards, principles). It is, more precisely, the relevance (force, influence, importance, consideration), in the ways and degrees determined by legal norms, that judicial precedents, as defined by legal norms, may or ought to have, according to positive law, while deciding subsequent cases of the same or like kind, at a certain time t, in a certain legal system LS. Contrariwise, de facto, or empirical, relevance is a matter of fact: it is the relevance (force, influence, importance, consideration) that, in fact, judicial precedents (of a certain kind) do have in the decision of subsequent cases of the same or like kind, at a certain time t, in a certain legal system LS. De facto relevance and de iure relevance are not necessarily at odds: in fact, they point to logically independent, though not necessarily unrelated, aspects of the legal world. From this standpoint, we may distinguish three kinds of de facto relevance: 1) secundum legem; 2) contra legem; 3) praeter legem. Empirical relevance secundum legem: the de facto relevance of precedents, whatever it is, is the effect of the following, by judges, of the positive legal norms concerning precedents – of their faithful allegiance to the positive doctrine of precedent. Empirical relevance contra legem: the de facto relevance of precedents, whatever it is, is the effect of the breaking, by judges, of the positive legal norms concerning precedents – of their violation of the positive doctrine of precedent. Empirical relevance praeter legem: by hypothesis, the law is assumed not to say anything about the relevance of judicial precedents, and to be, indeed, indifferent regarding it. In such a situation, the de facto relevance of precedents, whatever it is, is the effect of a judicial practice extra or praeter legem: for it has been established outside of positive law, possibly, but not necessarily, following some normative model of good judicial practice supported by the public opinion of the legal culture. Scholars frequently refer to precedents having persuasive force, as opposed to precedents having binding force. Sometimes, the distinction is meant to be tantamount to the distinction between the de facto and the de iure relevance of precedents: precedents having persuasive force have a merely de facto relevance – usually, we would say, a de facto relevance praeter legem; while, on the contrary, precedents having binding force have a de iure or formal relevance.

The Philosophy of Precedent

27

Sometimes, however, the distinction between the persuasive and the binding force of precedents seems to refer to two different forms of de iure, or formal, relevance of judicial precedents. From an analytical point of view, of course, confusions should be avoided. A final distinction that, like the previous one, belongs to the current theory of precedents runs between the horizontal and the vertical relevance of judicial precedents. Here, too, it may be worthwhile, for clarity’s sake, to combine this distinction with the basic de facto/de iure relevance divide: though, most of the time, the distinction refer to two forms of de iure or formal relevance of precedents. 2. THE BIELEFELDER KREIS The scholars of the Bielefelder Kreis (Aulis Aarnio, Robert Alexy, Zenon Bánkowski, Gunnar Bergholtz, Svein Eng, Francisco Laporta, Neil MacCormick, Geoffrey Marshall, Lech Morawski, Enrico Pattaro, Aleksander Peczenik, Alfonso Ruiz Miguel, Robert S. Summers, Michele Taruffo, Michel Troper, Marek Zirk-Sadowski, etc.3) have worked out a theoretical account of the practical relevance of precedents (better: of the rationes decidendi of precedent-judgments), where they single out four basic forms of relevance: 1. 2. 3. 4.

formal bindingness; de facto persuasive force; complementary justificatory force; mere illustrativeness or other like value4.

1. Formal bindingness = Df. “a judgment not respecting a precedent’s bindingness is not lawful and so is subject to reversal on appeal”. Three degrees of formal bindingness are singled out: (1a) strictly binding precedents: precedents are strictly binding, whenever they are subject neither to overruling, nor to any exception whatsoever; (1b) defeasibly binding precedents: precedents are defeasibly binding, whenever they are not subject to overruling, but they are subject to (well-defined or not-welldefined) exceptions; (1c) softly binding precedents: precedents are softly binding, whenever they are subject to overruling or modification, and they may also be subject to exceptions.

3 4

See N. MacCormick, R. S. Summers (eds.), Interpreting precedents. A Comparative Study (Ashgate: Aldershot, 1997), at vii. See Neil MacCormick and Robert Summers, ‘Appendix: Final Version of the Common Questions, Comparative Legal Precedent Study, September 1994’, in: Interpreting precedents. A Comparative Study, ed. N. MacCormick, R. S. Summers, 554–555; the account is given a wider scope in the essay of A. Peczenik, ‘The Binding Force of Precedent’, 461–479.

28

Pierluigi Chiassoni

2. De facto persuasive force = Df. “a judgment not respecting a precedent’s force, though lawful, is subject to criticism on this ground, and may be subject to reversal on this ground”. Two varieties of de facto persuasive force are singled out: (2a) defeasible persuasive precedents: precedents are defeasible persuasive, whenever they “should be applied unless exceptions come into play (exceptions may or may not be well defined)”; (2b) outweighable persuasive precedents: precedents are outweighable persuasive, whenever they “should be applied unless countervailing reasons apply”5. 3. Complementary justificatory force = Df. precedents have complementary justificatory force, whenever the subsequent decisions where they are not mentioned, though lawful and justified, are not “as well justified” as they would be if the precedents were invoked6. 4. Mere illustrativeness or other like value = Df. precedents have the force of a mere example, whenever the subsequent decisions, where they are not even mentioned, are lawful and well justified as well. The Bielefelder Kreis account is, as I said, perhaps the most sophisticated example of a net of concepts and distinctions worked out for capturing the wide variety of phenomena making up the practical relevance of precedents, and improving our understanding of it. Futhermore, the theoretical value of the Bielefelder Kreis account shines also as to its drawbacks, for they are nonetheless instructive. I see two major drawbacks in the Bielefelder Kreis account. The first and, to my view, more serious drawback concerns the way the de facto persuasive force of precedents has been characterized. Here, an unduly normative language has been used: one that would be more convenient when dealing with the formal, or de iure, relevance of precedents (see, for instance, where it is claimed that “a judgment not respecting a precedent’s force, though lawful, is subject to criticism on this ground, and may be subject to reversal on this ground”; or else where it is claimed that a precedent has persuasive force when it “should be applied unless countervailing reasons apply”). This point suggests that, perhaps, the distinction between formal bindingness and de facto persuasive force is not watertight: it does not point, as in my previous remarks (§ 5.1 above), to two neatly distinguished dimensions of the practical relevance of precedents. The normative dimension depending on positive legal norms, on the one hand; the empirical dimension of actual judicial practices, depending on legal culture and tradition, on the other. The second drawback of the Bielefelder Kreis account concerns the way of characterizing the three different forms of formal bindingness of precedents. Here, perhaps, a more articulated picture would be preferable (having to do, as we shall see in 5 6

Ibid., p. 555. Ibid., p. 555.

The Philosophy of Precedent

29

a moment, with two different kinds of defeasibility), as a tool for furthering our grasp of reality and classificatory capacities. I will try to outline such a picture in the following, and last, part of my paper. 3. EIGHT IDEAL-TYPICAL SYSTEMS OF FORMAL RELEVANCE Concerning the de iure relevance of precedents – or, better: of the ratio decidendi of precedent-judgments – eight formal, simple, ideal-type, systems may be singled out, basically by way of thought experiment: (1) (2) (3) (4) (5) (6) (7) (8)

prohibited-relevance systems; very-weak-argumentative-relevance systems; weak-argumentative-relevance systems; strong-argumentative-relevance systems; weak-binding-force systems; strong-binding-force systems; very-strong-binding-force systems; discretionary-relevance systems.

Let’s have a quick look at each of them. (1) Prohibited-relevance systems. The systems of this kind are characterized for prohibiting any practical relevance whatsoever to judicial precedents. On an ideal graph, they represent the zero degree of precedents’ practical relevance. How can such an outcome be accomplished? Let’s distinguish between the practical relevance outside of the individual lawsuits where the judgments are pronounced (external practical relevance) and inside of them (internal practical relevance). As to external practical relevance, this may be prevented by a strict regulation that: (a) rules out in principle any form of precedents’ authoritativeness – even of a persuasive, de facto, form, by prohibiting any reference to judicial precedents in judicial opinions; and (b) prohibits, to make such a ruling out effective, the publication, report, annotation, comment, etc., in whatever ways, of judicial decisions. This means that precedent-judgments would be doomed to a secret life in jealouslyguarded judicial archives. In this way, each judge having to decide a case would be able to draw no support at all from the previous decisions of other judges: each judge would be alone in front of a case at hand, provided only with her law-books (codes, regulations, and, perhaps, abstract juristic writings). What about a judge asking for advice and chatting with her colleagues, in the same judicial seat or elsewhere, about the cases they decided? On a very strict no-relevance system, even such contacts among judges would be prohibited. As to internal practical relevance, this may be prevented by a regulation according to which the appeal judge ought to act under a thick veil of ignorance about the

30

Pierluigi Chiassoni

way the case was decided by the lower court. In this way, each trial, though on appeal, becomes a totally new trial on the same issue. No-relevance systems appear quite weird to contemporary sensibilities. They seem, all things-considered, to be deeply irrational, unworkable, legal systems. On the one hand, the strict prohibition of precedents’ external relevance seems to result in a heavily impoverished legal culture. Case law is, to begin with, the privileged place where statutes, customs and other regulations may be tested as to their meaning, scope, and determinacy on the bench of the real facts of cases; case law is, furthermore, one of the major agency of law’s development, of its constant adaptation to the “needs” and “requests” coming from “society”; case law provides, finally, a further forum of principle, besides the political arena, where new solutions, and new rights, may be worked out by the cooperation of judges, jurists and lawyers. All this would be lost in a rigorous no-relevance system. On the other hand, the strict prohibition of precedents’ internal relevance may result in the total arbitrariness of judicial decision-making: due to the principle of the veil of ignorance, each judge, at each of the several degrees of a trial, may decide the case simply by flipping a coin. Of course, the legislator may impose to the judges a duty of justification, and make judicial opinions secret, to be read by a review court only, say, in trials for bribery and/or gross judicial negligence. But this would look foolish. By the way, actual systems (like, e.g., the French one), wishing to downsize the practical relevance of precedents, without incurring on (totally) absurd arrangements, chose, as it is known, a different path: (a) they stripped precedents of formal, or de iure relevance; (b) they did not prohibit, however, neither case law reports and literature, nor arguments from case law; (c) as to internal relevance, they imposed to any judge a duty of justification (as a check on judicial discretion), they allowed higher judges to read the opinions of the lower judges, they made the holding of the highest court, however, not automatically binding on the lower court to which the lawsuit has been referred for trying facts anew. (2) Very-weak-argumentative-relevance systems. In this kind of systems, the positive law doctrine of (the external relevance of) precedent contains two basic prescriptions: (2a) judges would do better to retrieve, and mention in their opinions, any relevant precedent; (2b) judges would do better to follow the relevant precedents, in order to promote the values of certainty, predictability, and the protection of people’s expectations. The two prescriptions are given, however, by way of mere recommendations for good judicial practice. Nothing unpleasant follows from disattending them. From a formal perspective, a judgment where these recommendations are not met, is perfectly lawful and valid, both as to its content (substance), and as to the formal adequacy of its motivation (argument), though the latter would have been a better piece of opinion, had the recommendations been followed. Accordingly, we may regard these systems as characterized by a formal relevance that is, at once, argumentative and very weak.

The Philosophy of Precedent

31

(3) Weak-argumentative-relevance systems. In this kind of systems, the positive law doctrine of (the external relevance of) precedent includes again two basic prescriptions: (3a) judges ought to retrieve, and mention in their opinions, any relevant precedent; (3b) judges would do better to follow the relevant precedents, in order to promote the values of certainty, predictability, and the protection of people’s expectations. In this case, the doctrine of precedent contains, by hypothesis, two heterogeneous ingredients: one duty-imposing norm and one recommendation. The violation of the duty affects the validity of a judgment: it may be appealed, and reversed, on that point. Once that duty has been fulfilled, however, the judgment is lawful and valid, even though the recommendation is not followed – and the judge, for instance, simply decide otherwise, without any argument, after having quoted the relevant precedents. Systems like these are characterized by a de iure relevance that is, at once, argumentative and weak. The presence of a relevant precedent, even one pointing to a different outcome from the one actually reached by the present judge, ought to be marked out; but the judge has no duty of providing arguments supporting her departure from it. Tacit distinguishing and overrulings are perfectly lawful. (4) Strong argumentative relevance systems. A system grants judicial precedents a strong argumentative de iure relevance, whenever its doctrine of (the external relevance of) precedent contains the following prescriptions: (4a) judges ought to retrieve, and mention in their opinions, any relevant precedent; (4b) judges ought to follow the relevant precedents, unless they can provide “good”, “serious”, “stronger”, “prevailing”, “overwhelming” reasons for departing from them or overruling them. In this system, the doctrine of precedent contains, by hypothesis, two duties. The violation of each one them affects the validity of a judgment, that may be appealed, and reversed, on both counts. In the terms of the Bielefelder Kreis, this system would be, roughly, a system characterized by softly binding precedents (or, perhaps, by outweighable persuasive force). I prefer to conceptualize this ideal-type in terms of de iure strong-argumentative force of precedents. Precedents impose on judges who want to get rid of them (overrule them, depart from them) a burden of argumentation: they have to find reasons for presenting a different ratio decidendi as better, stronger, (more) correct than the old one. (5) Weak-binding-force systems. A system grants judicial precedents a weak-bindingforce, or an open-defeasibility, relevance, whenever its doctrine of (the outside relevance of) precedent contains the following prescriptions: (5a) judges ought to retrieve, and mention in their opinions, any relevant precedent;

32

Pierluigi Chiassoni

(5b) Judges ought to follow the relevant precedents, even though they could provide “good”, “serious”, “stronger”, “prevailing”, “overwhelming” reasons for departing from them or overruling them; (5c) judges may nonetheless abstain from following the relevant precedents, if the case at hand falls within an exception belonging to an open list, the judge of the case may herself contribute to fill up (for, say, the list contains vaguely worded paradigmatic cases, open to analogical and a fortiori extensions). In the Bielefelder Kreis account, as we have seen (above, § 5.2), two situations are distinguished – defeasibly binding precedents and defeasibly persuasive precedents – sharing the common feature of the precedent-holding being subject to exceptions, either of a well-defined or of a not well-defined sort. I think, however, that in an analytical account of the several kinds, and systems, of de iure relevance of judicial precedents, the situations where a precedent-holding is subject to “well-defined exceptions” should be carefully kept apart from the situations where a precedent-holding may be subject, instead, to “not-well-defined exceptions”. To make things easier (but of course, the model could be made more complex and articulated, to be more realistic), I will assume the following: exceptions to a precedent-holding are well-defined if, but only if, they are a closed, precisely worded, set. Contrariwise, exceptions to a precedent-holding are not-well-defined if, but only if, they are a open, vaguely worded, set. Argumentative relevance has to do with the possibility of overruling (tacit, where it is very weak or weak; necessarily explicit, where it is strong). Binding force, which exists where overruling is formally precluded (e.g., by a positive norm like 5b), has to do with distinguishing (defeasibility-distinguishing, internal distinguishing, exception-distinguishing). Now, to any practical purpose, there is indeed a difference between strong-argumentative relevance and, as we shall see, strong and very-strong binding force; the difference however appears to be quite dim, when strong-argumentative relevance is compared with weak-binding-force. By adding some exception of her own invention, the judge may get to what she could not get to by means of overruling. (6) Strong-binding-force systems. A system grants judicial precedents a strong-bindingforce, or a closed-defeasibility, relevance, whenever its doctrine of (the external relevance of) precedent contains the following prescriptions: (6a) judges ought to retrieve, and mention in their opinions, any relevant precedent; (6b) judges ought to follow the relevant precedents, even though they could provide “good”, “serious”, “stronger”, “prevailing”, “overwhelming” reasons for departing from them or overruling them; (6c) judges may nonetheless abstain from following the relevant precedents, if the case at hand falls within an exception belonging to a closed, precisely worded, set of exceptions. The binding force of precedents, in systems of this kind, is stronger than in the previous case, for judges may defeat the precedent-holding if, but only if, an exception of a given set holds: i.e. an exception which, by hypothesis, they cannot them-

The Philosophy of Precedent

33

selves create or modify (but within the unavoidable fringes of vagueness coming along with every ordinary language expression). (7) Very-strong-binding-force systems. A system grants judicial precedents a very-strongbinding-force, or an absolute-binding-force, whenever its doctrine of (the external relevance of) precedent contains the following prescriptions: (7a) judges ought to retrieve, and mention in their opinions, any relevant precedent; (7b) judges ought to follow the relevant precedents, even though they could provide “good”, “serious”, “stronger”, “prevailing”, “overwhelming” reasons for departing from them or overruling them; (7c) judges ought to regard, and treat, the relevant precedents as subject to no exception whatsoever. (8) Discretionary-relevance systems. Finally, a system grants judicial precedents a discretionary-relevance, whenever its doctrine of (the external relevance of) precedent contains the following prescription: (8a) judges may take towards precedents whatever position they think proper, from case to case – e.g., they may feel absolutely bound by some precedent, or consider it as endowed with a strong or weak binding force, or regard it as endowed with a strong, weak or very-weak argumentative force, or even think precedents ought not to be given any relevance whatsoever as to the decision of subsequent cases. I do not know if any historical legal system is, or ever was, like this. There are, however, legal experiences the judicial practice of which, from an external or sociological viewpoint, may be described as informed by the doctrine of discretionary relevance, for judges seem to assume it, or something like it, as a de iure doctrine, in their decision-making. To conclude, I wish to make three remarks. 1) The account above can be made more complex, by adding to, and combining with, the several, ideal-type, systems of precedential de iure relevance the vertical and horizontal dimensions I mentioned earlier (§ 5.1). 2) The systems described above are ideal-types. This means that their explanatory and classificatory power always depend on adjustments to be made in the light of the legal experience that is the matter of inquiry. 3) Doctrines of precedents, strict and demanding as they may be on paper, have always to be taken, so to speak, salva interpretatione: i.e., against the background of interpretive techniques jurists and judges are used to, and/or may, employ, in a legal system, for identifying, and coping with, rationes decidendi having precedential value.

MARINA GASCÓN RATIONALITY AND (SELF) PRECEDENT: BRIEF CONSIDERATIONS CONCERNING THE GROUNDING AND IMPLICATIONS OF THE RULE OF SELF PRECEDENT1 A. CONCEPT

AND

TYPES

OF

PRECEDENT

In legal language the concept of judicial precedent is rather polysemous, referring to at least three different things. Firstly it refers sensu larguissimo to a previous judicial case which binds other courts when solving later cases. Of course a previous case can only have binding precedential value for a later one if they are similar, meaning that: 1) the legal issue to be dealt with in later case is essentially the same as that of the previous; and 2) later case rests upon facts which are relevantly analogous to those of the previous (they do not have to be identical). Secondly the concept of precedent can also refers sensu largo not to the case but to the actual judicial decision given to it. This definition of precedent seems to be more adequate, since it highlights the fact that what is actually binding in a precedent question is not properly the case but the decision made or, in particular, the judgment which it contains. But still, we have to be more precise. What is actually binding in a precedent case is not exactly the decision or judgment itself but the legal criteria, principle or ground supporting it. For this reason, the third and fundamental concept of precedent is the one which identifies sensu stricto as judicial precedent the legal criteria, principle or ground supporting a previous judicial decision which is used as a source for future decision making.2 In other words what is actually binding in a legal decision is the holding or ratio decidendi, i.e. the legal principles and grounds that hold or support the decision, as opposed to obiter dictum which does not necessarily have to be adhered to. In fact, one of the most common practical problems concerning the application of precedents is precisely the difficulty in distinguishing between ratio decidendi and obiter dictum when evaluating the effects of a particular decision. Hereafter, without further argument, we will use the term “precedent” in this third sense, that is as the legal criteria or principle for judicial decision making. Two types of precedents can be identified, depending on whether they are binding or merely persuasive. In the first case precedents must be applied to a novel case. In the second case precedents do not necessarily have to be followed but there are good reasons for doing so. Therefore the important feature that underlies the dis1 2

This paper has been written within the project framework PII1II09–0173–2296 and financed by the Junta de Comunidades de Castilla-la Mancha (Spain). Of course not everybody thinks so. Grant Lamond, for instance, holds that precedents does not create rules; and, more accurately, that common law can not be understood as a form of rule-based decision-making but as a form of case-by-case decision-making, and the doctrine of precedent constrains this decision-making by requiring later courts to treat earlier cases as correctly decided. (Grant Lamond, ‘Do Precedent create rules?’ (2005) 11 Legal Theory 1, 1–26).

36

Marina Gascón

tinction is the strength or the weakness of the reasons given to support the application of precedent. The question is thus placed on the analysis of the grounds for following the precedent. The reasons for following precedents may vary, depending on whether precedent derives from a court at a higher hierarchical level, or from a court at the same level, or indeed from the same court. Precedents are thus divided in three types: vertical, horizontal and self-precedent. Vertical precedent is the term used to define a legal precedent which derives from a prior decision made by a judge or court of a higher rank than the judge or court which later seeks to rely on the decision. The rule binding judicial decision makers to vertical precedent is known as stare decisis et quieta non movere (to stand upon decisions and not disturb what is settled) and whereby precedents become authoritative and therefore must be followed. Horizontal precedent is the one which stems from a prior decision made by a judge or court of the same hierarchical level as the current decision maker. Strictly speaking horizontal precedents are not “binding” on future courts, these precedents being merely persuasive3. Therefore it must be pointed out that horizontal precedent actually refers to the fact that courts reach a decision concerning a case by studying similar cases which have been solved by courts of equal rank. Finally I shall refer to self-precedent as the one stemming from prior decisions made by the very same judge or court dealing with today’s case. The rule binding judicial decision makers to their own precedents will be thus called the self-precedent judiciary rule. Naturally I am aware of the fact that the use of these expressions may be somewhat controversial. Firstly I have linked stare decisis only to the binding vertical precedent, but it is true to say that usually stare decisis is also linked to the doctrine stating that one’s own decisions are binding4. Moreover, some would hold that stare decisis only refers to the latter, arguing that binding vertical precedent is merely the natural consequence of the Courts hierarchical structure5. However, I have preferred to identify stare decisis only with the vertical precedent binding, given the fact, as I hold in this paper, that there is a dramatic difference between this rule and the rule of self-precedent. Secondly, it should also be pointed out that precedent deriving from the same court is not usually referred to as self precedent, but horizontal precedent. However I refrain from this common use of the term “horizontal precedent” and I consider it preferable to maintain both terms. This is due to the fact that the term “self ” precedent emphasizes, at least more than “horizontal” precedent, the obligation to abide by one’s own precedent. Additionally because I think it is important to keep or maintain an autonomous term (“horizontal precedent”) to refer to precedents, generally speaking merely persuasive, coming from analogous judges.

3 4

5

“As a general rule”, because it can also exist horizontal binding precedents. See. e.g., L. Goldstein (ed.), Precedent in Law (Clarendon Press: Oxford, 1991; Ana Laura Magaloni, El precedente constitucional en el sistema norteamericano (McGraw-Hill, Madrid, 2001); and Eduardo Sodero, ‘Sobre el cambio de los precedentes’ (2004) 21 Isonomía, 223 ff. In Spain, for example, Enrique Alonso, La interpretación de la Constitución (Centro de Estudios Constitucionales: Madrid, 1984), 165.

37

Rationality and (Self) Precedent

The binding nature of precedent is usually seen as one of the characteristic features of Common law systems as opposed to Civil law systems, where – it is held – judges are bound by law and not by judicial precedent. However, in actual fact this is not the case. It cannot be stated, unless of course we consider a simplistic if not fictitious view of the task of legal adjudication, that in Civil law systems jurisprudence holds no binding value whatsoever, nor on the other hand can it be stated that in Common law systems adhering to the statutory law plays a secondary or minor role to that of adhering to precedent. In any case, although substantial differences can be found in both systems related to the role and scope of adherence to vertical precedent,6, no relevant differences can be found regarding to the nature and implications of the binding to one’s own precedent. This is the case due to the fact that the self precedent rule, unlike the vertical precedent rule, is a tool for ensuring rationality or – which is the same – preventing arbitrariness, so being innate in the judicial activity, regardless of the special features of the legal system in which it takes place7. From now on I shall focus my attention on this rule. Firstly I will show how it could and should be understood as a rule of rationality and then I will identify the specific requirements deriving from this feature. Finally, based on this, I will be able to examine certain controversial questions relating to the application of the rule, as well as showing how these problems (like other common deficiencies in the application of the rule) are due to a misunderstanding of its nature and requirements. B. CHARACTERIZATION

AND

GROUNDING

OF THE

SELF-PRECEDENT RULE

1. VALUES PROMOTED BY BINDING TO PRECEDENTS The arguments supporting the principle of precedent (not only self-precedent but also the vertical precedent rule) are directly based on the values that it promotes. They can be summarized as follows. Firstly, precedent creates certainty and stability in the legal system, which essentially means that citizens can foresee or anticipate the probable outcome of their case; i.e., people make decisions expecting the legal rules previously used by judges to be applied to them. This feature of “following precedents” has been traditionally highlighted: when an earlier decision has been well thought out and grounded enough – this question has been noted by James Kent explaining the culture of precedent in American law – “the community have a right to regard it as a just declaration or exposition of the law, and to regulate their actions and contracts by it. It 6

7

In the same way that substantial differences between English Common Law and American Common Law related to vertical precedent system are noted. See differences in both Systems highlighted by Victoria Iturralde, El precedente en el Common Law (Civitas: Madrid, 1995). This idea is also held by Silvia Díez, even though alleging that breaking the rule is not an attack on rationality but an attack upon the most elementary idea of justice (the one of treating like alike). According to the author, “precedents must also be followed in Civil Law systems, even although a specific rule which imposes them doesn’t exist”. Therefore this doctrine “could be possible in most legal systems surrounding us” (cfr. El precedente administrativo. Fundamentos y eficacia vinculante (Marcial Pons: Madrid, 2008), 108).

38

Marina Gascón

would therefore be extremely inconvenient to the public if precedents were not duly regarded, and implicitly followed”8. Secondly, and very importantly, by following precedent the principle of formal equity is promoted, given the fact that like cases are treated alike. Thirdly, by following the rule of precedent courts save time, given that a legal solution already exists for the case and, what is more, it does not have to be justified. More precisely, when precedent legal criterion is applied no further justification is required, since it has been already justified in the precedent case where it was settled. Thus, by standing by precedent, a heavy burden is taken off courts. It is the discharge function fulfilled by following precedent9. Nevertheless, by adhering to ones own precedents, unlike binding to precedents stemming from external courts, an added value is promoted: that of rationality in adjudicating the law. Indeed, I will hold here the thesis that the self-precedent judiciary rule is not, strictly speaking, a legal tool for legal security, stability and equality, but mainly a guarantee against arbitrariness. In other words, precedent is essentially a rationality rule and hence a rule of legal argumentation10. 2. CHARACTERIZATION AND GROUNDING OF THE RULE 2.1. Legal Manifestation of the Principle of Universalizability Although on many occasions self precedent rule is considered a unifying mechanism for jurisprudence based on the need to guarantee legal security, stability and equality, it could be argued that this is not the most appropriate interpretation. I would argue that the doctrine of self precedent should be interpreted in the wider sense as a manifestation of the Kantian principle of universalizability in legal discourse for judges and courts, given that this principle states that one right answer is required for the same cases, and this is precisely, though in other words, what the rule of self precedent represents. What is more, even if it cannot be said that one right answer exists in law which can be found by every judge or court, the principle of universalizability requires the supposition that one right answer exists for every judge or court considered on an individual basis. The self precedent rule is precisely the manifestation of this requirement to act “as if ” one right answer actually exists in law. Universalizability, in effect, expresses a fundamental principle of practical Kantian reasoning, whereby one ought to act in accordance with criteria which, consid-

8

James Kent, Commentaries on American Law, vol. I, 11th ed. (Little, Brown and Company: Boston, 1867), vol. I, 519–20 (cited by E. Sodero, “El cambio de precedentes” (n. 4, p.525). 9 Concerning the discharge function of dogmatics and precedent, see. R. Alexy, A Theory of Legal Argumentation: The Theory of Rational Discourse as Theory of Legal Justification, trans. R. Adler & N. MacCormick (Clarendon Press: Oxford, 1989), 329, 336. 10 Leonor Moral also places the study of precedent in general and particularly of the self-precedent in the discourse of legal reasoning. I deem this approach to be right, which thus moves away from the traditional but fruitless analysis from the standpoint of the sources of law theory (See, El precedente judicial (Marcial Pons: Madrid, 2002).

Rationality and (Self) Precedent

39

ered as correct, can willingly be used in every future like case11. Consequently, in the legal sense this principle requires judges and courts, when solving a controversy, to make the decision according to criteria which can be used willingly in future relevantly analogous cases. It is therefore an extremely formal ethical principle which binds all cases: a non universalizable decision could not be considered morally acceptable. But the requirement of universalizability is particularly evident when judges act discretionarily; that is, when no conclusive or peremptory rules exist when solving a specific case; or which is the same, when all of the legal material available cannot offer the desired right answer. In these cases, where we need to choose between different options for a specific problem, it could be said that Kantian universalizability is almost the only one guarantee of rationality12. It is due to the fact that the natural domain of the rule is that of legal discretionality, the rule becomes meaningless where institutionally one right answer is guaranteed. This is precisely what happens when legal mechanisms culminating in a higher court are established in order to unify jurisprudence (that is, when a binding vertical precedent exists): in this given case one right answer does exist, and it is precisely the one offered by that higher court whose precedent binds the other (lower) courts. That is why binding lower judges and courts to the self precedent rule becomes meaningless, given that they cannot act discretionarily but have to adhere to jurisprudence from the higher court. However, on the other hand, the higher court must abide by the rule of precedent, in order to guarantee that, when establishing the right answer which will bind the other courts, it acts rationally. In short, in these cases lower courts would act correctly if they adhere to the higher court precedent. The higher court would act correctly following the self precedent rule. The thesis holding the self precedent rule is a rule for argumentation projecting the universalizability maxim in legal discourse has been held by some of the main representatives of the theory of legal argumentation. Robert Alexy, for instance, links self precedent to one of the basic rules of general practical discourse, whereby “every speaker who applies a predicate F to an object a must be prepared to apply F to every other object which is like a in all relevant respects”. Furthermore, Aulis Aarnio holds that “a discussant cannot refer to a value judgment that he is not prepared to generalize to cover other similar cases”13. Neil MacCormick also states that universalizability in this noted sense is one of the requirements which legal argumentation has to meet in hard cases.

11

It is well known that the first formulation of the Kantian categorical imperative states “act only according to that maxim whereby you can at the same time will that it should become an universal law”, Inmanuel Kant, Grounding for the Metaphysics of Morals (1785), trans. by James W. Ellington, 3rd. ed. (Hackett Pub: Indianapolis, 1993), 30. 12 I use the word “almost”, due to the fact that other criteria for rationality exist and must be considered. Particularly, the decision should be consistent and coherent with the principles of the system and the rest of the decisions taken, and its consequences should be acceptable and should not gravely endanger goods or things considered to be valuable. These principles, together with that of universalizability, are proposals put forward by Neil MacCormick as general criteria for rationality of decisions in practical and legal problems. (N. MacCormick, Legal Reasoning and Legal Theory (Clarendon Press: Oxford, 1978), chapter V). 13 Robert Alexy, A Theory of Legal Argumentation (n. 9), p.188; 262 ff; Aulis Aarnio, The Rational as Reasonable: a Treatise on Legal Justification (Reidel: Dordrecht, 1987), 198.

40

Marina Gascón

2.2. Content of the Rule In my view it is precisely MacCormick who offers the most suitable explanation of the content of this rule. He holds that the principle of universalizability should be applied to legal discourse both looking towards the future and looking back at the past, in such a way that “I must decide today’s case on grounds which I am willing to adopt for the decision of future similar cases, just as much as it implies that I must today have regard to my earlier decisions in past similar cases”14. The rule, in other words, works 1º) imposing a requirement to adhere to precedent, but especially and above all 2º) imposing at any rate a requirement to apply universalizable criteria. This second part of the definition boils down to the core of the rule and entails three important things which in my opinion must be highlighted in order to clarify some misunderstandings which could arise in its application. 1. The Self-Precedent rule is a rule for rationality (…) The noted feature (i.e. the rule as requirement of universalizable criteria) entails, firstly, that there is a dramatic difference between the self-precedent rule, on the one hand, and the vertical precedent rule, on the other. The latter merely offers a standard for solving the case, and therefore there is no relevant difference between “following law” and “following precedent”. In fact, it could be said that stare decisis carries out the same function as abiding by the law: to oblige courts to solve cases adhering to pre-established legal parameters, namely the principle of rule of law (and not of men) or principio de estado de derecho. The rule of self precedent however forces each Judge or Court to take into account the principle of universalizability, that is not so much as to give the same sentence as the day before, but to pronounce today one sentence which they would be willing to pronounce time and time again in the future, and furthermore would desire others to do so. That is to say, while the important feature in vertical precedent is that it offers a substantial criteria for solving the case to which the court must adhere, in the rule of self precedent the important thing is not that the court adheres to its own prior criteria, which could not exist, but that it adopts a universalizing attitude when dealing with a specific case and solves the case based on grounds which it would be willing to use in all future similar cases. To sum up, vertical precedent indicates the criteria to be used to solve a case: precisely the same used in a previous case. The rule of self precedent acts merely as a rationality check in solving the case, obliging the court to adopt universalizing criteria but not dictating what the criteria must be. 2. (…) that binds every judge and court (…) The characterization of the self-precedent rule as the obligation of universalizing ones own decisions entails, secondly, that it is not confined to a certain particular

14

Neil MacCormick, Legal Reasoning… (n. 12), 75. This is exactly the double, complementary nature which is at the very core of self precedent: “the decision made in a specific previous case represents a good reason to solve the analogous today’s case, and the today’s case decision must be made considering the future, that is wondering whether it could work as a general rule” (see Luis Prieto, Sobre principios y normas. Problemas del razonamiento jurídico (Centro de Estudios Constitucionales: Madrid, 1992), 65, interpreting Neil MacCormick).

Rationality and (Self) Precedent

41

group of judges and courts. Mainly being a rule of legal argumentation which guarantees rationality, it binds every judge and court, regardless of any legal prevision. 3. (…) and does not impede abandoning precedent Finally, and most importantly, this characterization implies that the rule of self precedent does not impede a court refraining from using its own precedent15. Quite the contrary, on many occasions by adhering to the principle of universalizability one must abandon his own precedent. This occurs, for instance, when it is thought (and can be firmly shown) that in previous cases the decision was wrong because the law was misunderstood or misapplied, and hence it is required to substitute the founded-in-error precedential legal criteria by the right and proper one16. But this also occurs when precedent criteria is no longer considered to be adequate, either because the normative context in which the law is inserted has changed, or because the circumstances or social estimations which justified the precedential criteria have changed; in actual fact this is the real sense of evolutionary interpretation17. Without this permanent possibility of adapting precedent, the law cannot reach its substantial goals18. Nevertheless, abandoning or adapting one’s own precedent is not something that can be done without further implications but it entails a burden of argumentation19 which supposes: 1) not only an ordinary explanation regarding the (new) rea15

16

17

18

19

Acccording to Francisco Laporta, what this rule actually does is impose upon the court decision maker the obligation to abide by previous rulings “provided that there is no relevant reason to modify them” (…) Precedents, as reasons for deciding, “are in the field of reasons, and can be argued and possibly abandoned” (Cfr. F. Laporta, ‘La fuerza vinculante de la jurisprudencia y la lógica del precedente’, presentation of Víctor Ferreres and Juan Antonio Xiol´s Work El carácter vinculante de la jurisprudencia (Fundación Coloquio Jurídico Europeo: Madrid, 2009), 34, 37). Thomas Hobbes expresses this idea very clearly when maintaining that if a judge deems that a precedent is wrong and “in another like case he find it more consonant to the equity to give a contrary sentence, he is obliged to do it. No man’s error becomes his own law, nor obliges him to persist in it” (cfr. Leviathan, (Andrew Crooke, at the Green Dragon: London, 1651), 170). “It is revolting – was held by Oliver Wendell Holmes highlighting this point- to have no better reason for a rule of law than that so it was laid down in the time of Henry IV. It is still more revolting if the grounds upon which it was laid down have vanished long since, and the rule simply persist from blind imitation of the past” (Oliver Wendell Holmes, ‘The Path of the Law’ (1897) 10 Harvard Law Review, 469). I totally agree with Karl Larenz in saying that: “The Judge should not (…) blindly abide by the precedent. He is not merely allowed but even obliged to depart from it, if convinced that it contains an incorrect or insufficiently grounded interpretation of law, or if the issue, rightly solved at one time, must nowadays be solved otherwise due to the fact that the normative context or even the entire legal system have changed” (Karl Larenz, Metodología de la ciencia del derecho, 4th ed. 1979, trans. M. Rodríguez Molinero (Ariel: Barcelona, 1994), 430). This aspect of burden of argumentation was underlined by Chaim Perelman when explaining the principle of inertia, according to which, adopted criteria and principles do not need to be discussed anew each time and will continue in the future, unless the change can be adequately justified through argument. It is, thus, change, and not inertia that needs justification (Chaïm Perelman y L- Olbrechts-Tyteca, Traité de l’argumentation. La nouvelle rhétorique (Brussels, 1970), 142 ff). Furthermore Alexy posits that the appeal to preexisting praxis does not require justification, but “whoever wishes to depart from a precedent carries the burden of argument”. This is one of the main rules regarding the use of precedent. (Alexy, Theory of Legal Argumentation, (n. 9), 171). Moreover, Aarnio states that it is not reasonable to reject without further justification what

42

Marina Gascón

sons grounding the decision, but also includes 2) a supplementary justification of departure from prior criteria. This justification must be precisely governed by the rule of universalizability, and consequently consists in showing that the new criteria is one which, considered to be correct (or at least more correct than the previous one), could be willingly used in the future. Furthermore, it is clear that the need to justify an overruling decision becomes more stringent when precedent has been based on solid grounds. And certainly it is true to say that if a decision has been made upon mature and solid argumentation there is a presumption in favor of it correctness, and therefore it cannot be abandoned without a particularly intense argumentative effort. In short, binding to one’s own precedent is not an inexorable command. It is perfectly plausible for a court not to adhere to its own precedent. However overruling is not allowed without further explanation, and this explanation must be guided by the principle of universalizability. Consequently an overruling would only be acceptable when it can be shown or demonstrated that the precedential criteria was wrong or inadequate and the current criteria is, on the other hand, the right and proper one which will be willingly used in the future. C. CONTROVERSIAL QUESTIONS. Although the rule of self precedent is reasonably accepted even in Statutory Law system, there are a number of questions regarding its application which could be considered problematic and are on many occasions surrounded by doctrinal debate. At least three of these controversial questions are worthy of remark, because they show clearly how in the majority of cases the “problem” is caused by the lack of a right and proper understanding of the nature of the rule. 1. CHANGES IN COURT MAKE-UP. One of the most common problems concerning self precedent is that of whether the rule could be applied when the individuals embodying the judicial law-maker (Judge or Court) who created the precedent have changed. The problem is displayed in the two following ways: 1ª) If the rule of self precedent establishes that each Judge or Court is bound by its “own” precedent, it is worthy to ask whether when a change in the make up of the court has taken place, should the court be considered the “same”. The doubt arising is whether a judicial law-maker, for purposes of applying the rule, can or cannot be defined in terms of the individual members who embody it. And this connects with the second formulation of the problem. 2ª) If the aim of the self-precedent judiciary rule is not to unify jurisprudence but merely to avoid arbitrary decision making, then it can only be applied when individuals embodying the judicial body are the same as those who created particuhas been previously adopted as prevalent, and therefore a change in the prevailing situation must be necessarily justified (see The Rational as Reasonable, n. 13, 202).

Rationality and (Self) Precedent

43

lar precedent. However, when these individuals are no longer the same, applying the rule would have no sense, given the fact that in the case it was applied it would work in actual fact as a rule to unify jurisprudence. Whether it is one or the other, it is often held that the rule cannot be applied when individuals embodying the judicial decision maker have changed. I however would argue that the noted issue does not represent a problem. Whether the individuals that embody a judge or a court have changed or not is not a relevant question in terms of determining the duty to obey the self-precedent judiciary rule. The rule is perfectly applicable, regardless of the changes made, and the reasoning behind it is extremely simple. The rule, not aiming to unify legal decision’s criteria but to avoid arbitrary decision making, and thus not obliging the court decision maker to adhere strictly in all circumstances to its precedents but only , binds every court no matter whether individuals that embody it are or are not the same who created the precedent. It is perfectly “logical” to expect a court to justify a (new) decision’s criteria where it differs from precedent, given the fact that it has not yet been justified in that court. And it is perfectly “logical” to require that this justification abide by the principle of universalizability. On the other hand, to say that the rule cannot be applied when the make-up of the court has changed would be equivalent to saying that the “new” court can establish a new criteria different from the one precedent without having to justify why that new criteria is more correct and appropriate than the precedent. This situation would be totally unacceptable. Moreover, it is not that the changes in make-up of courts are a problem in order to apply the rule but indeed they represent its “natural playing field”. To say more precisely: “overruling of precedent rarely occurs without a change in the Court´s personnel”20. This is, once again, due to a very simple reason. Justifying a change of precedent involves demonstrating that it adheres to the principle of universalizability, that is showing that, for whatever the reason may be, the new criteria is accepted as being more correct and appropriate and therefore will be maintained in the future. It is clear that this situation (namely, that the judicial law maker believes that a more correct or appropriate criteria exists which is the one to be used in future like cases) can arise when a judicial law maker is embodied by the same individuals who held the precedent line. But it is also clear that this tends to take place when there is a change in make-up of Courts. In short, there is no incompatibility whatsoever between rule of self precedent and changes in the composition of Courts, for at least two reasons. Firstly, because the rule does not impose on courts the obligation of following precedent at any rate, but merely imposes the need to provide justification for its overruling, an act which can and should be carried out regardless of whether members of the Court who today choose to abandon precedent criteria are or are not the same as those who initially adopted it. Secondly, due to the fact that changes in precedent generally occur when there is a change in Court’s personnel.

20

See. Justice Scalia, in the sentence South Carolina v. Gatheres, 490 US, 805 (1989).

44

Marina Gascón

2. APPLYING THE RULE WHEN A COURT WORKS AT SEVERAL PANELS OR SECTIONS. Another question regarding the application of the self-precedent rule arises when considering courts that work at several panels, or even at several sections within the same panel. In this case it could be asked whether the rule should only be applied to the court in question or indeed to each panel or section on an individual basis. The noted question is sometimes formulated as follows: what must be considered a “court” in terms of the self-precedent judiciary rule? In particular, must every panel or section of a court be considered only as an instance of this court or, on the contrary, must they be considered a court in terms of the self-precedent doctrine? It is worth pointing out that if each section or panel is considered as merely a part of the Court, thus applying the rule to the “Court” as one judicial body, then it would be necessary – as a logically previous stage to the rule application- to articulate mechanisms within the court in order to unify jurisprudence among the different panels or sections. However, on the other hand, if the answer to our question is that each panel or section is considered to be an independent judicial body, thus being the rule applicable to each one of them, then we would be accepting and validating contradictory jurisprudence to lie side-by-side within the same court, a situation that would not be fully understood or accepted by the general public. It is precisely this possibility of conflicting jurisprudence within the same Court which explains why those who consider the self-precedent rule as a mechanism for unifying jurisprudence tend to respond in the first sense (that is, considering that every section or panel must be seen simply as an instance of the court), then being critical of those jurisprudential practices which, in terms of the application of the rule, foster the autonomous nature of the different panels of the court, or even the different sections of each panel. The distrust of these critics is perfectly comprehensible to say the least, above all if we consider that jurisprudential unity promotes values worthy of protection such as legal security, stability and equality. However to give a proper response to our question it is important not to lose sight of the meaning and the basis of the rule. We must bear in mind that 1º) the rule, in the strictest sense, is a guarantee of rationality (or a tool for preventing arbitrariness) which in turn acts as a jurisprudential unifying mechanism; and 2º) it is articulated as a rule for legal argumentation which merely imposes court decision makers . Consequently, considering each panel or even each section of a court as an autonomous body (and thus allowing conflicting jurisprudence to coexist within the same Court) does not appear to be incompatible with the logic of the rule. However, by saying it is not incompatible does not mean that it is necessary: neither is it incompatible with the logic of the rule considering different panels as instances of one unique Court and establishing, as a previous stage in the application of the self precedent rule in this one unique court, some kind of mechanism in order to unify jurisprudence among the panels 21. I consider this second option to be preferable, since it at least also guarantees “one right answer” within the same Court. 21

With a similar model to that used in Germany to avoid possible contradictions between the five Federal Supreme Courts.

Rationality and (Self) Precedent

45

3. RETROACTIVE OR PROSPECTIVE EFFECTS OF OVERRULING Lastly, another relevant controversial question to deal with would be the prospective or retrospective effects of overruling. The exact question is the following: In the case of overruling, should the new criteria adopted by the Court only be applied to cases emerging hereafter (prospective overruling), or should it also be applied to cases stemming from the past and still to be solved (retroactive overruling)?22. This is an important point which involves the value of legal security. Effectively, one could answer, in accordance with the principle of universalizability, that the new criteria, once established, should be applied to every case to be solved23, consequently 1) not only to cases emerging after the change but also 2) to the case in which the change is made and, in general, 3) to all those that, even although stem from facts taken place before the change, are still unresolved. However if this was the case a retroactive application of law would be being admitted (given that in the last two cases the criteria for decision – which ultimately is the “law” for the case- would have been “created” ex post facto) and hence legal security would be being damaged. So, which effects are the most appropriate? In general terms, retroactive overruling is not usually associated with the principle of universalizability but with the theory of declaratory adjudication referring to Blackstone’s claim that judges are “the living oracles” of law. In virtue of this theory, judges, when solving cases, do not make new laws but simply verify and declare existing laws. As a result – it is held – an overruling sentence merely verifies and declares (in a correct manner) pre-existing law which had been wrongfully declared by the overruled precedent, and hence the new legal criteria would have to be applied retroactively24, that is, to all pending cases. On the other hand, prospective overruling thesis tends to be bound to the theory of creative adjudication. That theory emphasizes the role of judicial election in the law adjudication, and, in short, the law-making power of the judiciary. Thus – it is held – the “new jurisprudential criteria” constitutes in a certain sense new law, and therefore it must have prospective effects. I believe however these associations to be incorrect. The theory of adjudication used (declarative or creative) could be relevant in terms of accepting or not changing precedents. But in my opinion this has no relevance whatsoever with the effects attributed to such change. If it is held, in effect, the declarative theory of jurisprudence, then it is “logical” to understand that a consolidated precedent (that is, a previous verification and declaration of law for a certain case type) can only be abandoned if it is wrong or mistaken. Outside this case any other change would be equivalent to creating law, which is a task that can only be done by Parliament. This is why the House of Lords, governed 22

The question, as it will be seen, marks one of the most important differences between stare decisis in England and in the USA. However it is not only perceived as a relevant question in Common Law. In Germany, for instance, this question was already dealt with in the work of Wilhelm Knittel, Zum Problem der Rückwirkung bei einer Änderung der Rechtsprechung (Verlag Gieseking, 1965). 23 In fact, the principle of universalizability impedes solving cases in accordance with criteria intended as wrong or incorrect and justified so. 24 “The making power of the judiciary is in theory denied [which explains why] most civil law Courts thus far have failed to avail themselves of the technique of prospective overruling, Schlesinger-Baade-Damaska-Herzog, Comparative Law, 5ª ed., (Mineola, N. Y., 1988), 646.

46

Marina Gascón

by declarative theory, only allows precedent to be changed under exceptional circumstances and only if the abandoned precedent is ab iniuria, that is unlawful. If, on the contrary, the creative theory of jurisprudence is held there are no impediments to allow precedent to be abandoned also when it becomes obsolete or in any case less correct or less right than the new proposed criteria. This is why in the United States changing precedent occurs fairly often25. It is also true that in England retroactive effects are attributed to overruling whereas in the United States prospective overruling is accepted without any difficulty. But – let me insist – I think this is not an inexorable result of maintaining a declarative theory of jurisprudence in the first case and a somewhat creative theory in the second one, due to the following. The first equivalence (declarative theory = retroactivity) is not valid because, even if we held that the new criteria simply verifies and declares the will of law wrongfully declared in the previous criteria, what is clear is that citizens have up until now trusted that “the will of law is actually that declared by prior criteria”, and therefore our compromise to legal security could lead us to give only prospective effects to the new criteria. Neither is the second equivalence (creative theory = prospectivity) valid, because, although we state that new criteria constitutes new law, there is nothing, except the principle of legal security, which impedes us from attributing retroactive effects26 to it. That is, if prospective effects are given to new criteria it is only not to harm legal security and for no other reason. It should be noted, in short, that the kind of effects (retrospective or prospective) given to the new criteria is based on legal security considerations and not (or not mainly) on the underlying theory of legal decision. Taking this fact into account, I believe that the correct answer to the question mentioned above has to be the following. As a general rule, when there is no special need to guarantee legal security because the new jurisprudential criteria is more favorable (or less right restrictive) than the one precedent, its application should not be deferred: in virtue of the principle of universalizability retroactivity is imposed. That is, the new criteria should be applied to all cases to be solved in the future, regardless if they stem from facts that happened before the change; and, of course, it ought to be applied as well to solve the case in which the change is brought about. In this latter case for an additional reason: it would be absurd and unjust that the specific case in which a new more favorable criteria is established were solved using the previous (and less favorable) one. However, when the new jurisprudential criteria is less favorable (or more rights-restrictive) than the previous one, then, for the sake of legal security, prospectivity is required. That is, its application ought to be deferred to cases stemming from facts happening after the overruling27. Furthermore, when the norms 25

By affirming the general principle of adherence to precedent – observed the Court of Appeal of Pennsylvania – doesn’t mean that “we must consecrate the mere blunders of those who were before us, and stumble every time we come to the place where they have stumbled. A palpable mistake, violating justice, reason, and law, must be corrected, no matter by whom it may have been made”, McDowell v. Oyer 21 Pa 417, 423 (1853). 26 If we consider the creative theory of jurisprudence to be consustancial to prospectivity it is because we assume the legal system is committed to the individual rights and with the legal security as minimal requirement to warrant those. Things, however, are not necessarily so. 27 For these purposes, it is worth underlining that the duty of legislatures to respect legal security is also applicable to the judiciary. It cannot be allowed to reach through jurisprudence what is prohibited through the law. As noticed in Bouie v. City of Columbia, 378, U.S. 347, 353–354

47

Rationality and (Self) Precedent

to be applied are criminal law norms, sanctioning norms or in any case rights-restrictive norms, it is not only legal security but also the principle of legality which requires a prospective application of the new less favorable jurisprudence28. In conclusion, the effects of a change of jurisprudence are governed by the principle of universalizability (which requires retroactivity) and by the principle of legal security (which requires constraining retroactivity when the application of new criteria bears more right restrictive consequences than the application of the precedent one). Hence, if the new criteria is more favorable or less rights-restrictive than the previous one it will be applied retroactively. If it is less favorable or more rights-restrictive it will be applied prospectively29. It is worthy to state that holding – as it has just been done – the thesis of prospective effects of new jurisprudential criteria when this is less favorable, and justify it on the need to conserve legal security, is not incompatible with characterizing the rule of self precedent as a rule for rationality and not (at least primarily) as a tool for protecting legal security. And it is not incompatible due to the fact that the argument over the retroactive or prospective effects of the new criteria has little to do with the nature of the rule but only with the weight of the compromise with guaranteeing legal security. D. DEFICIENCIES

WHEN

APPLYING

THE

RULE.

Finally, it is also worthy of remark the most common deficiencies in the application of self-precedent rule, because many of them are also caused by a misunderstanding of the nature and requirements of it. These deficiencies arise in the following cases. a) When fulfillment of the rule is avoided through skilful and cunning distinguishing, that is through fictitious justification that the cases are not relevantly similar.

[1964], “If a state legislature is barred by the Ex Post Facto Clause from passing … a law, it must follow that a State Supreme Court is barred by the Due Process Clause from achieving precisely the same result by judicial construction”. 28 Hence, if the judicial interpretation of a criminal law is more favorable to the defendant than the precedential interpretation was, then that interpretation can be applied to today’s case. But if such interpretation is less favorable to the defendant, then he will not have to bear it, thus having this interpretation only prospective effects. To hold otherwise would be equivalent to infringe the principle of criminal legality, which states than no one shall be held guilty on the basis of an inexistent law at the time when the facts were committed. The position of Victor Ferreres concerning the scope of jurisprudential changes that are unfavorable to the defendant is not so forceful but also highlights the problem of giving retroactive effects to its: if we take the value of jurisprudente seriously, we have to recognize that variations in it against the defendant not always can have retroactive effects” (Victor Ferreres, ‘Sobre la posible fuerza vinculante de la jurisprudencia’, in: El carácter vinculante de la jurisprudencia (n. 15), at 72, emphasis added). 29 I thus disagree with Eduardo Sodero, who states that, following on from diverting jurisprudence of the U.S. Supreme Court holding that both prospective and retroactive effects however unfavorable they may be, “no formula could sufficiently answer this problem, hence it seems necessary to conclude that it will be the prudential judgment of the judge which ought to determine in each case the most just alternative hic et nunc” (cfr. ‘El cambio de precedentes’ (n. 4), 250).

48

Marina Gascón

Distinguishing indeed consists in reporting that the facts of the current case or the juridical questions involved in it are not similar enough to those of the previous case, thus holding that the previous case is not a precedent for the current one. Hence, distinguishing is not really an exception to the rule of precedent: using distinguishing is merely pointing out that the elements required in order to consider precedent do not exist. Even so it cannot be ignored that distinguising is also a ‘tool’ that allows a judge to avoid following a precedent which he would otherwise be bound to follow. It is remarkably easy to do so: one only has to hold that the current set of facts is sufficiently different from that of the earlier (precedential) case, thus justifying not to follow the decision settled that time. Obviously this behavior is highly objectionable, since precedent is not abandoned but at the same time its application in a specific case is avoided, which suggest that the decision in that case could be grounded on ad personam motives and it is in any case intolerable. For this reason, in order to avoid the misuse of distinguishing to get round the obligation to follow one’s own precedent, the decision would have to be sufficiently and rightly motivated. b) When the rule is not applied in all courts. Being a rationality rule, the self-precedent rule binds every court, also (and perhaps firstly) those who are higher up in the hierarchy of the judicial order. However certain higher courts do not regard themselves as bound by their own decisions. Most blatant and unacceptable cases occur when even the very court that settles the rule does not adhere to it. This is the case of the Spanish Constitutional Court, which despite the fact that has imposed this rule as an implication of the principle of equality when adjudicating law (art. 14 CE), delivers overruling sentences with certain flippancy, not only in terms of questionable frequency in departing from earlier precedents but also in terms of lack of adequate justification30. Overruling decisions without due justification are always dangerous, but even more so when carried out in a high court. And especially when it is the very same high court which created the rule that binds others to their own precedents. It is true that the question of applicability of self precedent when interpreting Constitution is a rather controversial question. In my opinion, however, this controversy lacks grounds. The rule of self precedent, at least regarding the characterization which has been made here, is perfectly applicable to constitutional interpretation. Concerning this, I believe Henry C. Black was correct in stating that “the principle of stare decisis applies with special force to the construction of constitutions, and an interpretation once deliberately put upon the provisions of such an instrument should not be departed from without grave reasons”31. A departure from constitutional precedent always requires a special justification. What is more, it could also be said that the self-precedent rule binds especially constitutional courts, given the particular burden of discretionality distinctive of the interpretation of such an open and undetermined text, which in turn entails that, in case of overrul30 31

To begin with, most common practice is simply not to motivate, considering in turn that there is an implicit motivation deriving from the sentence. Henry Campbell Black, Handbook of American Constitutional Law, 4ª ed. (St. Paul: West Publishing Company, 1927), at 90 (now published in Handbook on the Construction and Interpretation of the Laws (The Lawbook Exchange Ltd., Clark, NJ, 2008), at 34).

Rationality and (Self) Precedent

49

ing, a constitutional court must make an explicit, adequate and especially intense and clear justification. c) When the rule is misapplied. Let us be reminded that the rule of self-precedent requires courts to act in accordance with legal criteria which, considered as correct, they are willing to universalize. Therefore where the case to be solved has a precedent, the rule implies two alternative obligations: Either the obligation to follow precedent, if the legal criteria is considered to be correct. Or the obligation to depart from precedent, if the legal criteria is considered to be incorrect. Naturally the decision ought to be justified in both cases, that is whether precedent is abandoned or followed. What actually happens though is that while the decision to depart from precedential criteria and substitute it for a new one requires special justification, showing the reasons for its departure and complying with the principle of universalizability, the decision to follow it does not require justification of its correction and adequacy, since this has already been done in the precedent case and consequently it is understood as being implicit in today’s case. What this means is that there is a misapplication of the rule when failing to carry out one of the named obligations: on the one hand, when departing from one’s own precedent without adequate justification; but also, on the other hand, when the need to depart from it is held but in actual fact precedent is not abandoned. (c.1) Precedent is Abandoned Without Adequate Justification. This happens when precedent is abandoned in an arbitrary manner, that is without any explanation whatsoever regarding the reasons for doing so or with an inadequate or unsatisfactory explanation of it, which is the most common case. Regarding this, it is worth not to lose sight of the fact that departing from precedent ought to comply with the principle of universalizability, and therefore as deficient of justification is an overruling lacking of motivation whatsoever as one which has a motivation not inspired by (or even contrary to) that principle. (c.2) Precedent is Considered Incorrect but not Overruled. This happens in those sentences where incorrectness of precedential legal criteria is held but at the same time adhering to precedential legal criteria is decided. These are therefore sentences in which it is not the requirement of universalizability that counts but directly and merely the stare decisis doctrine. It seems to me this error can be made when the ground for self precedent is seen as essentially lying in the conservation of legal security and stability of law and not (at least primarily) in guaranteeing rationality in legal adjudication through the principle of universalizability. Effectively, when this is the case, it is easy to hold in addition that overruling requires balancing between legal security and legal stability, which weigh in favor of following precedent, and the other reasons, which weigh in favor of departing from precedent. But in this balancing the weight of legal security

50

Marina Gascón

is so great that it often prevails over the fact that precedent is wrong32. To sum up, according to this point of view overruling is only justified when the reasons for departing from precedent are really strong. That would explain why those who interpret the rule of self precedent in this way have as one as the main concerns related to its application the following: “how weighty” the mistake or incorrectness of the precedent must be as to justify the damage of legal security which departing from precedent entails33. On the contrary, when the self precedent rule is understood as a rule of rational argumentation, which primarily tries to achieve rational law adjudication where this is assumed as unavoidably discretional, the need to depart from precedent is clear when this is shown to be incorrect and legal security cannot balance against the change.

32

This happens in reasoning in Dickerson v. United States (530 U.S. 428, 429 [2000]), where sitting precedent in Miranda vs. Arizona was ratified: “whether or not we would agree with Miranda’s reasoning and its resulting rule (…) the principles of stare decisis weigh heavily against overruling it now”. And this also follows from Theodore M. Benditt statement: “a decision maker might change his mind about the relevance of the principle he has applied in the earlier case. In some such cases [he] should follow the prior decision none the less (…) In other cases it might be more important to correct what is now viewed as a mistake” (cfr. Theodore M. Benditt, ‘The rule of precedent’, in: Precedent in Law, ed. L. Goldstein ed. (Clarendon Press: Oxford, 1991), at 89). 33 See Eduardo Sodero, ‘Sobre el cambio de los precedentes’, (n. 4), at 234, note 73 and 236. That is the reason why, from this point of view, overruling is easier when precedents are “interpretative” (that is when they simply establish a certain way of interpreting law) than when they are “praeter legem” (that is when they create a norm ex nihilo by an analogical method or other integration mechanisms), given that “such mutation will be much less dramatic for legal security” (ibid., p.234).

CARLOS BERNAL PULIDO1 PRECEDENTS

AND

BALANCING

A. THE PRECEDENT-BALANCING PARADOX During the last two decades, the doctrines of precedent and balancing have been developing in civil law jurisdictions. On the one hand, in the recent European Continental and Latin American legal philosophical literature there is consensus to the extent that there are two basic kinds of norms: rules and principles. They are applied by means of two different juridico-methodological procedures: subsumption and balancing.2 While rules apply by means of subsumption, balancing is the way to apply principles. For this reason, balancing has become an essential methodological legal operation, especially for the adjudication of fundamental rights, entrenched in the last wave of Continental European and Latin American Constitutions. Many of these rights have the structure of principles. On the other hand, the doctrine of precedent has been transplanted from common law jurisdictions. According to this doctrine, previous judicial decisions are treated as “authoritative statements of the law which can serve as good legal reasons for subsequent decisions”. It also requires “judges in specific courts to treat certain previous decisions, notably of superior courts, as a binding reason”3. These authoritative statements – that are to be found in the ratio decidendi of previous judicial decisions – have the structure of rules. The ratio decidendi is a general utterance containing a general rule, which supports the decision. The ratio decidendi makes explicit the content of the legal provision, that is to say, what the relevant legal provision forbids, allows, commands, or authorises for certain kind of cases. The normativity that a ratio decidendi displays as an authoritative statements of the law is an all-or-nothing affair. It applies in its entirety or it does not apply at all for resolving a specific case. Every precedent envisages a de facto assumption which represents an abstraction of the facts of the previous case. If in a future case this de facto assumption has been proven, the judge should apply it in full. If the de facto assumption envisaged in the precedent is not verified, or despite the fact that it has been proven, an exception stipulated in it occurs as well, the judge should rule out the possibility of applying it.4 If this is the case, then puzzlement arises with the development, at the same time, of the doctrine of precedent, on the one hand, and the methodology of balancing, on the other, in civil law systems. This can be called: the precedent-balancing paradox. While the doctrine of precedent requires the application of rules, balancing is the way to apply principles.

1 2 3 4

The author thanks Thomas Bustamante and Lara Pratt for very valuable comments to this article. See Robert Alexy, ‘On the structure of legal principles’ (2000) 13 Ratio Juris, 294–304. J. Bell, ‘Sources of Law’, in: English Private Law, I, ed. Peter Birks (Oxford University Press: Oxford, 2000) 1–29. Ronald Dworkin, Taking Rights Seriously (Duckworth: London, 2000), 14.

52

Carlos Bernal Pulido

This paper aims to offer a solution to the precedent-balancing paradox. In order to accomplish this, I will explain the different relations of tension and support between precedents and balancing. The main claim is that in spite of the fact that the acceptance of balancing in certain way weakens the normativity of precedents, it makes the legal system as a whole more rational. At the same time, balancing of principles requires precedents to guarantee legal certainty, coherence and compliance with the requirements derived from the equality principle. I will argue for this claim in two sections: a system of precedents and principles and precedents in the structure of balancing. B. A SYSTEM

OF

PRECEDENTS

AND

PRINCIPLES

The doctrine of precedent accomplishes several functions in a legal system. When constitutional and legal provisions are open textured, precedents render concrete the content of those provisions with reference to particular facts (the facts of the case whose abstraction gives rise to the de facto assumption of the legal rule that every ratio decidendi generates). Most importantly, the doctrine of precedent is necessary to guarantee (1) the coherence of the legal system – if one case has been resolved in a certain way according to the law, it is coherent that an identical or a similar case should be resolved in the same way –; (2) its stability – the stability of a legal system requires that adjudication of legal rules determining the solution for cases remains stable through time –; and (3) the principle of equality – law should treat in the same way individuals concerned in identical or similar cases, precedent and future.5 The doctrine of precedent has the ability to fulfil these goals by means of the attribution of the character of legal rules to every ratio decidendi. Coherence, stability and equality can only be achieved if the ratio decidendi ought to be applied always when the same (or similar) facts of the case in which it arises take place. Notwithstanding, to make inflexible this character of rules would lead to very inconvenient results.6 Firstly, advantages in coherence for the legal order are paid with the high price of avoiding evolution of case law. This evolution is necessary for responding to challenges of new times and for correcting interpretative mistakes. Secondly, the principle of equality also sometimes requires treating differently two similar cases and, consequently, to reject precedents. Different cases deserve different solutions. Finally, flexibility of the legal system and justice in single cases must be guaranteed too. These purposes can only be reached if the judge can overrule precedents and distinguish between cases.

5

6

See on these arguments: Robert Alexy and Ralf Dreier, ‘Statutory Interpretation in the Federal Republic of Germany’, in: Interpreting Statutes. A Comparative Study, eds. Neil MacCormick and Robert Summers (Dartmouth: Aldershot et alt., 1993), 82 f.; Neil MacCormick and Robert Summers, ‘Interpretation and Justification’, in the same volume, 518 f.; Theodore M. Benditt, ‘The Rule of Precedent’, in: Precedent in Law, ed. L. Goldstein (Clarendon Press: Oxford, 1987), 89 f.; and Leonor Moral Soriano, ‘A progressive Foundation of Precedents’ (2000) Archiv für Rechtsund Sozial Philosophie, 327 f. See on these problems, for instance, Theodore M. Benditt, ‘The Rule of Precedent’, loc. lit.

53

Precedents and Balancing

Now, if distinguishing cases and overruling precedents are necessary operations for maintaining the good functioning of a legal system, then the question is how the judge should undertake them. Two possibilities arise. An irrational way would be granting to the judge an absolute discretion to perform this operation. This possibility is unacceptable in legal systems following the rule of law, in which every decision of legal officials (judges included) is ruled by the law. An alternative and more rational way is based on the idea that when a judge faces the question as to whether he or she should overrule a precedent or distinguish between cases, there is always a collision between principles that should be resolved by means of balancing. Coherence, stability and equality (in the dimension of treating equally people that are in situations alike) are principles speaking for applying the precedent. Updating the legal system, equality (in the dimension of treating different people in different way) and justice are principles that support overruling the precedent or distinguishing between cases. Balancing is the only way to solve this collision between principles. Despite the fact that accepting this kind of balancing would weaken the binding force of precedents, it would strengthen the rationality of the legal system as a whole. However, the question is: How does this kind of balancing should be carried out? Here I would like to suggest the following three criteria to undertake this specific kind of balancing: a. When, despite the similarities between the present and the precedent case, the differences between them concerning the ratio juris, that is the reason justifying that certain regulation should be given to certain de facto assumptions, are greater than the similarities, then the cases should be distinguished. Under these circumstances, the precedential decision cannot be applied as a rule to solve the present case. A new ratio decidendi ought to be found for it. b. A precedent should be overruled when, as result of social change, it was appropriate for ruling a past but not a present situation. Under these circumstances, the precedent became an inappropriate concretion of legal provisions, because the empirical effects (E) of its ratio decidendi (RD) produced in the initial situation (S1) cannot be reached in the present situation (S2). In the new situation (S2) the ratio decidendi (RD) produces negative new effects (E’). Therefore, the judge ought to overrule the radio decidendi (RD) and find a new one (RD’). c. A precedent should be overruled, when it is wrong, because it is the result of a mistaken balancing of principles. An example of this can be found in the Judgment STC 120/90 of the Spanish Constitutional Court. In this decision the Court overruled several precedents granting an absolute freedom of information and press to journalists under the Spanish Constitution 1978. The Court amended this case law and set constraints to the freedom of information with the purpose of protecting the constitutional rights to privacy and honour. C. PRECEDENTS

IN THE

STRUCTURE

OF

BALANCING.

It is not only the case that balancing enhances the doctrine of precedent by means of creating a rational procedure for distinguishing cases and overruling precedents.

54

Carlos Bernal Pulido

Precedents also enhance the doctrine of balancing by means of providing certainty and rationality to the structure of balancing. It is well known that balancing is at the heart of many theoretical discussions. One of the most important questions is whether balancing has a rational structure, and whether balancing is a rational procedure or a mere rhetoric device, useful to justify any kind of judicial decisions. This is the reason that explains the relevant theoretical efforts for provide balancing with a rational structure. The question is, however, what role should precedents play in such a structure? In order to answer this question, I will consider one of the most powerful accounts of the structure of balancing, namely, the one by Robert Alexy. 1. THE STRUCTURE OF BALANCING In A Theory of Constitutional Rights7 and in other articles, Alexy supports the thesis that balancing has a rational structure and offers a well-developed conception of it. In the last version, three elements form the structure of balancing: the law of balancing, the weight formula, and the burden of argumentation. These three elements aim to explain how a collision between principles ought to be resolved by means of balancing. According to the law of balancing, “The greater the degree of non-satisfaction of, or detriment to, one principle, the greater must be the importance of satisfying the other”.8

Consistent with this rule, the structure of balancing can be broken down into three different stages, which Alexy clearly identifies: “The first stage involves establishing the degree of non-satisfaction of, or detriment to, the first principle. This is followed by a second stage in which the importance of satisfying the competing principle is established. Finally, the third stage establishes whether the importance of satisfying the competing principle justifies the detriment to, or non-satisfaction of, the first”.9 It is important to note that the first and the second stages of balancing are rather analogous. Both operations consist of establishing the importance of the principles at stake, so we will refer to both as such.10 Indeed, in both cases Alexy claims that commensurability can be established by reference to a triadic scale: “light”, “moderate” and “serious”. The importance of the principles at stake is not the only relevant variable. A second one is the “abstract weight” of the principles.11 Different abstract weight might derive from the different legal hierarchy of the legal body in which the principle is affirmed or from which it stems, but it might be established by reference to 7

Robert Alexy, A Theory of Constitutional rights, Trans. Julian Rivers (Oxford University Press: Oxford, 2002). 8 Ibid., 102. 9 Ibid., 401. 10 Following the notation of Alexy, we will symbolize the degree of variation or non-satisfaction of the first principle in the case as IPiC and the importance in the satisfaction of the second principle in the case, as SPjC. See Ibid., 406. 11 Following the notation of Alexy, we will symbolize the abstract weight of the first principle as WPiA and the abstract weight of the second principle as WPjA.

Precedents and Balancing

55

positive social values. Thus, for instance, it could be claimed that the principle of protection of life has a greater abstract weight than that of liberty, if only because to be able to exercise one’s liberty, it is pretty obvious that one must be alive. Similarly, many national constitutional courts assigned a high abstract weight to freedom of speech, on account of its close connection with democracy, or to privacy, given its close association with human dignity. A third variable R should be added, which refers to the reliability of the empirical assumptions concerning what the measure in question means for the non-realization of the first principle and the realization of the second under the circumstances of the case. R is based on the recognition that the empirical assumptions relating to the importance of the competing principles can have a different degree of reliability, something which should affect the relative weight of each principle in the balancing exercise. According to Alexy, the weight formula is the appropriate structure to assess the importance of principles, their abstract weight, and the reliability of the empirical assumptions. This formula has the following structure:12 IPiC · WPiA · RPiC WPi,jC= SPjC · WPjA · RPjC This formula states that the concrete weight in a given case of principle Pi in relation to principle Pj results from the quotient between, on the one hand, the product of the importance of principle Pi, its abstract weight and the reliability of the empirical assumptions regarding its importance and, on the other hand, the product of the importance of principle Pj, its abstract weight, and the reliability of the empirical assumptions regarding its importance. Alexy says that it is possible to give a numerical value to the variables of the importance and abstract weight of the principles with the help of the triadic scale: light 2º that is 1; moderate 2¹ that is 2; and serious 2², that is 4. In contrast, the reliability of the factual premises must be given a quantitative expression in the following way: reliable, 2º, that is, 1; maintainable or plausible 2¯¹, that is ½; and not evidently false, 2¯², that is, ¼.13 Applying these numerical values, it is possible to determine the “concrete weight”14 of principle Pi in relation to principle Pj in the case at hand. If the concrete weight of principle Pi in relation to principle Pj is greater than the concrete weight of principle Pj in relation to principle Pi, the case should be decided according to principle Pi. On the contrary, if the concrete weight of principle Pj in relation to principle Pi is greater than the concrete weight of principle Pi in relation to principle Pj, the case should be decided according to principle Pj. If Pi supports the

12

Ibid., 408. See also Robert Alexy, ‘Die Gewichtsformel’, in: Gedächtnisschrift für Jürgen Sonnenschein, eds. Joachim Jickeli, Peter Kreutz and Dieter Reuter (De Gruyter: Berlin, 2003), 771–92. 13 See Ibid., p. 789 and f. 14 Robert Alexy, ‘On Balancing and Subsumption. A Structural Comparison’ (2003) 16 Ratio Juris, 433–49, at 433 ff.

56

Carlos Bernal Pulido

norm N1 that forbids ø and if Pj supports the norm N2 that commands ø, ø should be forbidden in the first case and ø should be commanded in the second case. The final element of the structure of balancing is the burden of argumentation. This burden operates when the application of the weight formula results in a stalemate, that is, when the weight of the principles is identical (or to express it formally, WPi, jC = WPj, iC). In the Postscript to A Theory of Constitutional Rights, Alexy defends that in stalemate cases, a restriction mandated by an act of Parliament should be considered as proportionate and therefore declared in accordance with the Constitution.15 2. PRECEDENTS AND THE WEIGHT FORMULA The structure of the weight formula sets out many interesting problems. Relevant for this article are, on the one hand, whether there are objective criteria to determine the value of the relevant variables which form the weight formula; and, on the other, were this not the case, whether precedents would help balancing to achieve if not objectivity at least rationality. In order to answer these questions, let us consider each of the variables in detail. i) The degree of importance of the competing principles It is certain that sometimes rational judgments about degrees of intensity and importance of competing principles are possible. Or what is the same, there are easy cases concerning the degree of importance of principles. For example, if a satirical magazine calls a handicapped officer a “cripple”, this clearly constitutes a serious offence against his honour (4), while at the same time contributes very slightly to the protection of freedom of speech (1). However, there are also hard cases in which the premises, both factual and normative, which should be considered in determining the importance of a principle are uncertain. This is typically the case when religious freedom is at stake. It can be doubted whether the degree of interference of a given measure with religious freedom can be determined in abstract terms, without taking into account subjective views on religious experience. Thus, the perceived degree of interference with religious freedom of a forced blood transfusion is clearly dependent on how the individual lives her religious faith. It might be fully negligible for most believers, and very serious for a Jehovah’s Witness. An assessment of the importance of the principle can only be made after taking a concrete stand which cannot be determined by the weight formula alone. In these cases, disagreement is at stake. Thus, reference to the weight formula implies a grant of discretion to the judge and to his critical moral views, as well as political ideology. There cannot be objectivity under such circumstances. However, even in such cases, the weight formula has a role to play, as it renders clear the margin of discretion left to the judge, and the room made to critical morality and political ideology in the balancing exercise. Likewise, the judge can exercise discretion when it is not clear if the case is an easy or a hard one with respect to the first variable in the weight formula, namely 15

Robert Alexy, A Theory of Constitutional Rights, op cit., 410 and f.

Precedents and Balancing

57

the importance of principles. This can be done with the help of a concrete case, the Tobacco Judgment of the German Constitutional Court, which Alexy tends to refer as an example of a clear case. The Tobacco judgment concerned the statutory duty imposed upon tobacco producers to make consumers aware of the health risks associated to smoking in the label of cigarettes, and more precisely, whether this was constitutionally sound or not. In principle, the judgment shows that there are easy cases in which “rational judgments are possible about intensity of interference and degrees of importance”, so that “an outcome can be rationally established in way of balancing”16. The labelling duty is a “relatively minor interference with freedom of profession”17, especially when compared to potential alternative measures, such as the prohibition of the sale of tobacco, or the imposition of restrictions on its sale. Moreover, it is clear that the measure fosters the protection of health. Therefore, Alexy concludes that: “The Federal Constitutional Court was not exaggerating when it stated in its decision on health warnings, that ‘according to the current state of medical knowledge, it is certain’, that smoking causes cancer and cardio-vascular disease”18. The minor interference with the freedom of profession would be balanced against satisfying the protection of health. However, different assessments of the relevant variables are not impossible. From a factual point of view, it could be said that it is not certain that the duty to advertise the heart risks stemming from tobacco in tobacco labels actually contributes to fostering consumers’ health. It could be the case that such a measure is inefficient, perhaps because consumers are already aware of what the labels tell them; or because tobacco addiction persists even if consumers are informed of its consequences, because it is to be traced back to weakness of will, and not to lack of information; or perhaps because providing information in the labels would render smoking more desirable. ii) The abstract weight of the competing principles Further room to judicial discretion derives from the measurement of the abstract weight of the principles. Abstract weight is a very singular variable, which always refers back to moral and ideological considerations. Its measurement requires the judge to take a position about the Constitution, the role of the State in the given society, and the very concept of justice. It is clear that the variable of abstract weight loses its importance when the competing principles are of the same nature. On the contrary, abstract weight becomes very relevant in solving the case when the nature of the competing principles is different. Even then, some cases might be relatively easy. It could be assumed, for example that the protection of life, or fundamental rights based closely related to the principles of human dignity and democracy should be given a higher abstract weight than others. 19 However, judges have a considerable discretion when determining the abstract weight of principles. Quite obviously, there is no complete pre-established graduation of abstract weights. The protection of life might be said to deserve the highest value (4), but one could discuss whether such a value should not also be granted to the rights closely connected to human dignity and democratic decision-making. Furthermore, should the value be 16 17 18 19

Ibid., 402. Ibidem. Ibidem. Ibid, 760, 770 and 772.

58

Carlos Bernal Pulido

the same for all rights connected to human dignity and democratic decision-making, or should it vary depending on the closeness of the connection? What about other principles like legal equality or the right to factual, and not merely legal, equality? It might be said at this point that the measurement of the abstract weight of principles according to the triadic scale clearly depends on the ideology of the judge. An individualistic judge will give the highest abstract weight to liberty, while a communitarian judge might give the greatest weight to the common good. The judge should solve the case according to the best moral argument, but sometimes it is not easy to know which the best moral argument is. Thus, the right answer is that there is no right answer. iii) The reliability of the empirical assumptions Some limits to rationality are also observable on what concerns the determination of the reliability of the empirical assumptions relating to the importance of principles. The importance can be said to depend on its efficiency, speed, probability, reach and duration. The limits of rationality are related to several factors. First, it is difficult to determine the reliability of the empirical assumptions from all these perspectives. The empirical knowledge of the judge is limited. Sometimes he does not know the right value of each one of these variables. Second, the combination of these variables is a highly complex affair. What should the reliability of an empirical assumption whose little efficiency is plausible (½), its high speed is not evidently false (¼), its high probability is reliable (1), its great reach is plausible (½) and its long duration is reliable (1)? And, correlatively, will this reliability be greater, if the same variables have the same values but in a different order? All this shows that, in the weight formula, there are diverse rationality limits that leave a margin of discretion to judges. However, it is precisely within this margin of discretion in which precedents matter and play an important role. The weight formula is an empty structure whose three variables that are to be filled with factual and normative premises. Factual and normative premises indicate whether a specific principle should have more or less relative weight in abstract or in a concrete case and whether a specific empirical assumption is more or less reliable. Fortunately, judges do not make decisions about weights of principles and reliability of empirical premises in a tabula rasa. They do it in the middle of a network of precedents. For instance, recent developments in the case law of civil law jurisdictions lead to the conclusion that the weight of the principle of equality should be greater when legal norms establish different legal treatment to individuals on the basis of their gender or sexual orientation. In the same way, the case law of European Continental Constitutional Courts of the last twenty years makes it explicit that the weight of the freedom of expression should be greater in cases in which its exercise is linked to the democratic participation of the political system. Consistent decisions made by Courts create a network of precedents. This network constraints the margin of discretion that the weight formula opens for the judge to the extent that, if a judge would like to depart from a network of precedent, rationality requires from him or her to justify its choice. In this way, precedents become necessary for the rationality of balancing. Precedents allow principles to be applied in a consistent and coherent manner, with the result that balancing becomes not only more rational but also predictable.

THOMAS BUSTAMANTE1 FINDING ANALOGIES BETWEEN CASES: ON ROBERT ALEXY’S THIRD BASIC OPERATION IN THE APPLICATION OF LAW A. INTRODUCTION The point of this essay is to discuss in a critical way one of Robert Alexy’s recent developments to his theory of legal argumentation. While in his earlier writings Alexy distinguished only two basic operations in the application of law, subsumption and balancing, he has recently changed this theory to claim that legal analogy, or comparison of cases, also figures as a third basic operation which shares the same general features of the former two. My purpose, therefore, will be to answer whether Alexy is right when he holds that the comparison of cases can stand besides balancing and subsumption as a basic operation in the application of law. B. THE TWO ORIGINAL BASIC OPERATIONS

IN THE

APPLICATION

OF

LAW

Alexy has held since his earlier writings that the distinction between rules, as definitive requirements, and principles, as optimization requirements, was correlative to a differentiation between the basic operations of subsumption and balancing. Since rules have a binary structure and contain fixed points which require one to do exactly what they say, they are norms which “are always either fulfilled or not.”2 Every legal rule can be applied, hence, according to a deductive structure in which a statement of the facts of the case can be subsumed under its universal formulation. In order to illustrate this basic operation, Alexy has proposed the following scheme: 1) (x) (Tx Ο ORx) 2) (x) (M1x Ο Tx) 3) (x) (M2x Ο M1x) . . . (n+2) (Sx Ο Mnx) (n+3) Sa (n+4) ORa This scheme was referred to by Alexy as the “Subsumption Formula”. It is meant to exhibit all the premises comprised in subsumption: (1) is a norm which attaches a consequence R to all class of individuals who share the predicate T; (2) – (n+2) are 1 2

The author would like to thank Carlos Bernal Pulido and Lara Pratt for the very helpful comments on this chapter. Robert Alexy, A Theory of Constitutional Rigths, trans. Julian Rivers (Oxford University Press: Oxford, 2002), 48.

60

Thomas Bustamante

semantic rules “connecting the concept used to give expression to the antecedent condition of the norm (T) with the concept used to describe the case (S)”; (n+3) is “the description of the case”; and (n+4) is the legal judgment expressing the solution to the legal problem.3 For Alexy, the subsumption formula has three basic features: it is “formal, necessary, and specific.”4 To say that it is necessary means that it must be employed “in all cases in which legal rules are to be applied”, and its specific character stems from that fact that it works according to the rules of logic, which are a specific kind of rule.5 Its formal character, in turn, means that it says nothing about how the rule should be interpreted or how the facts should be constructed, which implies that the scheme “stands in need of saturation by means of substantive arguments.”6 When it comes to an argumentation based on legal principles, however, the subsumption formula is not enough to express all the steps which need to be taken in the application of law. In Alexy’s theory of fundamental legal rights, principles are defined as optimization requirements which “can be satisfied to varying degrees.”7 As such, they “require that something be realized in the greatest extent possible given the legal and factual possibilities.”8 Since the legal possibilities can only be determined with regards to opposing principles, whose scopes need to be contrasted, this type of norm stands in need of balancing. For Alexy, there are at least three variables which need to be considered in the basic operation of balancing: (i) the degrees of interference in one principle and of satisfaction of the other; (ii) the abstract weight of each of the principles in opposition; and (iii) the reliability of the empirical premises relied upon to establish the degrees of interference in Pi and of satisfaction of Pj.9 To express the relationship between these variables, Alexy proposes the Weight Formula, which works as a basic scheme for the operation of balancing and runs as follows: Ii · Wi · Ri Wi,j = Ij · Wj · Rj Although it might be preferable to avoid a long quotation, the easiest way to explain this formula is to transcribe Alexy’s own words on it, which are very clear and concise: “This formula represents the core of a complex argument-structure. In standard cases, where only two principles are involved, balancing begins with the subsumption of the case under two competing principles (Pi, Pj), and continues with an assignment of values, first, to the intensity of interferences (Ii, Ij) with Pi and Pj, second, to the abstract weights (Wi, Wj) of both principles, and, third, to the degree of reliability of the empirical assumption (Ri, Rj) respecting what the 3 4 5 6 7 8 9

Robert Alexy, ‘On Balancing and Subsumption. A Structural Comparison’ (2003) 16 Ratio Juris, 433–49, at 434. Robert Alexy, ‘Two or Three?’, in: On the Nature of Legal Principles, ed. Martin Borowski, ARSPBeiheft 119 (Franz Steiner: Stuttgart, 2010), 9–18, at 10. Ibid., 10. Ibid., 10. Alexy, A Theory of Constitutional Rights (n. 2), 47–8. Ibid., 47. Alexy, ‘On Balancing and Subsumption’ (n. 3), 436–440 and 443–8.

61

Finding Analogies between Cases

measure in question means for the non-realization of Pi and the realization of Pj. Once numbers are assigned to these variables, calculation of the concrete weight (Wi, j) is no more difficult than deduction, once the class of premises is complete.”10

Once the concrete weight (Wi,j) is determined, the argument goes, one can demonstrate that under the circumstances C one of the principles (Pi) takes precedence over the other (Pj). At this point one has reached a new rule which has as its protasis C and as its apodosis the legal consequence (Q) of the prevailing principle (Pi).11 We can see, thus, that the basic operation of balancing is a process by means of which one creates an individual (Kelsen12) or derivative (Alexy13) rule upon which the case C can be subsumed. The basic operation of subsumption appears both at the beginning and at the end of the process of balancing. That does not mean, however, that balancing cannot be classified as a basic operation according to Alexy’s three criteria to identify it, since, first, it is specific, for it works according to a specific kind of rules (the rules of arithmetic), second, it is necessary, since “it must be employed in all cases in which legal principles are to be applied”, and, third, it is formal, since it can be connected to all other sorts of arguments with a view to determine the values to be ascribed to each of the variables of the Weight Formula.14 C. THE THIRD BASIC OPERATION: ANALOGY

OR

COMPARISON

OF

CASES

The idea of a third basic operation first appears in an article in which Alexy analyses Arthur Kaufmann’s theory of law-making, where he suggests the following schemes to represent the comparison of cases: “A1: In every case ci, each case cj may be adduced with the argument that ci shares with cj the features F j1, … , F jn, and that ci, for that reason and because the rule F j1, … , F jn Ο Q is valid, ought to be treated, as cj, to the effect that Q. A2: In every case in which an argument for the form A1 is put forward, it may be claimed that ci is distinguished by the features F i1, … , F in from cj, and that ci, for that reason and because the rule F i1, … , F in Ο ɑ Q is valid, ought not, in contradistinction to cj, to be treated to the effect that Q.”15

This formula has been heavily criticized by Bäcker16 and Bro˙zek,17 who argued, amongst other things, that Alexy’s comparison schemes presuppose the existence of a rule, either F j1, … , F jn Ο Q or F i1, … , F in Ο ɑ Q, which “applies directly and 10 11 12

Alexy, ‘Two or Three?’ (n. 4), 11. Ibid., 11. Hans Kelsen, Pure Theory of Law, trans. Max Knight (University of California Press: Berkeley, 1967), 269. 13 Alexy, A Theory of Constitutional Rigths (n. 2), 33, 54. 14 Alexy, ‘Two or Three?’ (n. 4), 10–1. 15 Robert Alexy, ‘Arthur Kaufmanns Theorie der Rechtsgewinnung’, in: Verantwortetes Recht, ed. Ulfrid Neumann, Wingrid Hassemer, and Ulrich Schroth (Franz Steiner: Stuttgart, 2005), 47–66, at 65–66; Alexy, ‘Two or Three?’ (n. 4), 9. 16 Carsten Bäcker, Begründen und Entscheiden (Nomos: Baden-Baden, 2008), 298, cf. Alexy, ‘Two or Three?’ (n.4), 9. 17 Bartosz Bro˙zek, ‘Analogy in Legal Discourse’ (2008) 94 Archiv für Rechts-und Sozialphilosophie, 188–201, at 198–200.

62

Thomas Bustamante

explicitly to both cases”, in such a way that “there is no need for analogical reasoning here, as we are not in a situation in which there is a case for which there exists no relevant legal rule.”18 Rather than create a new rule by analogical reasoning, the scheme would merely subsume cases ci and cj under another rule. This objection was strong enough to lead Alexy to reformulate his Analogy Scheme, replacing the clause “because the rule F j 1, …, F jn Ο Q is valid” in A1 and “because the rule F i1, … , F in Ο ɑ Q is valid” in A2 by “because there are reasons for the rule F j1, …, F jn Ο Q” and “because there are reasons for the rule F i1, … , F in Ο ɑ Q.” Thus, the revised Analogy Scheme now reads as follows: “A1: In every case ci, each case cj may be adduced with the argument that ci shares with cj the features F j1, …, F jn, and that ci, for that reason and because there are reasons for the rule F j1, …, F jn Ο Q, ought to be treated, as cj, to the effect that Q. A2: In each case in which an argument of the form A1 is put forward, two counter-claims may be raised: A2.1: It may be claimed that ci is dintinguished by the features F i1, …, F in from cj, and that ci, for that reason and because there are reasons for the rule F i1, …, F in Ο ɑ Q, ought to be treated, in contradistinction to cj, to the effect that ɑ Q. A2.2: It may be claimed that ci shares with ck the features F k1, …, F kn, and that ci, for that reason and because there are reasons for the rule F k1, …, F kn Ο ɑ Q, ought to be treated, as ck, to the effect that ɑ Q.”19

In Alexy’s view, the Analogy Scheme, like the Subsumption Formula and the Weight Formula, has the three distinctive features which are necessary to qualify it as a valid scheme for a basic operation in the application of law: “it is formal, necessary, and specific.”20 Its formal character lies in the fact that the formula says nothing about “which features are to be classified as relevant”, and equally nothing on “the question of whether the argument according to A1 or the argument according to A2 prevails.”21 The necessity of the Analogy Scheme, in turn, “stems from the fact that it is not possible to refer in a rational way to other cases without using the scheme.”22 And finally, its specific character stems from the “dialectic of reference to features of other cases.”23 According to Alexy, this is sufficient to qualify analogy or comparison as an autonomous basic operation in the application of law. D. A CRITICAL ASSESSMENT

OF

ALEXY’S ACCOUNT

ON

LEGAL ANALOGY

The first thing to be noticed in Alexy’s Analogy Scheme is that it is, in fact, a formally correct structure to explain the way jurists raise claims and put forward arguments on the basis of analogical reasoning. If one is to criticise Alexy’s views on analogy, this shall not be done on the basis of the formal aspects of the Analogy 18 19 20 21 22 23

Ibid., 199. Alexy, ‘Two or Three?’ (n. 4), 17. Ibid., 17–8. Ibid., 17. Ibid., 18. Ibid., 18.

Finding Analogies between Cases

63

Scheme or of the lack of accuracy in its description of precedent-based legal reasoning, but rather on the basis of the unsoundness of the claim that such scheme is actually “necessary” and of the inability of the scheme to contribute to the justification of legal decisions. Furthermore, one can successfully argue that Alexy’s conception of “basic operations” is ambiguous and needs to be redefined in a way that excludes analogy between cases from the scope of the notion of a “basic operation in the application of law.” In the following sub-sections, I attempt to explain these objections. 1. OBJECTIONS TO THE NECESSITY AND RELEVANCE OF THE ANALOGY SCHEME One of the great contributions Alexy has made to contemporary jurisprudence was to clarify the relations between rules and principles in legislation and adjudication in a way that was not entirely evident in, for instance, Dworkin’s earlier writings about legal principles. With his so-called Law of Competing Principles (LCP), Alexy demonstrated that principles are the fundamental reasons which justify legal rules and operate as standards on the basis of which every rule of the system should be interpreted. In addition, he was able to unveil that a rule is constructed as a result of a balancing of principles. In fact, the LCP shows that from every balancing of principles one necessarily reaches a derivative norm which has the character of a rule, as it is visible in its canonical formulation: “(LCP): If principle P1 takes precedence over principle P2 in circumstances C: (P1 P P2) C, and if P1 gives rise to legal consequences Q in circumstances C, then a valid rule applies which has C as its protasis and Q as its apodosis: C Ο Q.”24

This meta-rule demonstrates that balancing principles always gives rise to a rule which has as its apodosis the same consequences of the principle that takes precedence in the light of the concrete case. Hence, for every rule one can think of, there is a principle as its leading justifying reason. Each legal rule “may be presented as a result of a weighing and balancing of principles, performed by the lawgiver.”25 When it comes to a legal analogy there is no other way to justify it other than through a balancing of the principles which were considered in the precedent case. In the case of precedent-based reasoning, the rationes decidendi should not be regarded as rules which may be detached from the principles which figure in their justifications, as it happens in a system of binding precedents inspired by legal positivism,26 but as “rulings on law stated as necessary parts of justifications of deci-

24 25

Alexy, A Theory of Constitutional Rigths (n. 2), 54. Aleksander Peczenik. ‘Jumps and Logic in the Law: What can one expect from Logical Models of Legal Argumentation?’(1996) 4 Artificial Intelligence and Law 297–329, at 298. 26 This is, in fact, the traditional way that precedents figure in the discourse of British jurisprudence, as one can see in the works of Goodhart, who holds that “the reason which the judge gives for his decision is never the binding part of the precedent” (Arthur L. Goodhart, ‘Determining the Ratio Decidendi of a Case’ in: Essays in Jurisprudence and the Common Law (Cambridge University Press: Cambridge, 1931), 2), or Cross, who thinks that precedents may be authoritative even in cases in which “no reasons are given for the decision” (Rupert Cross; J. W. Harris. Precedent in English Law. 4th ed. Clarendon: Oxford, 1991), p. 47).

64

Thomas Bustamante

sions relatively to cases and the arguments put by given parties.”27 Once this connection between rules, reasons, and the principles which underlie them is recovered, it becomes clear that no rule can be analogically extended without a balancing of principles. As Alexy recognized while he was speaking of his Analogy Scheme, “the features in A1 and A2 acquire their relevance by virtue of the fact that they are related to reasons undergirding the rules that contain them as antecedents”, and “these reasons normally have the character of principles.”28 All this seems to me to be very straightforward and correct, and this is, in fact, one of the merits of Alexy’s theory of norms. Yet this explanation of the relationship between rules and principles loses much of its practical significance if the new formal scheme for the basic operation of analogy is adopted. In fact, the so-called Analogy Scheme seems to fall back into the traditional way to explain analogies, which focused entirely on the “similarities between cases” or “features” and was unaware of the fact that best way to justify analogies is by means of an argument based on the principles which underlie the precedent rule and the new rule to be ascribed to the instant case.29 Whenever an old rule is to be extended to cover situations which were not originally subsumed into its operative conditions, this interpretive process has to be done on the basis of the principles which justify it, for, as MacCormick correctly points out, “the relevance of analogy depends on the broader principle of which each of the cases can be shown to be an instantiation.”30 As I had a chance to express in a short essay published in the same volume as Alexy’s new paper, the notion of “similarities between facts” is not enough to face the challenges that appear in analogical reasoning, since the value judgments required to establish the “similarities” that allow an inference in the Analogy Scheme remain obscure and cannot be objectively controlled.31 As I wrote in my previous essay, “Instead of comparing the facts of the precedent with the facts of the current case it is better to compare the rules that are presupposed in the reasoning that leads to the decision of each of the two situations. We should look not to the ‘similarity of facts’, but to the ‘similarity of rules’. In fact, when arguing by analogy one requires ‘a base line or point from which we can start and by reference to which we can compare and contrast any particular problem case or situation’ [Cfr. Bankowski]. This ‘baseline’ for the analogy between cases is much more visible when we compare rules than when we compare brute facts: rules are immediately linked to legal principles,

27 28 29

30 31

Neil MacCormick, ‘Why Cases Have Rationes and What These Are’, in: Precedent in Law, ed. Laurence Goldstein (Clarendon: Oxford, 1987), 155–181, at 171. Alexy, ‘Two or Three?’ (n. 4), 14. Indeed, the Analogy Scheme holds that analogies are properly justified “because there are reasons” for a rule, more specifically, a general rule that covers both ci and cj (Ibid., 17), but it neither makes clear what kind of reasons these are nor expounds how they are to be weighed. This reference to ‘reasons’ is so vague that it does nothing more than stating that the process of comparison should be rational, which is in fact true, but trivial. Since these reasons are unspecified, it can be said that analogical reasoning in the Analogy Scheme is entirely driven by the comparison between the features of the cases. Neil MacCormick, Rhetoric and the Rule of Law (Oxford University Press, Oxford, 2005), 207. Thomas Bustamante, ‘Principles, Precedents and their Interplay in Legal Argumentation: How to Justify Analogies Between Cases’, in: On the Nature of Legal Principles, ed. Martin Borowski, ARSP-Beiheft 119 (Franz Steiner: Stuttgart, 2010), 63–77, at 71.

Finding Analogies between Cases

65

while facts are not. Principles can bridge the gap between rules, and nothing seems to be able to do the same job with mere facts [or, as I now add, with any ‘feature’ of the facts of the case].32

Hence, if I am right in this, when Alexy formulates his Analogy Scheme without incorporating the steps needed to balance the principles to be compared, he is hiding away one of his most relevant contributions to the theory of norms, which is the explanation of the precise relationship between principles and rules. By missing the interactions between these norms, the Analogy Scheme fails to explain how one is to choose between the alternative solutions that may be available to a legal problem. To be sure, that scheme would be much more useful to a sceptic who denies the existence of legal principles than to a jurist who decides to apply Alexy’s own theory of norms to decide a case by means of an analogical construction. For the sceptic the scheme would suffice, although she might struggle to justify the ‘sameness’ of the cases or the ‘relevance’ of the features of the situations that must be compared; but for the follower of Alexy the scheme would lead us only half-way to the conclusion, since she would not be able to avoid balancing principles in order to arrive at the rule that has to be issued for the case. It makes more sense, thus, and it is also more coherent with Alexy’s previous theoretical commitments (which were never dismissed in the papers in which he advocates a third ‘basic operation’), to endorse Bro˙zek’s ‘Partial Reducibility Thesis’, according to which one shall weigh the “principles standing behind the rules that govern both similar cases”33 with a view to justify a conclusion by legal analogy. It should be clear, however, that to endorse the Reducibility Thesis a follower of Alexy’s theory of norms does not need to accept all the criticisms that Bro˙zek directs towards his theory, including that which invites us to abandon the LCP on the assumption that it is ‘unnecessary’ if a system of defeasible logic is adopted.34 As I wrote in a previous attempt to analyse the structure of analogical argumentation, “it is precisely the Law of Balancing that makes the ‘Partial Reducibility Thesis’ plausible, for it is the fact that the collision of principles leads to the formulation of a rule that makes it possible to determine, with a certain degree of objectivity, the principles that stand behind the rules which need to be compared in order to validate the argument by analogy.”35 This argument provides, therefore, a good reason for us to claim that the Analogy Scheme, at least in the current formulation, is not actually relevant for a jurist who is up to the task of justifying analogies, for it works under the assumption that legal analogy is only a judgment about the relevant features of different cases, rather than a judgment about the relevant principles contributing to the rules assumed for the solution of these cases. Furthermore, I do not think that Alexy is right when he argues that the Analogy Scheme is “necessary” in the sense that “it is not possible to refer in a rational way to other cases without using the scheme.”36 If he were claiming that it is not possible to refer to other cases without using a formal scheme, rather than the scheme that he 32 33 34

Ibid., 71. Bro˙zek, ‘Analogy in Legal Discourse’ (note 17), 195. Bartosz Bro˙zek, Rationality and Discourse – Towards a Normative Model of Applying Law (Oficyna: Warsaw, 2007), 150. 35 Bustamante, ‘Principles, Precedents and their Interplay in Legal Argumentation’ (note 31), p. 37. 36 Alexy, ‘Two or Three’ (note 4), 18.

66

Thomas Bustamante

advocates, it would be quite easy to agree with him. None the less, this is obviously not the case, and this weak claim (the latter) would not be enough to vindicate the view that the Analogy Scheme should be adopted. Hence, in order to show that Alexy’s argument is implausible one needs merely to demonstrate that it is possible to imagine any alternative scheme to compare the features of the cases that need to be contrasted in analogical reasoning. In the remaining part of this section, I attempt to provide an alternative scheme which is, nevertheless, entirely based on Alexy’s ideas on the inter-relations between rules, principles and cases, but which takes into consideration Laura Clérico’s developments of the LCP into the direction of a “model of balancing oriented toward rules”37 which takes very seriously the “universalization of the results of a balancing operation”38 and the balancing exercise required to move from one rule to the other. In fact, while Alexy’s Analogy Scheme expounds only the structure of the judgment about the consequences of the relevant features of the cases, it is equally feasible to adopt an alternative scheme which describes the structure of the balancing exercise that is needed to vindicate the universal norms that stem from the solution of the case. In effect, the move from Normi, which stems from Ci, to Normj, which stems from Cj, can be described according to the following argument form: Ni: (x) (Ci x Ο ORx) (P1 P P2) Ci (P1 P P2) Cj Nj: (x) (Cj x Ο ORx).39 This formal scheme may be expanded so the norms Ni and Nj include more predicates in their operative conditions, in which case it would read as: Ni: (x) ([C 1i ࣿ C 2i ࣿ … ࣿ C ni ]x Ο ORx) (P1 P P2) C 1i ࣿ C 2i ࣿ … ࣿ C ni (P1 P P2) C 1j ࣿ C 2j ࣿ … ࣿ C nj Nj: (x) ([C 1j ࣿ C 2j ࣿ … ࣿ C nj ]x Ο ORx).40 37

Laura Clérico, El examen de proporcionalidad en el Derecho Constitucional (Eudeba: Buenos Aires, 2009), 180. 38 Ibid., p. 180. 39 If one concludes that the ‘claim to generalization’ (Clerico, La estructura… [n. 36], 184) which is raised to justify an analogical extension of Ni cannot be vindicated, then a negative version of the scheme applies, which runs as follows: (1) Ni: (x) (Cix Ο ORx) (2) (P1 P P2) Ci (3) (P2 P P1) Cj (4) Nj: (x) (Cjx Ο ɑ ORx). 40 As it happens in the simpler version of the scheme, one may need a negative version, which can be formulized thus: (1) Ni: (x) ([C1i ࣿ C2i ࣿ … ࣿ Cni]x Ο ORx) (2) (P1 P P2) C1i ࣿ C2i ࣿ … ࣿ Cni (3) (P2 P P1) C1j ࣿ C2j ࣿ … ࣿ Cnj (4) Nj: (x) ([C1j ࣿ C2j ࣿ … ࣿ Cnj]x Οɑ ORx).

Finding Analogies between Cases

67

This demonstrates that it is not very difficult to develop an alternative scheme for comparison of cases. Furthermore, we can see that this argument form is more interesting than Alexy’s Analogy Scheme because it shows the structure not only of the rules and the features of the cases, but also of the priority relations between the principles that are used to justify these rules. Moreover, there is yet another advantage in this scheme, since it illustrates more explicitly that in analogical legal reasoning we are comparing rules, not merely unspecified features of the cases. Here we can see the that Clérico is right when she states that “a ‘rule resulting from balancing’ (‘regla-resultado de la ponderación’) is [considered as] universalizable if it is applicable to the solution of a new case, and if the operative conditions of the ‘rule resulting from balancing’ and the circumstances of the (new) particular case can be justified as equal or similar.”41 This connects in a reasonable way the “feature of the cases” to the justificatory and universalistic dimensions of legal argumentation, which are characteristic of the rules. One may be wondering why it is important to stress that any scheme to represent analogies shall be based on the comparison of rules, rather than features of cases. Why does it make a difference, it may be asked, to come up with an Analogy Scheme which compares cases on the basis of their “relevant features” or on the basis of the “universal rules” that follow from the decisions? This question may be answered in different ways depending on how one applies the operation of balancing. If the point of our inquiry is merely descriptive, then it does not make any difference at all to focus on “features of the cases” or on the “rules” that stem from precedent decisions. From the analytical point of view, it is irrelevant if we concentrate our analysis on the rules (i.e. on the major premise of the legal syllogism) or on cases (i.e. on the minor premise of the legal syllogism). For a rule to be instantiated in a case it must necessarily be true that there is semantic equivalence between the predicates which figure in the operative conditions of the norm and in the description of circumstances of the case. As Wróblewski explains, the uses of precedent “oriented towards facts” and precedent “oriented towards rules” are interconnected, “since rules always rule upon facts, and facts in legal discourse cannot be understood apart from rules.”42 So it can be said, for instance, that “distinguishing, overruling and fact-adjusting are functionally equivalent”, since “they serve as alternative means to the same end, the result desired by the judge.”43 The only differences between these alternative approaches to the ratio decidendi of a precedent are the type of justification that the judge provides for her decision and the influence of this decision in future cases. None the less, if the point of the inquiry is normative – for instance, when one is using the structure for the purpose of providing a justification, rather than a description, of a case – then it is more advisable to concentrate the comparison on rules than on the features of the case. Apart from the argument that one can see better the connection between cases and the principles which justify the decision, 41 42

Ibid., p. 182. Jerzy Wróblewski, ‘Il Precedente nel Sistemi di Civil Law’, trans. Riccardo Guastini, in: La Giurisprudenza per Massime e il Valore del Precedente, ed. G. Visintini (CEDAM: Padova, 1988), 25ff, at 29–30. 43 Svein Eng, ‘The Doctrine of Precedent in English and Norwegian Law: Some Common and Specific Features’ (2000) 39 Scandinavian Studies in Law, 275–324, at 316.

68

Thomas Bustamante

stated above,44 the model that describes precedents as rules should be adopted in order to avoid misunderstanding the universalist character of legal argumentation. While a comparison of rules is clearly directed towards the universalizability of the solution of cj, which is of key importance for the rational justification of a legal decision, a comparison of “cases” or “features” appears to provide a context that is very attractive for particularism, which does not take the principle of universalizability seriously enough. Although it is not necessarily the case that the user of a conception of analogy centred on the ‘similarities of features’ would embark on a particularistic approach to legal reasoning, she would find a context that is favourable to such an approach, since the Analogy Scheme would emphasize the comparison of the features of the cases and would say nothing about the character of the reasons that will be used to justify the decision. This is, indeed, another reason to stick with a ‘rule-comparison’ description of legal analogy. In this sense, it should be reminded that to justify a decision means to construct it in a rule-based way, as MacCormick argues when he states: “To justify requires one to put one’s claim, defence, or decisions on the footing that because the facts F1, F2, … Fn, the judgment j ought to be pronounced”, and that because-clause “requires a commitment to universal.”45 If one describes analogies without using the language of rules, one is exposed to the risk of losing track of the fact that justificatory discourses always claim universality, as Alexy has convinced us in his Theory of Legal Argumentation, where it is stated that any legal judgment “must follow logically from at least one universal norm together with further statements.”46

44 45 46

See supra, text above note 31. MacCormick, ‘Why Cases have Rationes and What these are’ (n. 27), 162. Robert Alexy, A Theory of Legal Argumentation: The Theory of Rational Discourse a Theory of Legal Justification, trans. Ruth Adler and Neil MacCormick (Oxford University Press: Oxford, 1989), rule n. J.2.2., at 223. The risk of particularism has always been considered a problem by Alexy, who warns us, for instance, against the risk of misunderstandings that may arise if Habermas’ distinction between “application discourses” and “justification discourses” is accepted. The idea of an application discourse is “easy to misunderstand” because it “houses the danger of a nonuniversalist practice of decision making” (Robert Alexy, ‘Jürgen Habermas’s Theory of Legal Discourse’ (1996) 17 Cardozo Law Review, 1027–1034, at 1032–3). As Alexy persuasively argues, “a harmonization as unification of the law in the sense of a universalist practice of decision making is only possible if, in deciding individual cases, rules are formed which can win the power of precedence” (ibid, 1033). This is, hence, a good argument that can be deployed to substitute a model based on the “features of the cases” by a model oriented toward the justification of universal rules, since the latter is better equipped to defend the most important claims of Alexy’s general theory of legal discourse. I must stress, however, that I am not accusing Alexy of advocating particularism. In fact, I am well aware that his Analogy Scheme even says that ci and cj ought to be seen as similar “because there are reasons for a rule” which prescribes that the features that they share lead to a common consequence. Yet, I argue that the Analogy Scheme still contains an invitation to focus on the “features of the situation”, which seem to drive the reasoning process, and that’s why it may lead to misunderstandings (see, also, supra, note 27).

Finding Analogies between Cases

69

2. A NOTE ON THE CONCEPT OF A ‘BASIC OPERATION IN THE APPLICATION OF LAW’ The argument developed in the previous section provides an objection to the formal scheme that Alexy has recently offered to describe the structure of analogy or comparison in legal argumentation. However, even if the objection is sound Alexy may still be right in the assertion that analogy figures as a third basic operation in the application of law. To decide whether this is the case, one no longer needs to consider the formal properties of the argument by analogy, but rather to concentrate on the meaning of the notion of a “basic operation in the application of law.” To begin, we should notice that Alexy does not define in precise terms what he means by a “basic operation in the application of law”, although he tells us that it must be “formal”, “specific” and “necessary.” There is a great deal of ambiguity with regards to the contents of the notion of a “basic operation” and with regards to its object. From the point of view of modern hermeneutics, the notions of understanding, interpreting and applying an object, in our case a norm, are seen as “an integral part of the hermeneutical process.”47 Nevertheless, from the analytical point of view we can distinguish the “psychological process of judicial decision-making” and the “justification” of a legal decision,48 which can be studied separately for the purposes of the construction of legal knowledge and evaluation of legal arguments. Hence, although at the hermeneutical level it may be true that the activities of interpretation and application of law are inseparable, these steps should be neatly differenced for the purpose of the analysis of the justification of legal judgments. With regards to the discourses of justification of legal decisions, the notion of application of law in a broad sense can be divided into four groups of “fractional decisions” that need to be justified as separate steps in legal argumentation. These fractional decisions are: (i) decisions of validity, in which one identifies the valid legal norms that are relevant for the decision at stake; (ii) decisions of interpretation, which concern meaning of the norms applicable to the case; (iii) decisions of evidence, which concern the material facts considered to be proved; and (iv) decisions of choice of the consequences, which concern the establishment of the legal consequences appropriate to the proven facts.49 The last group of decisions (decisions of choice of the consequences) can also be defined as “application of law in a narrow sense”, which refers to the application of a particular norm. In its sense, to say that a norm is applied means “that in decisions in which its conditioning facts are assumed to exist, it forms an integral part of the reasoning underlying the judgment and has been therefore one of the decisive factors determining the conclusion at which the court has arrived.”50 There are thus two alternatives open for us: either we accept that Alexy’s basic operations in the application of law take place in the fourth group of fractional deci47

Hans-Georg Gadamer, Truth and Method, trans. Joel Weinsheimer and Donalg G. Marshall, 2nd ed. (Continuum: London, 2004), 307. 48 Jerzy Wróblewski, ‘Legal Syllogism and Rationality of Judicial Decision’ (1974) 3 Rechtstheorie 1, 33–46, at 35 ff. 49 Jerzy Wróblewski, The Judicial Application of Law, trans. Zenon Bankowski (Kluwer: Dordrecht, 1992), 11. 50 Alf Ross, On Law and Justice (The Lawbook Exgange: Clark, NJ, 2004), 42.

70

Thomas Bustamante

sions (or the so-called “application of law in the narrow sense”), which takes place when all the difficulties concerning “validity, interpretational and evidence decisions” are solved;51 or we have to claim that when we speak of basic operations in the “application of law,” this expression must be understood in its wider sense, which comprises all the four groups of fractional decisions stated above. Bro˙zek, for instance, accepts the first alternative and holds that the notion of application of law is used by Alexy as consisting “on a decision [on] which of the valid and interpreted legal norms to apply in the given case.”52 This alternative seems to be attractive when we speak of the basic operations of subsumption and balancing. To subsume a fact under the description of a rule one must have already asserted the validity and the final interpretation of the norm in analysis, as well as decided that this norm has the structure of a rule. By the same token, to balance rights or principles one must have decided, first, that the norms applicable to the case are valid, second, that they have the structure of principles, and third, that the facts of the case fall within the scope of application these norms. As we can see, the operations of balancing and subsumption can easily be comprised as formal operations which are necessary to determine the consequences of (previously interpreted and indisputably valid) norms and (previously constructed) facts in a given case. Nevertheless, the same cannot be said about legal analogy. In fact, analogy or comparison is not about the application of a legal norm in strict sense or a decision at the final stage of the process of application of law, but rather about the creation of a new rule which cannot be considered a mere act of determining the consequences of the rule which served as a point of departure for the argumentation. If we think of analogy as a form of argument by means of which one applies the consequences of a norm to cases that cannot be subsumed in its operative conditions, what seems to be a necessary implication of the reference to “analogous” or “similar” cases, then it is analytically true that there is no legal analogy when the “conditioning facts” of that norms “are assumed to exist.”53 We can talk of analogy only when, after there is no more disagreement about the contents of the rule R (properly interpreted), we cannot resolve the case by simply subsuming it under R. We can see, therefore, that the very concept of “application of law in the narrow sense” by definition excludes the possibility to include analogy within its range of basic operations. Hence, we are only left with the second alternative: the locution “application of law” is understood in its wider sense, which includes the four classes of fractional decisions described by Wróblewski. Here, the creation of a new rule by comparison or analogy is viewed as a decision about the validity of a new rule, which is developed on the basis of the principles underlying the rule that served as a paradigm for the analogical reasoning. None the less, although this alternative is not impossible, it is in my opinion unattractive because it brings nothing but confusion into the theory of law. If we understand the locution “application of law” in its wider sense, then not only analogy, but all other argument forms that can be employed to reconstruct the fractional decisions of a judge would equally be classified as a basic operation in the 51 52 53

Bro˙zek, ‘Analogy in Legal Discourse’ (n. 17), 38. Ibid., 38. Ross, On Law and Justice (n. 50), 42.

Finding Analogies between Cases

71

application of law. In effect, one can think of many operations in legal reasoning that are formal, necessary and specific, and all of them can equally be classified as operations of “application of law” in the broad sense. To name some of them, these would include decisions on the proof of facts and on the normative qualification of these facts (which are both decisions of evidence); on the formal and material validity, derogation, constitutionality, analogical extension, reduction etc. (which are decisions of validity); and decisions on which canon of interpretation is to be applied and on which of these canons must prevail in case of conflicts or oppositions (which are decisions of interpretation). In fact, these decisions can all be classified as “formal”, “necessary” and “specific”, and can equally be formalized by an analytical scheme (or argument form) that can be used in legal justification. None the less, nothing is gained if we call all these fractional decisions “basic operations in the application of law”, for this very notion would be banalised and would make reference to decisions which have nothing to do with the application of norms as we normally understand it in legal discourse (where the expression is usually used in the narrow sense). To escape these unattractive consequences, the best alternative seems to be to claim that, at the end of the day, analogy is not a “basic operation in the application of law”, but rather one of the forms of legal reasoning by means of which one establishes a new norm that was not previously settled in the legal system. To put it very simply, it is a form of judicial construction or development of the law that is, at the end, based on a balancing or principles. E. FINAL CONSIDERATIONS We can conclude, thus, that Alexy’s theory of norms is still a very powerful device for the justification of analogies in legal reasoning. The LCP makes it very clear that there is a connection between subsumption, balancing and analogy or comparison of cases. This connection shall not be missed by those who intend to justify in a satisfactory way the making of a novel decision by analogy with precedents or statutory rules. None the less, it is perfectly possible, and even preferable, to do so without qualifying analogy as a basic operation in the application of law. The very idea of a basic operation in the application of law was useful, in my opinion, because it made clear the structural differences between rules and principles, and the modes of application of these norms, but nothing is gained if we extend this notion to a form of reasoning which is not a case of “application of law in a narrow sense.” “Application of law”, therefore, shall be interpreted for a user of Alexy’s theory of norms as equivalent to “application of norms”, for this is the only way to avoid confusion and obscurity, and to interpret his theory at its best light.

II ON

THE

STRENGTH

AND THE

USES

OF

PRECEDENTS

LARRY ALEXANDER PRECEDENTIAL CONSTRAINT, ITS SCOPE AND STRENGTH: A BRIEF SURVEY OF THE POSSIBILITIES AND THEIR MERITS INTRODUCTION In this essay I examine how broadly precedents constrain (their scope) and how strongly they do so (their strength). With respect to scope, I look at three models of precedent and conclude that only one of these is both constraining and coherent. With respect to strength, I affiliate overruling precedents with the more general issue of legal transitions and conclude that no algorithm for when to overrule is available. Overruling precedents turns out to be only an instance of the more general problem of the rationality of rule following. Precedent in Anglo-American jurisprudence is understood as something that constrains present decision makers. (From here on I’ll refer to present decision makers as courts, though it should be kept in mind that precedents may be thought to constrain other decision makers besides courts.) The constraint that precedent is thought to impose is this: A present court that, in the absence of the precedent, would decide A v. B in favor of A is, because of the precedent, constrained to decide in favor of B. Precedential constraint has two aspects. The first is its scope: How many potential decisions are affected by the precedent? The second is its strength: Under what conditions may a precedent be overruled so that it no longer constrains? Most of my remarks will address the scope question, though I will have a few things to say about the strength question at the end. A. THREE MODELS

OF

PRECEDENTIAL CONSTRAINT

AND

ITS SCOPE

How is it that a decision by a court at T1 is supposed to affect a court at T2, so that the T2 court is constrained to decide its case differently from how it would have decided it in the absence of the T1 decision? One way the decision at T1 can affect the T2 case is by changing the facts on the ground. Suppose, for example, that in the T1 case of A v. B, the decision is in favor of A. And suppose, had the T2 court decided A v. B, it would have decided in favor of B. In other words, the T2 court believes the T1 court decided A v. B incorrectly. Still, now that A v. B has been (incorrectly) decided in favor of A, people, including A’, may have relied on that decision in arranging their affairs. If the T2 court decides A’ v. B’ in favor of B’ – as it would have in the absence of the earlier decision – A’ will suffer a loss because of his reliance on the decision in A v. B. On balance, then, the case of A’ v. B’ may now tip in favor of a decision for A’ because of the erroneous decision in the T1 case. In previous works I called this way in which an incorrect decision might constrain subsequent ones as the natural model of precedent.1 I labeled it “natural” to em1

See L. Alexander, ‘Constrained by Precedent’ (1989) 63 Southern California Law Review, 1; L. Al-

76

Larry Alexander

phasize the fact that the later court – the constrained court – is merely deciding its case as it naturally should, that is, by taking into consideration all relevant facts and then reaching a decision it deems correct in light of those facts. The precedent decision, or, more precisely, the behavior induced by it, is a fact, though it is just one fact among many that might be relevant. The natural model of precedential constraint is not the model Anglo-American courts employ. Although they may look at reliance on precedents in deciding whether to overrule precedents – something I take up in the discussion of precedential strength – they do not look at reliance in determining the scope of precedential constraint. In other words, Anglo-American courts view precedential constraint to be broader than the natural model would entail. A second model of precedential constraint is one that I have labeled the rule model of precedent.2 It can be easily described. On this model, when the precedent court decides A v. B, it announces and promulgates a rule – “if X, Y, and Z, then decide for A” – and subsequent courts are bound to follow this rule unless and until the precedent is overruled by a court or other body with the authority to do so. The rule model of precedent views a precedent court as a legislative body as well as an adjudicative one. The precedential constraint in A v. B is not located in the decision in favor of A but in the rule promulgated in the precedent that purports to justify that decision (“if X, Y, Z, then decide for A”). The rule promulgated is the precedent case’s “holding.” Everything else said by the court is “dicta” and has no binding effect. There are three principal objections to this model. One objection is that courts are not competent to or not authorized to legislate. A second objection is that it is often difficult or impossible to locate a “rule” in a court’s opinion. A third objection is that the rule model obliterates one feature that is thought to be present in a system of precedent, namely, the feature of distinguishing prior decisions as opposed to overruling them. For the later court either follows the rule laid down in the precedent case or else overrules or amends it (which is a partial overruling). Joseph Raz disagrees with this last point.3 He suggests that precedent rules can be distinguished without being overruled. According to Raz, a judge seeking to distinguish a precedent rule must restate the rule in a way that meets two conditions: the modified rule must be the precedent rule with some further condition added, and the modified rule must support the outcome of the precedent case. He illustrates with an example in which the precedent case involved facts a, b, c, d, and e, the result was X, and the opinion announced a rule “if A, B, and C, then X.” The new case involves facts a, b, c, d, and f, but not e. The court can distinguish the new case and announce a modified rule “if A, B, C, and E, then X,” or a modified rule, “if A, B, C, and not F, then X.” But it cannot announce a modified rule, “if A, B, C, and not D, then X,” because this rule does not support the outcome of the precedent case.

2 3

exander and E. Sherwin, E., The Rules of Rules: Morality, Rules and the Dilemmas of Law. (Durham: Duke University Press: Durham, 2001), chapter 8. See id. See J. Raz, The Authority of Law: Essays on Law and Morality (Oxford University Press, Oxford, 1979), 183–189.

Precedential Constraint, Its Scope and Strength

77

This constraint is illusory, however. Assume a precedent rule: “Wild animals in residential neighborhoods are nuisances.” And suppose it is announced in a case involving a pet bear. In a later case, Jerome is keeping a crocodile in his house, and the court, in response to a suit seeking to enjoin Jerome from doing so, and sympathetic to Jerome, announces its modification of the precedent rule: “Furry wild animals in residential neighborhoods are nuisances.” This rule may not be ideal, and it authorizes a result that seems contrary to the values the precedent rule was designed to protect; but it meets Raz’s two conditions: it is the precedent rule with a condition added, and it justifies the outcome of the precedent case. Nor do Raz’s conditions guarantee that the modified rule will be similar in effect to the precedent rule. The later court could announce a rule, “Wild animals that are three-year-old ocelots with one lame foot are nuisances,” without running afoul of the supposedly constraining conditions. But the pattern of future nuisance decisions under the rule will be radically different from the pattern one would have expected under the earlier court’s rule, “Wild animals are nuisances.” Precedent rules cannot, therefore, be “distinguished.” They can only be followed or overruled, either in whole or in part. So all three of the stated objections to the rule model of precedent are valid. Despite these objections, the rule model is the only model of precedent that can constrain a later court to decide a case in a way that it believes is incorrect at the time it decides it. (The natural model does not do this because it asks the later court to reach the correct all-things-considered decision in the case before it in light of the effects of the precedent decision.) However, in order to show that only the rule model of precedent can constrain, I need to put on the table a third model of precedent, one that seems to possess all the features that we associate with a system of precedent. Its only problem is that it is conceptually confused. This third model is what I have labeled the result model of precedent.4 On this model, the constraint imposed on later courts by the precedent decision comes from its having decided in favor of A in a case in which the facts were F1, F2, and F3. Suppose in the later case of A’ v. B’, the facts are F1, F2, F3, and F4. Suppose the later court believes that these facts in total weigh in favor of a decision for B’. However, F4 is a fact that weighs in favor of a decision for A’. Therefore, the later court believes that although in the absence of the A v. B precedent, it should decide the case in favor of B’, the facts weigh more heavily in favor of a decision for A’ than they did for A in A v. B. And because the precedent court decided for A on weaker grounds than exist for A’, A’ v. B’ is an a fortiori case for a decision for A’. In other words, if the precedent decision is to be followed, then A’ v. B’ must be decided in favor of A’ even though in the absence of A v. B, the decision would be in favor of B’. The result model of precedent appears to possess all of the features we commonly associate with a system of precedent. Unlike the natural model, the result model’s constraint does force subsequent courts to decide differently from how they would decide even after taking into account reliance on the precedent. In other words, the result model views precedent decisions as themselves constraining even in the absence of any reliance on other relevant effects. Unlike the rule model, the result model does not view precedent courts as legislating rules. And also unlike the 4

See note 1, supra.

78

Larry Alexander

rule model, the result model retains the distinction between overruling a precedent and distinguishing it. For a precedent can be distinguished by citing to some fact in the present case that did not obtain in the precedent case and that counts in favor of a decision in favor of the party analogous to the losing party in the precedent case. (If F5, a fact in the case of A’ v. B’ that was not present in A v. B, counts in favor of B’, then the present court might decide for B’ and cite F5 as the fact that distinguishes A v. B.) The only problem with the result model is that it cannot work. To begin with, the later court has only filtered access to the facts of the precedent case. It knows only what the precedent court has told it. A precedent court that wanted to bind subsequent courts quite broadly could just cite a few facts, described at a high level of generality. In other words, it might issue the following kind of opinion: “In A v. B, B made a promise to A, supported by consideration, and B breached it. Decision for A.” The later courts would know nothing of the particulars of B’s promise, A’s consideration, or B’s breach. In another case involving a promise, consideration, and breach, but one in which the court believes the promisor should prevail, how can the court distinguish A v. B, given that it knows only the generalities and not the particulars of that case? There is a way for a court to distinguish a precedent despite the precedent court’s spare and highly abstract recitation of the facts it considered material. The later court might assume the presence of an unmentioned fact in the precedent case, a fact that would have justified the precedent court’s decision and that is absent in the present case. Suppose, for example, that the precedent court’s decision for A in A v. B would have been correct if a decision for B would have triggered a great calamity. (Nuclear terrorists would have detonated an atom bomb had the decision been for B.) Call that fact, unmentioned by the precedent court, F6. If the precedent court recites facts F1, F2, and F3, in support of its decision for A, the later court can read the precedent as F1, F2, F3, and F6. And because F6 does not exist in A’ v. B’, the precedent court can decide for B’ and distinguish A v. B on that ground. This tack, however, makes precedents incapable of constraining, as one can always come up with some fact, unmentioned in a precedent case, that would have justified its decision, even if one has little reason to surmise such a fact was present. Moreover, it seems quite uncharitable towards the precedent court to assume that it would fail to mention those facts that would have justified its decision while mentioning facts that did not do so. But instead of distinguishing A’ v. B’ on the ground that a fact that was possibly present in A v. B is absent in A’ v. B’, the court in A’ v. B’ might distinguish A v. B on the ground that a fact that is present in A’ v. B’ was not mentioned in A v. B and can therefore be assumed to have been absent in that case. This seems more charitable to the precedent court than assuming it failed to mention a fact that was present and necessary for justifying the outcome. But it leads to the same result, namely, that the court in A’ v. B’ is unconstrained by A v. B. For if the court believes A’ v. B’ should be decided in favor of B’, it believes the facts justify that result. And because the facts justify it, the court can assume that at least some of them were absent in A v. B. All of this is quite abstract, so let me try to give a concrete example. Suppose in A v. B, A sought to enjoin B’s keeping a caged pet bear in a residential neighbor-

Precedential Constraint, Its Scope and Strength

79

hood, and the court sided with A, finding a bear to be a dangerous nuisance. Suppose in A’ v. B’, a similar injunction is sought, but the court believes that in the absence of A v. B, B’ should win. In other words, it does not believe keeping a caged bear in a residential neighborhood to be a nuisance. Suppose there is evidence in A’ v. B’ that many neighborhood children enjoy seeing the bear. No such evidence is cited in the opinion in A v. B. The court in A’ v. B’ can then distinguish A v. B on that ground and avoid the precedent’s constraint. And it will always be able to do so if it describes the facts in its case with sufficient particularity. If the later court cannot escape the constraint of the precedent by assuming either the presence or the absence of unmentioned facts in the precedent case, then it must assume that the reasoning of the precedent court – that (mentioned) facts F1, F2, F3 by themselves justified a decision for A – is correct reasoning, even though, in the opinion of the later court, it is not. The later court then must ask, what would be correct outcomes if the incorrect reasoning of the precedent court were correct? I do not think that such a question is coherent. It is like asking how seals should be classified if hippos were correctly deemed to be mammals but whales were incorrectly deemed to be fish. Or what would be the correct view of discrimination against gypsies be if discrimination against Jews were correctly deemed immoral but discrimination against Hispanics were incorrectly deemed morally proper. The reasoning of an incorrectly decided precedent cannot cohere with other normative views, at least some of which are correct. It is like a misshapen piece in a jigsaw puzzle. To fit it in renders other pieces incapable of fitting. Moreover, any case that is an a fortiori case for following an incorrect precedent can also be an a fortiori case for following other precedents that were correctly decided. The result model of precedent is the model Ronald Dworkin appears to endorse. Dworkin distinguishes between a precedent’s enactment force – its rule – and its gravitational force.5 The latter refers to the precedent’s result, which now constrains future courts along the axis of fit. Future courts must make their decisions the best they can be, but they must also “fit” their results with the precedent results. Yet that approach is tantamount to courts’ asking what results would be correct were certain earlier incorrect results correct, a question that I do not believe can be coherently answered.6 My conclusion then is that although the result model has all the features we associate with a system of precedent — on it, precedents constrain more than they do on the natural model, and unlike precedents on the rule model, they can be distinguished as well as overruled — it suffers from the fatal defect of incoherence. There is no sensible answer to the question that the result model asks a court to pose: What would the correct decision in our case be should an earlier incorrect decision have been correct? Only the rule model survives as a truly constraining model of precedent.

5 6

See R. M. Dworkin, Taking Rights Seriously (Harvard University Press, Cambridge, MA, 1977) 110–115. See note 1, supra.

80

Larry Alexander

B. INTERLUDE: PRECEDENTIAL CONSTRAINT UNDER STATUTES AND CONSTITUTIONS How does precedential constraint work when the precedent deals with a statutory or constitutional text? There are two aspects to such precedents. The first is the interpretation of the statutory or constitutional text. That amounts to the precedent court’s substituting its language for the language of the text. The court’s language is supposed to clarify the meaning of the text and not alter it, as when the court substitutes a non-ambiguous term for the text’s ambiguous one. The precedent court’s new formulation of the text constrains later courts to treat that formulation as the text’s correct interpretation even if the later courts disagree that it is. The precedent court’s formulation is now the rule the later courts must follow in place of the later court’s own interpretation of the statutory or constitutional text. The rule model of precedent captures this aspect of precedential constraint in interpreting statutory and constitutional texts. The second aspect looks at what the precedent court does once it has interpreted the text. If it has interpreted the textual provisions to be a determinate rule, then it will have applied the rule in the precedent case. Presumably, that will not be the aspect of the precedent case with which the later court disagrees. On the other hand, if it interprets the textual provision to be a standard, then either it will have merely decided the case in all its particularities under that standard, or it will have created rules for implementing the standard and decided the case under such rules. The later courts would be bound by both the interpretation of the text and by the implementing rules announced by the precedent court. If no implementing rules were announced, and the precedent case was merely decided directly under the announced standard, only the interpretation of the text to be a standard would constrain later courts. C. THE STRENGTH

OF

PRECEDENTIAL CONSTRAINT

The question of strength is a question of when precedents are properly overruled. An answer of “never” is too extreme, as it could entrench catastrophically poorly decided cases. On the other hand, an answer of “whenever following the precedent decision leads to an incorrect result” reduces precedents to their status under the natural model, which is one of complete non-entrenchment. Too little entrenchment results in too little settlement of issues that need to be settled, just as too much entrenchment results in too many potentially disastrous mistakes. The question of how much to entrench precedents against overruling is one aspect of the more general issue of legal transitions. Law should be changeable in order to adapt to new circumstances and information and to purge itself of past errors. But law should also be relatively stable and predictable so people can confidently plan their affairs. The problem of legal transitions appears in controversies over what changes in regulations of property and contracts should lead to compensation. Such

Precedential Constraint, Its Scope and Strength

81

controversies surface in constitutional litigation under the takings clause, the impairment of obligations of contracts clause, and the due process clauses.7 At the deepest level, these controversies boil down to the problem of the rationality of rule following: For although law’s primary moral function is to settle moral controversies, and law achieves this settlement through determinate rules (“standards” leave matters unsettled), determinate rules’ determinacy comes from their bluntness and thus their over and under-inclusiveness vis-à-vis their background moral justification. This in turn means that some applications of legal rules will depart from what an all-things-considered direct application of the background moral norms would require, leaving actors with moral reasons for departing from rules that they have moral reasons to maintain.8 But the problem of legal transitions also appears in other guises, including the question of the proper strength of precedential constraint, i.e., when should precedents be overruled. I know what the wrong answers to this question are: That we should never overrule precedent is wrong, and also that we should overrule it whenever we disagree with it. (I should say that the latter answer is not wrong if you do not believe precedent should constrain in the domains of statutory and constitutional adjudication, and there are vocal opponents of precedential constraint in those domains, especially the latter.9 I just happen to disagree with those people. For the logic of the argument that erroneous judicial interpretations of statutes and constitutions should never bind later courts also entails that higher courts’ misinterpretations should not bind lower courts — a repudiation of so-called vertical precedent — and even more radically, that judicial misinterpretations should not bind anyone, officials and citizens alike, which would undermine all principles of repose, including res judicata and the force of judicial orders.) * * * * * Let me summarize. I have argued that the rule model of precedent is the only model of precedent that is capable of constraining later courts. On the rule model, the holding is the announced rule, and the dicta are everything else said by the court. Moreover, the holding, if it applies, must either be followed or overruled, at least in part. It cannot be distinguished. The result model of precedent, on the other hand, is incapable of constraining in any coherent way. Although it, unlike the rule model, leaves room for distinguishing precedents, it in fact leaves room for extinguishing them through distinguishing them. Finally, with respect to the strength of precedential constraint, I have argued that precedent rules should be overrulable, at least by certain bodies. Absolute entrenchment is too strong. But no entrenchment is also ill-advised. And there can be 7

8 9

See U.S. Const. art. I, § 10; U.S. Const. amend. V; U.S. Const. amend. XIV, § 1. See also Symposium, Legal Transitions: Is There an ideal Way to Deal With the Non-Ideal World of Legal Change? (2003) 13 Journal of Contemporary Legal Issues, 1. See Alexander & Sherwin, note 1, supra, at Ch. 4. See M. S. Paulsen, ‘The Intrinsically Corrupting Influence of Precedent’ (2005) 22 Constitutional Comment, 289; G. Lawson, ‘The Constitutional Case Against Precedent’ (1994) 17 Harvard Journal of Law and Public Policy, 23.

82

Larry Alexander

no algorithm for determining when erroneous precedents should be overruled. For if a court is instructed to overrule erroneous precedent rules whenever, all-thingsconsidered, the costs of overruling are less than the benefits of doing so, then the rule model of precedent would collapse into the natural model, under which the only “constraint” is “do the right thing.”10 Moreover, if a judge concludes that, all things considered, doing the right thing requires entrenching rules more than the natural model would dictate, the natural model would collapse back into the rule model, thus generating a practical paradox.11 To avoid the Scylla of absolutely entrenching mischievous rules and the Charybdis of undermining the entrenchment of rules and thus the value of rules, courts must rely on a presumption in favor of following even mischievous rules the strength of which cannot be captured in any formula.12

10 11 12

See Alexander & Sherwin (n. 1), supra, at Ch. 4. See Alexander, ‘Constrained by Precedent’ (n. 1), supra, at 15–16, 48–51. See Alexander and Sherwin (n 1), supra, at Ch. 8 (but see Ch. 4, discussing presumptive positivism).

PATRÍCIA PERRONE CAMPOS MELLO THE ROLE

OF

PRECEDENTS

AS A

FILTER

FOR

ARGUMENTATION

A. INTRODUCTION This essay examines the role and the limits of a theory of precedent in promoting the values of legal certainty, equality, legitimacy, and efficiency of the courts. It demonstrates, further, that precedents have another major role: they serve as a filter for legal argumentation, guiding litigants and judges on the issues to be discussed and considered in the decision of a case. With this objective, section I clarifies the relationship between the use of precedents and the values mentioned above. Section II demonstrates the manner in which precedents are applied. Section III analyses the subjective and institutional elements that also influence judicial behavior. Finally, section IV presents the conclusion, highlighting the importance of precedents for legal reasoning. B. REASONS

FOR ADOPTING BINDING PRECEDENTS

Imagine that the eldest daughter of a couple has asked her father to go out, on a weeknight, during the academic term, in order to attend her best friend’s birthday party, and that her father has consented. The following week, her younger brother also asked their father to go to a party with a friend, but had the request denied. The father then explained to the son that, in the first case, the occasion for going out was exceptional: the birthday celebration of a person very dear to his sister. He explained further, that, unlike the son, his sister was doing well in school and, therefore, the father could make an exception to the general rule that forbade his children from going out on weeknights during the academic term. A week later yet, the son asked for permission to go to a classmate’s house, to complete a group project for school. His father consented on the grounds that it was an exceptional case and had academic purposes. These precedents indicate the parents’ criteria regarding weeknight outings during the academic term. The general rule is that they are prohibited. The purpose for the said rule is to maintain an orderly life, as a way of promoting the children’s health and academic performance. The first exception holds that weeknight outings are allowed if they are extraordinary, provided that the child is doing well in school. The basis for the exception is that, under such conditions, weeknight outings do not compromise the purpose of the general rule. The second exception stipulates that weeknight outings are allowed when they are extraordinary and required for school work. In this case, as well as in the previous one, going out at night does not compromise good health because it is exceptional. And it is not detrimental to academic performance because it is necessary for completing a group project for school. Thus, the analysis of the three precedents reveals that behind the rule and its two exceptions lies a basic principle: the promotion of good health and academic performance. The exceptions are justified because they are in harmony with said

84

Patrícia Perrone Campos Mello

principle. If the children fully understand the precedents as well as the reason underlying these precedents, they will be able to predict situations in which their father will allow them to go out at night, and situations when he will not. They will notice that the same criterion applies to each and all of them. They will understand that such criterion has a rational basis and they will be capable of adjusting their conduct without having to consult their father. The raison d’être for binding precedents and the reasoning one formulates based on them can be understood from examples of daily life, the logic of which is reproduced in the process of judicial decision making. First of all, imposing upon judges the obligation of following past decisions concerning a certain matter ensures the predictability, stability, and continuity of the law, as the decision applied to one dispute will determine the outcome of similar claims. Binding precedents are tools to guide peoples’ conduct on certain matters. They are, to this extent, an element of legal certainty. Furthermore, the universe of judicial decision making, different from the example above, does not rely on a single decision maker with just one well-established criterion for deciding cases. On the contrary, it is composed of a multiplicity of judges who may have divergent views on similar cases. To this extent, the adoption of binding precedents preserves equal treatment among litigants. All those who find themselves in the same situation will (or should) obtain the same legal response. This argument again harbors the notion of legal certainty and also of justice as equality: even if a certain ruling is not the best, it is fair because it is the same for all. The equality imperative determines that, when analyzing a dispute, judges must take into account not only the legal response to be given to the case at hand, but also to all the other disputes that are similar. It requires, thus, the search for an objective legal basis. “Under the principle of objectivity, courts are obliged to reason from propositions that are universal, that is, propositions the courts are ready to apply not merely to the parties to the immediate dispute, but to all similarly situated disputants who may come before them in the future. Stare decisis gives effect to this concept too. Under stare decisis, a court is on notice that if it chooses to apply a given proposition to resolve a dispute between these litigants today, it may be obliged to apply the same proposition to all similarly situated disputants in the future. Thus stare decisis discourages a court from deciding cases on the basis of propositions it would be unwilling to apply to all similarly situated disputants”.1

Due to such claim of equality, normative precedents limit the power and the discretion of judges, confining them to the conclusions already arrived at. Therefore, normative precedents narrow judicial activity to the traditional notion that said activity consists of deciding cases, with objectivity and neutrality, based on pre-existing norms. Even if such is not an accurate description of the decision-making process, it corresponds, to some extent, to the layman’s expectations and views on judicial activity. Moreover, the litigants’ perception that judges decide based on logical, fair, and established criteria that predate the appreciation of their case preserves the legitimacy and the credibility of the courts, contributing to their institutional stability. 1

Melvin Aron Eisenberg, The Nature of the Common Law (Harvard University Press: Cambridge, 1988), 47–48. Also: Laurence Tribe, American Constitutional Law. 3. ed. (Foundation Press: New York, 2000), v. 1, at 82; Benjamin N. Cardozo, The Nature of the Judicial Process (Dover Publications: New York, 2005), 17.

The Role of Precedents as a Filter for Argumentation

85

Finally, the use of binding precedents may render the legal system more efficient. In the domestic example above, precedents would make it unnecessary for children to ask for their father’s permission to go out every time they were invited to, because, in most cases, they will be capable of predicting their father’s response, sparing him from new requests and reiterated explanations about the matter. The same occurs in the judicial context. The application of binding precedents saves time and resources because it renders unnecessary new considerations about matters that are already decided. It also dissuades potential claimants from bringing frivolous claims, since they already know how such actions would end.2 It avoids the issuance of conflicting rulings and, to some extent, judicial errors. Therefore, in synthesis, the values that justify the adoption of a system of binding precedents are: legal certainty, equality and the promotion of legitimacy, and efficiency of the judicial process.3 C. DEALING

WITH BINDING PRECEDENTS

1. DEFINING THE NORM THAT WILL GOVERN FUTURE CASES Legal reasoning based on precedents is grounded in certain essential concepts: the notions of holding, justification, and obiter dictum.4 The holding (or ratio decidendi) is the norm extracted from the case that binds inferior courts. It corresponds precisely to the legal thesis, that is, the premise that justifies the court’s conclusion. The no2

3

4

Earl Maltz, ‘The Nature of Precedent’ (1988) 66 North Carolina Law Review 367–392, at 371 ff; Thomas Hansford, and James Spriggs II, The Nature and Timing of the U. S. Supreme Court’s Interpretation of Precedent. Available at . Accessed on Jan. 29, 2011, p. 6 ff; The Politics of Precedent on the U.S. Supreme Court. Available at: . Accessed on Sept. 20, 2005, p. 163 ff. This assertion is generally valid for common law as well as for civil law countries. It is true that binding precedents are the main source of law in common law jurisdictions. In contrast, precedents usually have a mere persuasive efficacy in civil law countries, where the primary source of law is statute law. However, many civil law jurisdictions tend to adopt binding precedents, especially in constitutional matters (but not necessarily restricted to them). Constitutional law employs rules/principles with abstract and potentially conflicting meanings for the protection of fundamental rights (e.g.: the right to privacy vs. right to freedom of expression and press). Several civil law countries have attributed to the Judiciary the task of defining the reach of these rules/principles when deciding disputes. The notion of constitutional supremacy and the need to preserve the Constitution’s normative force have justified the granting of binding force to the decisions about constitutional matters. This is currently happening, to a greater or lesser extent, in Germany, Italy, Spain. It is also a tendency in Brazil, Argentina and Colombia (cf. Robert Alexy and Ralf Dreier, ‘Precedent in the Federal Republic of Germany’; Michelle Taruffo and Massimo La Torre, ‘Precedent in Italy’; Alfonso Ruiz Miguel and Francisco Laporta, ‘Precedent in Spain’, in N. MacCormick and R. Summers (eds.) Interpreting Precedents: A Comparative Study. (Ashgate: Dartmouth, 1997), 26–27, 154–155 and 272; René David, Os Grandes Sistemas do Direito Contemporâneo (Martins Fontes: São Paulo, 2002), 160–161; Luís Roberto Barroso, ‘Neoconstitucionalismo e Constitucionalização do Direito (O triunfo tardio do direito constitucional no Brasil)’ (2005) 240 Revista de Direito Administrativo, 1–42, at 9–10). The mentioned concepts are useful tools for a legal reasoning based on binding precedents despite the differences in legal systems.

86

Patrícia Perrone Campos Mello

tion of holding is one of the most important concepts for the application of binding precedents; however, its identification may not be as simple as it appears. The interpreter shall examine the decision, with special attention to its justification.5 He shall seek to identify which facts the court considered to be important, the legal issues discussed and, finally, the court’s understanding of these facts and issues.6 There are cases when the court raises more than one question of law or more than one reason for deciding. In these circumstances, if the reasons for adopting certain solutions are cumulative, they will generate a single binding rule. If they are independent from each other, and capable of autonomously leading to the same result, each one of them may be considered a concurrent holding. Finally, if one of the reasons is not a determinant factor for the decision of the case, it will not constitute a holding and, therefore, will not bind future claims.7 Back to the example mentioned above, when the father denied authorization for his son to go out at night with a friend, the father mentioned that, his daughter had been allowed to attend her best friend’s birthday party because, in that case, going out was exceptional as the birthday party would only happen that night. But, still, he added that, differently from the son, his daughter was doing well in school. This last statement could generate questions about the meaning of the rule that arises from the decision. If the reasons for the father’s decision were cumulative, the holding would be: going out on a weeknight during the academic term is only allowed when the occasion is exceptional and the child has good academic perform5

6

7

Henry Paul Monaghan, ‘Stare Decisis and Constitutional Adjudication’ (1988) 88 Columbia Law Review 4, 723–773, at 763–767; Larry Alexander, ‘Constrained by Precedent’ (1989) 63 Southern California Law, 1–64, at 9–10; Charles Cole, ‘Stare Decisis na Cultura Jurídica dos Estados Unidos. O Sistema de Precedente Vinculante do Common Law’, trans. Maria Cristina Zucchi (1998) 87 Revista dos Tribunais, n. 752, 11–21; Maltz, ‘The Nature of Precedent’ (n. 2), 372–383; Geoffrey Marshall, ‘What is Binding in a Precedent’, in: Interpreting Precedents: A Comparative Study (n. 3), 503–518; Edward D. RE, ‘Stare Decisis’ (1994) 73 Revista de Processo, 47–54; Robert Summers, ‘Precedent in the United States (New York State)’, in: Interpreting Precedents: A Comparative Study (n. 3), 355–406; Isabelle Rorive, ‘La Rupture de la House of Lords avec un Strict Principe du Stare Decisis dans le Contexte d’une Réflexion sur l’Accélération du Temps Juridique’, in: L’Accélération du Temps Juridique, ed. Philippe Gerard, François Ost; and Michel Van De Kerchove, Michel (Facultés Universitaires Saint-Louis: Brussels, 2000), 801–836, at 807; Arthur L. Goodhart, ‘Determining the Ratio Decidendi of a Case’ (1959) 22 Modern Law Review, 117–124, at 162; Celso de Albuquerque Silva, Do Efeito Vinculante: sua Legitimação e Aplicação (Lumen Juris: Rio de Janeiro, 2005), 182–183. There is some controversy about the methods for identifying the holding, which sometimes emphasize the facts raised by the case and, other times, the importance of the decision justification for this purpose. On the subject: cf. Henry Paul Monaghan, ‘Stare Decisis and Constitutional Adjudication’ (n. 5), 763–766; Larry Alexander, ‘Constrained by Precedent’ (n. 5), 10 ff; Ronald Dworkin, Uma Questão de Princípio, trans. Luis Carlos Borges (Martins Fontes: São Paulo, 2005), 453 ff; Laurence Tribe and Michael Dorf, On Reading the Constitution (Harvard University Press, Cambridge, 1991), 114–117; Frederick Schauer, ‘Precedent’ (1987) 39 Stanford Law Review, 571–606; Frederick Schauer, ‘Rules, the Rule of Law, and the Constitution’ in: (1989) 6 Constitutional Commentary, 69–85;Earl Maltz. ‘The Nature of Precedent’ (n. 2), 367 ff; Marhsall, ‘What is Binding in a Precedent’ in Interpreting Precedents: A Comparative Study (n. 3), 503–518; Summers, ‘Precedent in the United States (New York State)’ in Interpreting Precedents: A Comparative Study (n. 3), 355–406. Marshall, ‘What is Biding in a Precedent’ in Interpreting Precedents: A Comparative Study (n. 3), 515.

The Role of Precedents as a Filter for Argumentation

87

ance. If the reasons were independent from each other, the norm would be: going out on a weeknight during the academic term is allowed when the occasion is exceptional or when the child has good academic performance. However, if the statement about school performance is considered a mere criticism, and not a determinant factor to the father’s decision, the precedent would hold that: going out on a weeknight during the academic term is only allowed when the occasion is exceptional. In this case, school performance would not determine future decisions. Despite the fact that the holding definition rests largely on the decision’s justification, these concepts are not coincident. The justification includes all the arguments and questions of law mentioned by the court. Part of such arguments and questions may not have been determinant to the decision, being a mere obiter dictum.8 Obiter dicta are precisely those remarks that were not necessary to the decision of the case at hand. For example, arguments put forth by a judge of a collegiate court that were not taken into account by his fellow judges, or remarks about related matters that do not refer directly to the issue at hand (e.g: the son’s performance in school) do not constitute the holding. Obiter dicta are not binding because judge-made law derives its rules from concrete cases. If certain remarks are not necessary to reach the decision, they do not concern the claim presented before the court. Yet, they derive from hypothetical formulations, whose particularities the Judiciary would be unable to consider in the abstract sphere, given that it is not the role of the courts to do so, but rather to resolve the dispute at hand. Notwithstanding, dicta have an argumentative role. In a case that poses, at the same time, various questions of law, a certain statement by the court, perceived as an obiter dictum, ultimately may become an alternative holding. Further, the lower courts may confer to the obiter dictum a considerable weight when there is evidence that it resulted from an attentive scrutiny by the higher court, even though such level of scrutiny was unnecessary for the decision.9 It should be noted that, upon defining the legal thesis that governed the solution of a case, there still remains the question of the level of generality of the language employed to announce the thesis. In order to illustrate the issue, legal literature usually refers to Donoghue v. Stevenson [1932] A. C. 562. In this famous case, decided by the House of Lords, Mrs. Donoghue’s friend purchased a bottle of ginger beer. After ingesting part of the content of the opaque bottle, Mrs. Donoghue noticed the remains of a snail in the drink, which later caused her gastroenteritis. She sued the beverage producer, asking for indemnification, on the basis of an existing (extra-contractual) duty of care in avoiding his products from inflicting harm to final consumers. In its decision, the House of Lords recognized there was such a

8

9

Precisely because of this, Julius Stone proposes a distinction between descriptive ratio decidendi and prescriptive or binding ratio decidendi, the first corresponding to the argumentative process by which a concrete decision has been reached and the second to the content that will bind courts in subsequent cases, i.e., the holding (Julius Stone, ‘The Ratio of the Ratio Decidendi’ (1959) 22 Modern Law Review, 600–601). Summers, ‘Precedent in the United States (New York State)’ in Interpreting Precedents: A Comparative Study (n. 3); Marshall, ‘What is Binding in a Precedent’ in Interpreting Precedents: A Comparative Study (n. 3), 515.

88

Patrícia Perrone Campos Mello

duty, and remarked that the producer could foresee that the lack of care in such circumstances may cause harm. 10 The rule that emerges from the case could be described in narrow or broad terms. One could argue that the precedent determines that beverage producers who cause harm to the health of final consumers must indemnify them, as long as the product is purchased in circumstances that prevent the consumers from discovering any defect (opaque bottle) and there is reasonable knowledge that the absence of care may cause harm. Litigants could also allege that the decision establishes that any producers or service providers who cause harm to final consumers for negligence should have the duty to indemnify, or, yet, that any person who causes damage to another for negligence should also have this duty. The definition of the holding and its reach rests upon an evaluation of the relevant facts, the justification of the decision, how broad the language employed in the ruling was, and also, eventually, the treatment that the ruling received in subsequent cases. Subsequent decisions help define the normative content of past cases to the extent that the confrontation of the latter with new situations clarifies the standard or principle the court employed in its first decision. 2. DECIDING WHETHER TO APPLY (OR NOT TO APPLY) THE NORM TO NEW CASES The application of the ruling established in a precedent to a new dispute depends upon a comparison between the essential elements of the precedent to those of the new case. The interpreter confronts the relevant facts in each dispute; the values they refer to; and the question of law to be resolved by the court. He/she also will examine the basis of the first decision. The existing similarities or differences will be debated. In the end, the interpreter decides whether the new dispute is identical or not, and, therefore, if it must be governed by the precedent. The application of the precedent to new cases will serve to confirm the holding and its reach, increasing the force of the ruling and its level of determination.11 On the other hand, if the court considers that the new case presents factual particularities that evoke a substantially different question of law, it will not apply the precedent, distinguishing the second case from the first. This distinction is a consequence of the criteria that guide the application of precedents, according to which: “the rule follows where its reason leads”; “where the reason stops, there stops the 10

The case heard by the House of Lords concerned simply whether there was a case to be tried. The question was whether the beverage manufacturer, who was not in a contractual relationship with Mrs. Donoghue, had an extra-contractual duty to indemnify her, considering that his defective product was sold to a third party, not to Mrs. Donoghue. The Court decided there was a case to be tried, since the producer could be liable in the mentioned circumstances, despite the non existence of a contract, because: “a manufacturer of products which he sells in such a form as to show that he intends them to reach the ultimate consumer in the form of which they left him with no reasonable possibility of intermediate examination, and with the knowledge that the absence of reasonable care in the preparation or putting up of the products is likely to result in injury to the consumers life or property owes a duty to the consumer to take that reasonable care”. 11 Karl Llewellyn, ‘The Case Law System in America’ (1988) 88 Columbia Law Review, 989–1020, at 1006–1007.

The Role of Precedents as a Filter for Argumentation

89

rule”.12 Thus, if for the sake of fairness and integrity similar cases shall receive the same treatment, for the very same reasons, different situations must also be treated differently.13 In the domestic example, if the father had to authorize a nightly outing for a reason other than attending an exceptional event, and his son was doing poorly in academics, but was on vacation from school, it is likely that the father’s decision would be different. In this situation, a relevant fact, present in all of the other scenarios, would be absent: the circumstance that the nightly outings would take place on weeknights during the academic term. This factual difference would render inapplicable the principle that guided the previous decisions. Thus, the reasoning based on precedents permanently produces new opportunities for deciding: “[B]y maintaining at the centre of ‘the rule of stare decisis’ a notion of the ratio decidendi of a case which is almost a perfect medium for the creation of multiple and competing references. While thus leaving room for the play of contemporary insight and wisdom, however, the notion also directs the attention of the later court to the contexts of earlier cases, and to the views of logical consistency, experience and values displayed by judges in the earlier contexts”.14

The possibility of modifying the holding in light of new situations brought before the court is one of the essential characteristics of the common law. The tension between applying or distinguishing from among precedents in subsequent disputes is permanent in that legal system. A party’s claim to benefit from an understanding in his/her favor will be opposed by the other party’s argument for distinguishing, and, from this confrontation, judge-made law will evolve.15 Dealing with precedents is not a mechanical process. It is markedly argumentative and value dependent. Despite being justified on reasons of, among others, promotion of legal certainty, predictability, and stability of the law, the application of precedents faces constraints in limiting the power and discretion of judges. 3. LEADING AND HARD CASES16 The influence a precedent exerts over the ruling of a subsequent case is not confined to its holding. Certainly, the holding will embody the norm that will govern similar cases in the future, and only them. However, even if only the holding has binding effects, the underlying principle behind it has a justifying force capable of influenc12

Statements attributed to Karl Llewelyn (apud Summers, ‘Precedents in United States (New York State)’ in Interpreting Precedents: A Comparative Study (n. 3) 390). 13 Tribe and Dorf, On Reading the Constitution (n. 7), 71 ff; Karl Llewelyn, The Common Law Tradition: Deciding Appeals (Little, Brown and Company: Boston, 1960), 77 ff; Goodhart, ‘Determining the Ratio Decidendi of a Case’ (n. 5), 166; Ronald Dworkin, O Império do Direito, trans. Jefferson Luiz Camargo (Martins Fontes: São Paulo, 2003), 300 ff; Summers, ‘Precedent in the United States (New York State)’ in Interpreting Precedents: A Comparative Study (n. 3), 390–394; Schauer, ‘Precedent’ (n. 6), 571–605; Schauer, ‘Rules, the Rule of Law, and the Constitution’ (n. 6), 69–85. 14 Stone, ‘The Ratio of the Ratio Decidendi’ (n. 8), 619. 15 Llewellyn, Gewirtz and Ansaldi, ‘The Case Law System in America’, 989–1020. 16 The first decision that faces a certain legal issue is called a leading case. Leading cases are good candidates for hard cases, cases of difficult solution in which there are references to principles that could result in divergent conclusions.

90

Patrícia Perrone Campos Mello

ing the judgment of other cases. The general principle of law that serves as basis to the rule created by the precedent engenders a kind of gravitational force capable of interfering in the decision of future cases. In Dworkin’s words: “[…] when a precedent does have enactment force, its influence on later cases is not taken to be limited to that force. Judges and lawyers do not think that the force of precedents is exhausted, as a statute would be, by the linguistic limits of some particular phrase […]. He would urge that the earlier decision exerts a gravitational force on later decisions even when these later decisions lie outside its particular orbit. […] In adjudication, unlike chess, the argument for a particular rule may be more important than the argument from that rule to the particular case.”17

Imagine that Parliament passed a law prohibiting the use of contraceptives by married couples, and that, after examining such law, the Constitutional Court found it to be unconstitutional, affirming that the decision about whether or not to have children is a matter confined to every citizen’s realm of privacy, in which the State cannot interfere. The holding extracted from such a case could be worded as follows: “The State cannot intervene in a couple’s decision to use contraceptives, in order to avoid having children, because their decision is protected by the right to privacy”. The principle that served as basis to the holding, in its turn, would state that: “Citizens have a sphere of privacy, concerning intimate choices and personal projects, in which the State cannot intervene, at the risk of violating the right to privacy.” Since only the holding would bind judges when deciding subsequent cases, the ruling would only govern cases about the right of couples to use contraceptives to avoid having children. Suppose then that the court examined a law that criminalized sexual intercourse between people of the same gender. This new case could not be decided by applying the holding of the previous case, nor of any other, and, for this reason, it would constitute a leading case on the subject. Nonetheless, in arguing the unconstitutionality of the mentioned law, it would be possible to evoke the principle that served as basis for the holding in the claim about contraceptives, because in the new case, as in that of contraceptives, the law interferes in people’s intimate and personal choices as well as their sex life, thus violating the right to privacy.18 A similar argument could be used to challenge the constitutionality of a law regulating marriage and marital property regime between individuals of different genders, if said law failed to award similar status to homosexual couples. One could argue that the law adopts an illegitimate criterion of discrimination, as it is incompatible with the principle according to which people’s sex lives are a private matter immune from State interference. 17

18

Ronald Dworkin, Taking Rights Seriously (Harvard University Press: Cambridge, MA, 1977), 111– 112. According to Dworkin, the meaning of norms may change to the extent the comprehension of the values on which they are based advances. Even if one could argue that real life judges cannot be compared to judge Hercules, the above cited passage describes a form of argumentation that lawyers often use, and that, therefore, is in some measure reflected in judicial decisions (in constitutional matters this is evident, as subsequent examples will show). Through argumentation, one seeks to extend the incidence of the principle that served as basis for the holding of a previous case to new cases that are not identical, but that are related to similar values, hoping to guide the court to decide in harmony with its previous decisions. Cf., and: Tribe, and Dorf, On Reading the Constitution (n. 6), p. 109–110; Dworkin, O Império do Direito (n. 5), 123–125. Tribe and Dorf, On Reading the Constitution (n. 6) 78–79.

The Role of Precedents as a Filter for Argumentation

91

A law prohibiting abortion could be challenged with basis in similar reasons. Litigants could allege that such law should be declared unconstitutional because it is not compatible with the principle that the Court upheld in deciding the previous cases, as it interferes in a pregnant woman’s right to make decisions about her own body and personal projects. This argument could be contested on the grounds that, in this latter case, the element “intrauterine life in gestation” would engender the incidence of yet another constitutional right: the foetus’ right to life. In deciding, the Court would need to address both allegations, issuing its opinion on the conflict between the two rights and deciding which of them should prevail, and to what extent, in the case at hand. In leading and hard cases, in which courts face, for the first time, questions they have never considered before, principles and values that served as a basis for previous rulings will provide elements for the new decision. To this extent, precedents may serve as important argumentative marks in leading cases, although one must inevitably recognize that these cases entail a broad space for judicial discretion, for which a theory of precedent is unable to offer other solutions. In Justice Cardozo’s words: “Finally there remains a percentage, not large indeed, and yet not so small as to be negligible, where a decision one way or the other, will count for the future, will advance or retard, sometimes much, sometimes little, the development of the law. These are the cases where the creative element in the judicial process finds its opportunity and power. […]. In a sense it is true of many of them that they might be decided either way. By that I mean that reasons plausible and fairly persuasive might be found for one conclusion as for another. Here come into play that balancing of judgment, that testing and sorting of considerations of analogy and logic and utility and fairness, which I have been trying to describe. Here it is that the judge assumes the function of a lawgiver. I was much troubled in spirit, in my first years upon the bench, to find how trackless was the ocean on which I had embarked. I sought for certainty. […]. As the years have gone by, and as I have reflected more and more upon the nature of the judicial process, I have become reconciled to the uncertainty, because I have grown to see it as inevitable. I have grown to see that the process in its highest reaches is not discovery, but creation; and that the doubts and misgivings, the hopes and fears, are part of the travail of mind, the pangs of death and the pangs of birth, in which principles that have served the day expire, and new principles are born”.19

In such instances, the decision making process is based on pragmatic considerations regarding the various values at stake, customs, benefits and disadvantages of a certain decision and its consequences. In the same author’s words: “My analysis of the judicial process comes then to this and little more: logic, and history, and custom, and utility, and the accepted standards of right conduct, are forces which singly or in combination shape the progress of the law”.20 In hard cases, courts will make choices concerning values, in deciding whether or not to infer a fundamental right from a constellation of precedents. 21 Despite all of the law’s efforts to promote legal certainty, and despite this being one of the essential roles of a theory of precedent, it is unable to eliminate the wide space left for judicial discretion. Such space results from the imprecision of language, the vague19 20 21

Cardozo, The Nature of the Judicial Process (n. 1), 161–163. Cardozo, The Nature of the Judicial Process (n. 1), 108. Tribe and Dorf, On Reading the Constitution (n. 6), 116; Cardozo The Nature of the Judicial Process (n. 1), 23–26.

92

Patrícia Perrone Campos Mello

ness and abstraction of certain essential legal concepts, the complexity of real life and the law’s inability to foresee and regulate every possible dispute of interests. This space may also be generated by the dynamism of legal systems, which are constantly in change, in what concerns both the comprehension of the content of existing norms and also the way in which such norms should be interpreted in light of new situations of real life. In this latter case, new decisions may depart from existing precedents, creating new spaces for judicial discretion. 4. OVERRULING A precedent is overruled when the binding court, which established a certain holding, changes its understanding regarding the matter, thus deciding to rule it in a different way. Overruling results from the perception that a decision is incompatible with the other legal norms and principles that govern the subject or with certain social standards related to the values at stake.22 Nevertheless, the decision about overruling a precedent shall still consider reasons of legal certainty, equality, legitimacy, and efficiency that could recommend its maintenance. 23 This is so because the change of the rules usually applied to certain situations may surprise the citizens. It is detrimental to the predictability of the law and results in unequal treatment to similar cases. It suggests that other discussions can be reopened, thus fomenting new disputes before courts. Overruling also lends uncertainty to the criteria used by the court in deciding, and may cause losses to their credibility. There are subject matters that are specifically sensitive to changes of understanding, such as property rights and tax law. If peoples’ conduct and their fundamental choices concerning these matters are informed by previous judicial decisions, grave consequences and damages can arise from overruling them. In those cases, courts will analyze whether the benefits of correcting the precedent justify the burdens imposed to the mentioned values. 24 There are a number of situations that may lead to the overruling of a precedent, as, for example: a) decisions that are unenforceable in practice, because the rule is inoperable, obscure in meaning, or has been undermined by arbitrary distinctions; b) the recognition that an interpretation was mistaken from the very beginning; c) the current understanding that a doctrine is unjust by virtue of cultural, political, social, economic, or technological changes; and d) the obsolescence of a decision caused by the evolution of applicable legal principles.25 22 23

Llewellyn, Gewirtz, and Ansaldi, ‘The Case Law System in America’ (n. 11), 1012–1014. The understanding that a holding is not correct does not necessarily imply the decision to revoke it, at least not in any system of binding precedents. Eisenberg, The Nature of the Common Law (n. 1), 104 ff. 24 The rigor with which the courts will consider whether to overrule or not precedents that are incorrect or outdated, balancing the decision with reasons of legal certainty and equality, will vary according to the legal system. In those in which prospective overruling is admitted, overcoming an understanding tends to be easier because this technique allows the mitigation of part of its burden. 25 Cf. Planned Parenthood of Southeastern PA. v. Casey, 500 U.S. 833 (1992); Tribe, American Constitutional Law (n. 1), 236 ff; Rorive, ‘La Rupture de la House of Lords avec un Strict Principe du Stare Decisis dans le Contexte d’une Réflexion sur l’Accélération du Temps Juridique’ (n. 5),

The Role of Precedents as a Filter for Argumentation

93

When the holding does not offer a precise guideline, as when a tribunal formulates a binding rule by employing vague concepts, the precedent is not able to ensure safe and equal treatment to all litigants. The same happens when a certain ruling established by the binding court is disfigured by arbitrary distinctions made by the bound judges, as a form of tacit insubordination against its application (normally for perceiving it as unfair or mistaken).26 In these situations, the burdens of revoking the precedent tend to be small, as it already constitutes a fragile reference for judges and citizens alike. In the three last situations – recognition that a ruling was mistaken from the beginning, obsolescence of the decision due to the evolution of the applicable legal principles, and social changes – the arguments and benefits that favor the modification of an understanding shall be weighed against other reasons and burdens that recommend its maintenance. If arguments of legal certainty and/or equality advise against the overruling of the precedent, the court may appeal to intermediary forms of decision, as a technique for optimizing the efficacy of all the values at stake and minimizing the sacrifice of each of them. An intermediary form of decision, in the sense mentioned above, is the prospective overruling, through which, despite the application of the old precedent to the case at hand, the court announces that it will no longer be applicable from that day onwards or from a certain future date stipulated in the decision. Another intermediary form of decision consists in the signaling technique, whereby the binding court applies the old precedent to the new case, but signals to the legal community its intention of changing that precedent, what discards, as of such notice, the justified belief in its application, opening a track for its revocation. 27 Overruling precedents is essential to the development of law and to the preservation of legal authority. The law seeks to regulate a changing reality. If a theory of precedent is not able to provide mechanisms that allow it to evolve along with the society it regulates, the law itself would become outdated. This way, as it would not be able to coordinate responses to the various conflicts of interests, it would eventually be overtaken by reality. Some time ago, racial segregation in schools was not regarded as a violation to the right to equality. Women and men did not enjoy the same basic rights. It was impossible to know with certainty the paternity of a child before the existence of the DNA test. There was no way of verifying that a foetus in the womb would not be viable, due to malformation of a vital organ. One could not even imagine that the world would be interconnected through the internet and that

816 ff; Monaghan, ‘Stare Decisis and Constitutional Adjudication’ (n. 5), 758 ff; Summers, ‘Precedent in the United States (New York State)’ (n. 3), 374 ff; Eisenberg, The Nature of the Common Law (n, 1), 104 ff; Cardozo, The Nature of the Judicial Process (n. 1), 147; Cole, Stare Decisis na Cultura Jurídica dos Estados Unidos. O Sistema de Precedente Vinculante do Common Law (n. 5), 18; Alexander, ‘Constrained by Precedent’ (n. 5), 1–64; Hansford, Spriggs II, ‘The Politics of Precedent on the U.S. Supreme Court’ (n. 2), 19; Silva, Do Efeito Vinculante: sua Legitimação e Aplicação (n. 5), 262–284. 26 The distinction between cases will be arbitrary when it leads to the non application of the precedent by invoking facts and arguments that do not justify a differentiation. Not all factual differences are legally relevant. Schauer, Rules, the Rule of Law, and the Constitution (n. 6), 69; Cardozo, The Nature of the Judicial Process (n. 1), 19. 27 Eisenberg, The Nature of the Common Law (n, 1), 122.

94

Patrícia Perrone Campos Mello

new types of crimes would be committed with this tool. These are relevant social changes that may interfere with the content of the law. In deciding to overrule a precedent, a court shall justify its reasons surpassing the arguments for its adoption in the past. The court may also demonstrate that the modification of the law brings more benefits than burdens. To this extent, the previous precedent will impose a kind of argumentation that restricts the reasoning of the court in its decision to overrule. It is necessary to acknowledge, however, that a considerable space for discretion remains, one that is difficult to control and that, like the decisions in leading and hard cases, is prone to the interference of subjective and institutional factors. 5. PRACTICAL REASONING AND CONCERN WITH CONSEQUENCES For all the above, dealing with precedents implies a practical reflection focused on the problem and on the issues raised by the case. Past holdings are applied considering the principles and purposes that justified their adoption, and, therefore, the law is redefined in every new case at hand, according to the solution that best achieves the objectives of the norm. Courts will identify such solution based on considerations regarding the Court’s experience, arguments raised by the parties, values and social interests to be served, moral patterns of the community, analogy, custom, tradition, historical and teleological concerns, integrity and harmony of the legal system, social standards of conduct, and the perception of the most adequate and fair decision.28 A pre-defined script does not exist. Those elements are tested and balanced. In some circumstances, the legal system will have a relatively ready answer for the situation. In others, the final result will consist of a truly judicial creation. The consequences of adopting a certain position are also considered in the decision making process. The very idea of a system of binding precedents requires the judge to reflect on the outcome of rulings for new disputes, since he will be setting a norm that will govern similar claims in the future. In deciding a leading case, the decision to adopt a given solution will take into account the principles already affirmed by the court in previous decisions, but also the practical results of such a solution. In overruling a precedent, the court will assess its incorrectness or obsolescence, but also the impact that the change will produce on other important values, such as legal certainty and fairness. If the court concludes that a change of understanding will bring about more burden than benefit, it is possible that it will refrain from overruling its precedent or it will alter the precedent with prospective effect only. Thus, judicial decisions are taken with an eye to the past and another to the future.29 On the other hand, the consequences, as well as the facts, are not important in themselves. To identify the ones that must be taken into account in a decision and their weight, as well as to reflect on the various elements described above, the judge has, inevitably, to resort to cultural, moral and historical, references, reopening the space for choices. 28 29

Cardozo, The Nature of the Judicial Process (n. 1), 39. Richard Posner. How Judges Think (Harvard University Press: Cambridge, 2008), 119, 197–198, 230, 243–245; The Problems of Jurisprudence (Harvard University Press: Cambridge, 1990), 454– 469.

The Role of Precedents as a Filter for Argumentation

D. THE

95

OTHER FACTORS THAT INTERFERE ON THE DECISION MAKING PROCESS

Dealing with precedents is an activity that involves various levels of indetermination. As already mentioned, precedents may provide incomplete answers or simply not provide an adequate response to the decision of a given case. On these occasions, the judicial decision making process becomes more vulnerable to two kinds of factors which, in addition to the orthodox legal materials30, may interfere in the judicial rulings: a) subjective factors and b) institutional factors. The subjective factors include judges’ psychological make-up, specific groups with which they identify themselves, and their ideological preferences. The institutional factors are related to any and all considerations a judge takes into account in forming his/her conviction, based on the awareness of being a member of a tribunal and/or on the concern to preserve the court as an institution. These considerations encompass the relationship among the judges of the same panel, as well as the relationship of the appeal court with the other lower ranking courts, the branches of government (Executive and Legislative) and the public opinion.31 1. SUBJECTIVE FACTORS WHICH INTERFERE IN THE DECISION MAKING PROCESS Judicial decisions are issued by human beings, based on perceptions and knowledge acquired throughout their lives, which are not purely technical or legal, and that interfere, in an unconscious manner, in their comprehension of the problems and in the solutions they propose for them. Differences in gender, culture, education, religion, social environment, and professional experiences are determinant factors in how a judge assesses certain situations brought to trial. Human perceptions are the product of an interaction between impressions produced by the impact of the external world on the senses, and a mental classificatory apparatus that springs from each person’s own background. This apparatus is responsible for unconscious preconceptions that interfere with the cognition and the judging ability of every person.32 Precisely because the pre-understandings are unconscious, even when the judge believes he is acting objectively and with neutrality, he will, to some extent, be projecting his own subjectivity onto the solution of the problem. “We may try to see things 30

The term “orthodox legal materials” refers to a set of precedents, other rules (laws, decrees, regulations) and interpretive theories recognized by the law. 31 Legal literature offers several theories that try to explain the decision making process. They will not be addressed because they are beyond the scope of this work. What is important here is to demonstrate that subjective and institutional factors may be equally involved in this process, as well as to clarify the role that a theory of precedent plays in such circumstances, to the extent that the interference of same factors can compromise the values that justify its adoption. 32 Posner, How Judges Think (n. 29), 67–70; Lawrence Baum, The Supreme Court. The Supreme Court. 9. ed. (CQ Press: Washington, 2007), 120–121; Luís Roberto Barroso, ‘Fundamentos Teóricos e Filosóficos do Novo Direito Constitucional Brasileiro (Pós-Modernidade, Teoria Crítica e Pós-Positivismo)’, in: Temas de Direito Constitucional (Renovar: Rio de Janeiro, 2003), t. II, 9; and Luís Roberto Barroso, ‘Constituição, Democracia e Supremacia Judicial: Direito e Política no Brasil Contemporâneo’ Available at . Accessed on Jan 29, 2011.

96

Patrícia Perrone Campos Mello

as objectively as we please. Nonetheless, we can never see them with any eyes except our own.”33 Therefore, the judge’s background is a powerful influence on the formation of his conviction. A person who lives in a highly dangerous city, in constant fear, and believes that criminal conduct, as a rule, goes unpunished, will tend to view criminal laws that restrict the rights of the accused differently from a person who does not live in such conditions or who has been, him/herself, unfairly criminally prosecuted. A judge who has served as State Attorney will probably see certain matters involving the Treasury Department with different perspectives than others without that same experience. The religious beliefs of a magistrate possibly may influence his assessment on certain moral issues that are brought before him for trial. The understanding of the influence that a judge’s background exerts on his cognitive capacity finds a good parallel in the way the elderly tend to analyze problems and propose workable solutions to them. Their experiences, though forgotten, are accessible sources of knowledge. New situations are in some ways similar to past experiences. The unconscious accumulation of these experiences enables them to intuitively know how to act in new situations. In a similar manner, judges’ experiences feed their intuition and, over time, constitute a practical subjective framework with its own “normative force” on their judicial behaviour. 34 Moreover, people worry about the opinion that others may have of them and, thus, such opinions can interfere with their attitudes. The idea one has of himself does not develop in isolation from other individuals. It develops from the interaction with them. Therefore, one’s identification with a group shapes his thoughts and attitudes.35 The general community is just one of the many groups judges care about. There are others: his fellows from court, other judges, social, political and professional groups, close friends, and family. These smaller and partial groups may be even more influential on the decision of a judge than the general public because of the importance of their esteem. Relationships with advocacy groups for women’s rights or against racial segregation, for example, can play an important role in the decision of some disputes. 36 There are also those circumstances in which the judge consciously projects his own political preferences when deciding a case. This decision model is often explored by American legal scholars, who have labeled it attitudinal model. 37 Its scholars argue that the very way of selecting judges in the U.S. constitutes an institutional factor that allows permeability between ideological convictions and judicial rulings. Many State judges are elected.38 Federal judges, including those who sit on the Su33 34 35

Cardozo, The Nature of Judicial Process (n. 1), 13. Posner, How Judges Think (n. 29), 107; Baum, The Supreme Court (n. 32), 120–121. Lawrence Baum, Judges and Their Audiences: A Perspective on Judicial Behavior (Princeton University Press: Princetown, 2006), 26–28. 36 Ibid, 60–63, 118–123; Baum, The Supreme Court (n. 32), 142–149. 37 Posner, How Judges Think (n. 29),19–29; Baum, The Supreme Court, (n. 32), 120–132; Judges and Their Audiences: A Perspective on Judicial Behavior (n. 35), 5–8; Cass Sunstein, David Schkade, Lisa Ellman and Andres Sawicki, Are Judges Political? An Empirical Analysis of the Federal Judiciary (Brookings Institution Press: Washington, 2006), 17–45; Saul Brenner and Joseph Whitmeyer, ‘Strategy on the United States Supreme Court’ (Cambridge University Press: New York, 2009), 11–18. 38 According to Posner, the concern with reelection may interfere with the judgment of such state judges, reducing their independence and making them more vulnerable to public opinion or

The Role of Precedents as a Filter for Argumentation

97

preme Court, are appointed by the President and approved by the Senate, after eventful hearings. The President’s selection of the candidate considers the professional’s ideological convergence with his political party and the perspectives of the candidate’s approval by the Senate, which tends to be an ideologically more heterogeneous group. 39 In such conditions, in which the very way of selecting judges attracts those professionals whose performance combine political and legal elements, it is claimed that one cannot expect that, in performing their activities, judges would be politically unbiased. If they were politically neutral, they would not have been selected in a process that is eminently political. The aforementioned scholars prove their point by analyzing the way in which judges vote on certain matters and comparing it to the orientation of the political party of the President that has nominated them. They demonstrate that judges appointed by Democratic presidents tend to have a voting posture that is markedly more liberal, whereas judges appointed by Republican presidents tend to be more conservative.40 Statistical surveys evidence the existence of legal matters that are highly susceptible to ideological voting, especially in constitutional claims, such as, for example, cases of sexual discrimination, sexual harassment, racial segregation, pornography, and abortion.41 This is so because constitutional decisions touch on fundamental moral and political issues, which provoke intense emotions and subjective reactions, even among the interpreters of the law. Constitutional norms, in turn, are vague, open to antagonistic interpretation and, therefore, may provide uncertain orientation to courts. The relevance of some issues and the lack of objective guidance leave space for the interference of subjective factors in the decision making process. 2. INSTITUTIONAL FACTORS The judges’ ideological tendencies can be intensified or mitigated according to the composition of the court in question because of three characteristics of collegiate deliberations: a) ideological polarization in politically homogeneous groups; b) ideological neutralization of a dissenting judge; and c) ideological moderation of the majority, the latter two in politically heterogeneous bodies, as a consequence of the phenomenon known as dissent aversion.42 Panels entirely composed of judges with the same ideo-

39 40

41

42

to the opinion of certain interest groups that support them. (Posner, How Judges Think (n. 29), 134–139). Posner, How Judges Think (n. 29), 57–59, 134–135. Sunstein, Schkade, Ellman and Sawicki, Are Judges Political? An Empirical Analysis of the Federal Judiciary (n. 37), 8. Although this ideological structure is indigeneous to the North American political culture, the argument itself, about the projection of political preferences in the decision of morally relevant cases, especially where the legal orthodox references are uncertain, is plausible in several other legal systems. Sunstein, Schkade, Ellman and Sawicki, Are Judges Political? An Empirical Analysis of the Federal Judiciary (n. 37), 8–61. This list of subject matters is based on U.S. law. Obviously, the most controversial issues may vary from one community to another since they relate to cultural moral disagreement. Sunstein, Schkade, Ellman and Sawicki, Are Judges Political? An Empirical Analysis of the Federal

98

Patrícia Perrone Campos Mello

logical profile tend to produce ideological polarization and, thus, decisions that are more liberal or conservative than those rendered by bodies with a heterogeneous composition. The union of homogeneous minds may lead to extremes. Ideologically heterogeneous bodies, in turn, are known to produce more moderate decisions. In the mentioned bodies, the dissenting judge acts as an antidote against the radicalism of the majority, either because he plays the role of the spokesman for different arguments, capable of moderating the understanding of his fellows, or by the majority’s interest in avoiding dissent, and, consequently, by its willfulness to adjust its understanding in order to accommodate some of the dissenting judges’ concerns. On the other hand, due to an aversion to dissent, the dissenting judge who sits in a heterogeneous court is usually ideologically neutralized, abstaining from directly manifesting his divergence. This attitude stems from several factors. First, many magistrates become upset when a fellow judge dissents from their decisions. The need to preserve a good relationship among colleagues who sit together in court for years may render them less prone to dissent, especially if said divergence does not interfere with the outcome of the case. And dissenting is demanding insofar it requires the elaboration of a vote combating the position of the majority. The judge who abstains from dissenting in such circumstances makes a cost-benefit analysis and only dissents in disputes where his belief is sufficiently strong to compensate for the attrition with the other judges as well as the increased workload. On the other hand, in dissenting with self-restraint, a judge hopes for reciprocal behavior on part of his fellows. There are cases, however, where ideological voting does not suffer the interference of other members of the panel, possibly because the intimate belief of the judge is sufficiently strong to bar external influences.43 Thus, depending on the issue, the vote of a judge may be evaluated exclusively on the basis of his ideological preferences or on the basis of the interaction of such preferences with those held by the other members of the court where he sits. These remarks apply more effectively to bodies composed of few members. As the number of judges in the group increases, the bargaining power of the dissenting judge decreases,44 and the effects of the collegiate decision become more difficult to identify, since they require an understanding of the dynamics of the interactions among several judges, as it is the case, in general, of a Supreme Court.45 Despite this, the interference of group dy-

Judiciary (n. 37), 8–9; Posner, How Judges Think (n. 29) 31–35. Brenner, Whitmeyer, ‘Strategy on the United States Supreme Court’ (n. 37), 53–58. 43 The formation of conviction with intensity to justify dissent is the result of a number of other variables that interfere on decision making, as psychological (emotional and personality) aspects and the background of the judges. Sunstein, Schkade, Ellman and Sawicki, Are Judges Political? An Empirical Analysis of the Federal Judiciary (n. 37), p. 10–13. 44 Posner, How Judges Think (n. 29), 57. 45 Ibid, 57. Furthermore, the dissent aversion tends to be lower in a Constitutional Court, what is justified by the following benefits of dissenting: a) it provides visibility; b) it addresses the concern of the Justices to present a coherent judicial philosophy; c) it is more likely to influence the development of law, because of the greater instability of Supreme Court precedents. (Sunstein, Schkade, Ellman and Sawicki, Are Judges Political? An Empirical Analysis of the Federal Judiciary (n. 37), 43–45).

The Role of Precedents as a Filter for Argumentation

99

namics in the decision making process of an individual judge of a Constitutional Court should not be neglected either. The courts are also sensitive to situations that could either undermine the courts’ stability or compromise the courts’ ability to perform their roles.46 Therefore, courts will avoid issuing decisions that they believe the Executive and Legislative branches manifestly will abstain from enforcing. The classic example of this type of situation is found in West Coast Hotel Co. v. Parrish. In that case, the U.S Supreme Court overruled the judicial interpretation consolidated in Lochner v. New York. 47 In Lochner, the Court had invalidated a law of the State of New York, which limited the number of working hours for bakers, on the basis that such law would constitute an undue restriction on the freedom to contract, initiating a period that was known as the Lochner Era, during which the Supreme Court annulled successive laws enacted in order to regulate labor relations. In order to address the negative effects of the Great Depression of 1929, which peaked in the United States in 1933, U.S. President Franklin Roosevelt approved a series of measures, known as the New Deal, which aimed at regulating and recovering the economy. However, the doctrine the Supreme Court established in Lochner stood in the way of such measures, as it was responsible for the declaration of unconstitutionality of important statutes enacted for such purposes. In reaction to these decisions, President Roosevelt proposed, in 1937, a federal judicial reform plan, which became known as the Court-Packing Plan, allowing himself to nominate six new Justices to the Supreme Court (who, obviously, would be chosen according to their ideological convergence with the measures the President intended to implement). In the same year, the Court ruled West Coast Hotel Co. v. Parrish, in which, by a narrow margin, it changed its former position on freedom to contract in an episode that became known as “the switch in time that saved nine”.48 Public opinion also constitutes a relevant element in the consideration of different points of view. It is capable of interfering with the institutional stability of the courts, with their power to politically influence the law, and also with the effectiveness of their decisions. The Judiciary is subject to the constraints of the political game. There is a limit up to which judges are willing to confront unpopular decisions. The acute and realistic risk of destabilization and demoralization could motivate a judge to bow to the public opinion, at the expense of his own convictions.49 46

Brenner, Whitmeyer, ‘Strategy on the United States Supreme Court’ (n. 37), 128–135; Baum, Judges and Their Audiences: A Perspective on Judicial Behavior (n. 35),50–87; Cass R. Sunstein. A Constitution of Many Minds (Princeton University Press: Princetown, 2009), 125–164; Luís Roberto Barroso, Constituição, Democracia e Supremacia Judicial: Direito e Política no Brasil Contemporâneo (n. 32), 33–46. 47 198 US 45 (1905). 48 Available at . Accessed on Jan. 29, 2011. 49 However, it is believed that such cases are very rare, because: a) a popular reaction able to interfere with the judicial conduct of the Court would need to be a negative reaction of great intensity; b) it can be difficult for judges to anticipate that a reaction of such magnitude will occur, as a result of their decision; c) even if there is a negative reaction, it will not necessarily generate a weakening of the Court’s position before the public opinion, especially if one considers the long term; d) even if such weakening occurs, it will not necessarily signify loss of effectiveness of the decision, unless compliance depends on public opinion or the public opinion is able to

100

Patrícia Perrone Campos Mello

In these cases, the courts tend to adopt a passive stance, under the following attitudes: a) dismissing the claim, where possible, b) adopting a posture of deference towards the Legislative, expressing the view that the matter should be decided by the democratic process, or c) deciding the dispute in a minimalist fashion, issuing a very particular ruling, bound to the case at hand, so that its impact is minimized. 50 On the other hand, the support of public opinion and its alignment with the tendencies fostered by the courts enhances jurisprudential shifts. According to Sunstein, many of the innovations of the U.S. Supreme Court were based on social verdicts that had been established before or that were already in process of emerging, therefore, the conduct of the Court has reflected such changes instead of promoting them. The author illustrates his point of view listing understandings shifts regarding freedom to contract, racial segregation, and women’s rights, among others.51 Judges also can craft their opinion in a strategic way, in terms that allow them to achieve, as best as possible, the decision they consider to be adequate for the case at hand, and generate the least possible resistance to it. In acting strategically, a judge or a court can issue opinions that do not correspond to their real convictions, or to the decision they consider to be ideal.52 To issue a decision upholding certain constitutional values to the largest possible extent, the judge has to consider the reaction that decision will produce, for example, on the other members of the court, the Legislature, or the general public. The need to influence others may lead a judge to deliver his opinion grounded on bases that are different from those that actually inform his conviction by using, for example, a legalistic argument in order to conceal his political orientation. If a judge’s goal is to influence the evolution of the law in a particular subject, he may have to moderate his decision so as not to provoke a reaction from the Legislative, either by the enactment of a constitutional amendment, or by the alteration of the composition of the Court. A similar situation may occur with respect to public opinion, if the issue at hand is able to mobilize it.53 E. JUDICIAL PRECEDENTS

AS A

FILTER

FOR

ARGUMENTATION

The above considerations show that, although the adoption of a theory of precedent is justified as a way of promoting values that are important to the law, such as legal certainty, equality, legitimacy, and efficiency, it faces significant limitations in pre-

50 51

52 53

influence an action of the Executive or Legislative; e) it is more likely that a Judge worries about legal issues and political goals than about a potential damage to the Court, that he does not even know if will actually occur; f) the difficulty to anticipate and valuate the consequences may lead the Court to disregard them; g) the Judiciary must be independent from the public opinion (Brenner, Whitmeyer, ‘Strategy on the United States Supreme Court’ (n. 37), 128–129; Baum, Judges and Their Audiences: A Perspective on Judicial Behavior (n. 35), 63–66). Sunstein, A Constitution of Many Minds (n. 46), 129, 135. Sunstein, A Constitution of Many Minds (n. 46), 140; Brenner, Whitmeyer, ‘Strategy on the United States Supreme Court’ (n. 37), 128–129; Baum, Judges and Their Audiences: A Perspective on Judicial Behavior (n. 35),134–135. Baum, Judges and Their Audiences: A Perspective on Judicial Behavior (n. 35), 7; Posner, How Judges Think (n. 29), 29–30. Baum, Judges and Their Audiences: A Perspective on Judicial Behavior (n. 35), 77–81.

The Role of Precedents as a Filter for Argumentation

101

serving these values. In fact, if one examines the reasoning of a decision, with the aim of identifying its holding, one finds a space of conformity with a considerable level of uncertainty. The same occurs in the definition of the holding’s reach, when its scope can either be reduced or enlarged, thus extending or reducing the holding’s impact on the development of the law. The decision whether or not to apply a precedent in a subsequent case can also be influenced by the judge’s subjectivity. The discussion and comparison of the relevant facts of both claims, the debate about the circumstances that could give rise to a different treatment may engender value judgments that lack safe guidelines. As already noted, a judge’s assessment of the facts and their relevance can be influenced by personal experiences. Finally, the decisions in leading cases and the overruling of precedents allow judges ample room in which to exercise their judgment. In the first case, judges unequivocally create law, voicing their opinion on new issues. In the second case, judges change the law, as expressed in previous decision, and also may determine the point in time when the new understanding will begin to produce effects. Such limitations do not weaken the value of a theory of precedent. As noted by Benjamin Cardozo in a passage quoted above, uncertainty is inevitable; it is inherent to the limits of language and to the impossibility to standardize all the facts of life in detail.54 Instead, a theory of precedent, further to contributing, to the extent possible, to the reduction of uncertainty, promotion of equality, legitimacy, and efficiency of the courts, plays another very important role: it operates as an argumentative filter, by offering guidelines for the selection of the issues that will be subject to debate. If the understanding of a particular legal issue has already been established, this circumstance will guide the parties to discuss the holding’s reach or to attempt to demonstrate its inapplicability to the case at hand because of its peculiarities. The parties will know that they will not succeed in discussing the ruling itself, which has already been determined, unless their arguments meet the requirements necessary to evidence that the precedent should be overruled. In the last situation, litigants will have the burden to explain why the previous standard is inadequate and shall demonstrate the reasons why other solutions are better. They shall discuss the positive and negative consequences of abandoning the old understanding and, lastly, shall demonstrate that the former outnumber the latter. In leading cases and absent other normative references, litigants shall try to: a) argue from principles that served as basis for previous holdings; b) challenge other principles that could possibly have some bearing on the case and that could produce negative outcomes; c) present an analysis focused on the problem, in its relevant facts; and d) draw on practical considerations of the values involved, customs, accepted social standards, and the consequences of the different solutions that could be applied to the dispute. In extreme situations, as when previous precedents are lacking, when they are overruled, or when the court makes inconsistent and arbitrary distinctions – in short, in any occasion when the principles previously stated by a court do not serve as basis for its decisions; and, additionally, when the judges fail to articulate other reasons to justify their conclusions, a theory of precedent will serve to evidence that 54

Cardozo, The Nature of the Judicial Process (n. 1) 161–163.

102

Patrícia Perrone Campos Mello

the decision is possibly motivated by other factors, which are alien to the orthodox legal material. This finding will enable a critical assessment of the works of the court, as well as the elaboration of a theory describing the decision making process, in which the elements that bear influence in its conviction are effectively represented, even though one understands that, from a prescriptive point of view, such process should be improved. The adequate description of the way the court operates is, once again, crucial to guide the litigants’ argumentation. If subjective and institutional aspects play a role in the decision rendered to some cases, the litigants should consider these factors in deciding whether to bring (or not) a dispute before a court and how best to argue their claims. Thus, binding precedents constitute a powerful mechanism in guiding the argumentation brought forth by litigants. They entail the exclusion of discussions on issues already settled, direct the debate to issues that remain open, indicate the type of argument that litigants may raise, depending either on the existence or on the absence of precedents, or, yet, on the need to have them overruled. They may also evidence subjective and institutional elements capable of interfering in the judicial behavior, enabling the development of a descriptive model of the decision making process and, most of all, providing litigants with sound references with which to uphold their arguments.

III PRECEDENTS

IN

CONTEMPORARY LEGAL CULTURES

VICTORIA ITURRALDE PRECEDENT

AS SUBJECT OF INTERPRETATION (A CIVIL LAW PERSPECTIVE)

“It seems hardly surprising that in common law systems a very large part of the discussion of the workings of precedent has historically been concerned with the problem of what is meant by the ratio decidendi of a case and how the ratio decidendi is to be established” 1

A. PRECEDENT

AND

CIVIL LAW

When one undertakes the application of law from the perspective of legal systems pertaining to civil law, one begins with the premise that legal rules come from the legislative power, that is, the parliament and the government, and that the sole function of the courts is interpretation and adjudication. For that, the only subjects of interpretation are the legal rules than come from the legislator In this way, it seems logical that legal theory has centred its attention almost exclusively on problems of the interpretation of legal rules and has omitted examination of interpretation of judicial decisions. As Visitini points out, we continue to proclaim ourselves interpreters of the statutory law and to elaborate theories about interpretation, but in fact we find ourselves operating more frequently as interpreters of the judgments.2 Even so, the distance between the two judicial systems is not actually established by the different role that case law has as source of law, but by a certain conception about what the function of judges should be. In this sense, it seems unavoidable to mention an article by Genaro Carrió3 published several decades ago which analyzed some aspects of jurists’ debates a propos the proclamation that “judges make the law”, showing (among other things) that there was not a disagreement between beliefs (that is to say, about what judges actually do), but instead a disagreement between attitudes (about what things should be preferred and how they should be done). The objective of this paper is to emphasize the need for analysis of judicial decisions as a subject of interpretation in civil law systems. This is necessary for two reasons: on the one hand, because these decisions are a formal source of law (similar to the legislation that comes from the classic legislative bodies); and, on the other hand, because they are a material source (that is to say, they have immense relevance in the determination of what law is, and they may have more weight than that which is generally attributed to them, as I will illustrate herein). In Spain, there is a widely shared view that judicial decisions are not a formal source of law. This comes from the insistence to continue using Article 3.1 of the Civil Code as a reference to what our system of sources is, disavowing (or better yet 1 2 3

G. Marshal, ‘What is Binding in a Precedent’ in: Interpreting precedents, ed. N. MacCormick and R. S. Summers (Ashgate: Dartmouth, 1997), 505–506. G. Visitini, ‘Il modelo Italiano’ in: Lo stilo delle sentencze e l´utilizzazioine del precedente, ed. L. Vida, (Giappichelli: Torino, 2000) 164. G. R. Carrió Notas sobre derecho y lenguaje (Abeledo-Perrot, Buenos Aires, 1976) 110–111.

106

Victoria Iturralde

silencing) the fact that since the Constitution in 1978 the system of sources is that which is dictated therein and that which is derived from its development, and additionally, omitting an obvious fact: that there are assumptions with which the courts make law in the strict sense of the term, in virtue of the standard of the order which establishes their obligatory power erga omnes. As has been illustrated, “we have seen that in countries, civil law precedents have greater importance than is officially recognized; and that the conception of precedent in civil law is distinct to that which governs in common law. Now, in some civil law systems, pathological situations occur which perhaps do not receive the attention they deserve as the normative value of precedents is not fully established in the manifest consciousness.”4 To sum up, we shall say that the courts’ developed legislative function is manifested in two different ways: through positive and negative legislation.5 A) Negative Legislation. This function occurs since the tribunals have the competence to eliminate laws from the legal order (a function similar to that which takes place when the legislature repeals a law). In Spanish law, this takes place through: (a1) The judgments of the Constitutional Court which considers an appeal against the alleged unconstitutionality of acts (sentencias estimatorias en un recurso de inconstitucionalidad) (articles 38.1 and 2 of the Constitutional Court Act and 164.1 of the Constitution); (a2) The judgments of the Supreme Court (and the Superior Courts of Justice of the Autonomous Communities, the National Audience, and the Contentious-Administrative Courts of Law) which considers a contentious administrative appeal against provisions (reglamentos) (articles 71 and 72 of the Administrative Procedure Act). Other questions aside, in a1 and a2, the result of the cited judgments is the annulment of certain legal rules. B) Positive legislation. This takes place when the courts create or modify pre-existing legal rules. This occurs by means of: (b1) The judgments which consider an “appeal in the interest of law” (sentencia estimatoria en un recurso en interés de ley), delivered by the Supreme Court (in the civil order) and by the Supreme Court and Superior Courts of Justice of Autonomous Communities (in the contentious-administrative order) (articles 490 and 493 of the Civil Procedure Act, and articles 100.7 and 101.4 of the Administrative Procedure Act). Article 493 expressly states that in the case of a summary disposition that is dictated in appeals with respect to the law, the court “will settle the jurisprudential doctrine in the verdict. In this case, it will be published in the Official Gazette of the State, and, after its introduction into the Gazette, it will complement the judicial order, binding in such a concept all the judges and courts of the civil jurisdictional order apart from the Supreme Court.” In the contentious-administrative realm the law states: “In the case of a summary disposition,…, the court will settle the legal doctrine in the verdict. In this case, it will be published in the ‘Official Gazette of the 4 5

V. Ferreres Comella, El principio de taxatividad en materia penal y el valor normativo de a jurisprudencia (Cívitas: Madrid, 2002) 223–224. I omit analysis of European Union Law and the erga omnes rulings of the Court of Justice of the European Union, which, should be incorporated in a complete analysis of this question.

Precedent as subject of interpretation (a civil law perspective)

107

State,’ and, after its introduction into the Gazette, it will bind all the judges and courts inferior to the degree of this jurisdictional order” (article 100.7). (b2) The Constitutional Court’s judgments (and orders -autos) in all types of processes (articles 13, 42.1, and 50.1 of the Constitutional Court Act; 161 and 164 of the Constitution; article 5 of the Judicial Power Act). Apart from the fact that there are decisions which have by their nature a particular scope (for example, an individual complaint (recurso de amparo) or a conflict of competencies (conflicto de competencias), the Constitutional Court understands that the doctrine contained in all of the judgments has erga omnes effect (among others the STC 150/1997). (b3) The application of equality under the law contained in article 14 of the Constitution. With regards to this, we must recognize that the Constitutional Court “in an original and ingenious construction, without a model in Constitutional Law”6 has continuously put so many limits on said principle that it has nearly been left completely aside. According to the Constitutional Court, for the principle of equality under the law to be found to be broken in the judicial organs’ application of the law it is necessary: a) that the plaintiff contributes the judicial resolutions that constitute the term of comparison, b) that the case does not deal with an isolated precedent; instead, that a jurisprudential line which reunites the notes of generality, continuity, and constancy must exist, c) that the case has to do with resolutions dictated by the same judge that was the author of the contested resolution, d) that an absolute similarity between the supposed facts of the resolution exists; that is to say, that is has to do with substantially similar cases, and e) that there is no justification for the criteria change.7 (b4) There are cases in which the law does not explicitly determine the scope of certain judgments, but the character of the source of law can be affirmed since the doctrine of the Supreme Court permits laying the groundwork of an appeal. As such, in Spanish law, this is the case when submitting: (a) an cassation appeal for “cassational interest” (recurso de casación por interés casacional) (civil order) before the Supreme Court and the Superior Courts of Justice of Autonomous Communities, when the appeal judgment “is contrary to jurisprudential doctrine of the Supreme Court” or “opposes the jurisprudential doctrine” (of a Superior Court of Justice) (article 477.3 of the Civil Procedural Act);

6

7

M. Gascón Abellán “Igualdad y respeto al precedente” (1993) 2 Derechos y Libertades 236. As Ruiz Miguel points out “considering that equality in the application of the law is one of the two facets of article 14… Although the court has been very careful and even excessively temperate in designing the scope of equal application of the law, which has not only delimited abstractly, but also, a more than reasonable delimitation, it has restricted to the point, in my opinion, of breaking the coherence of our judicial system and the constitutional doctrine itself.” A. Ruiz Miguel ‘La igualdad en la jurisprudencia del Tribunal Constitucional’ in El principio de igualdad, ed. L. García San Miguel (Universidad de Alcalá de Henares/Dykinson: Madrid, 2000), p. 186. E. Fernández Igualdad y Derechos Humanos (Tecnos: Madrid, 2003) 64–68; cfr. M. Gascón Abellán, La técnica del precedente y la argumentación racional (Tecnos: Madrid, 1993) 56–77.

108

Victoria Iturralde

(b) a cassation appeal “due to infringement of the norms of the judicial order or jurisprudence applicable to resolve questions being debated” (contentiousadministrative and social orders) (article 88 d) of Administrative Procedural Law and article 205 of the Labor Procedural Act, and (c) a cassation appeal for unification of doctrine (contentious-administrative and labor orders) before the Supreme Court and the Superior Courts of Justice in which contradiction between specific judgments is produced upon resolution of similar cases (article 217 of the Labor Procedural Act, and articles 96 and 98 of Administrative Procedural Act.).8 I do not intend to compare the Spanish system with that of Anglo-Saxon precedent; for the moment, I think that the information presented above is sufficient to demonstrate that in Spanish law there are judicial decisions with erga omnes scope and that the cases of summary disposition (b2, b3, and b4) have much in common with the precedent-based law. Along with the previous information, we must not obviate the importance of judicial decisions as a material source of law through their interpretation and application (or better yet, through the decisions that count as such).9 I am referring to those judicial decisions through which the courts exercise legislative function by modifying the applicable law (as there is no judicial norm that attributes this competence to them). To cite a recent example, on March 10th 2009, the STS, in relation to the application of the regulation of the ONCE (contained in a Royal Decree) according to which, and without any interpretive doubt, states that winning tickets may only be cashed in upon presentation of said tickets. So, the judgment says that: “the ruling recurred here (…) rejected the contentious-administrative appeal understanding that, conforming with the norms of the ONCE, one may only obtain a prize when the prized lottery ticket is presented, and that declaration, and conclusion, which was established by the Supreme Court upon hearing the cassation appeal interposed against the ruling on the 12th of May, 2004 of the Superior Court of Justice of Madrid, that had deposed the same doctrine, that is, that one can only obtain the prize by means of the opportune presentation of the winning ticket, has been left without effect due to the cited ruling of the Supreme Court on the 11th of June, 2006” (RJ\2006\6000), that its legal justification (fundamentos jurídicos) refers to the possibility to obtain a prize won on a misplaced ONCE lottery ticket “when it can be proven beyond the shadow of a doubt that the plaintiff acquired the lottery ticket, that is was misplaced, and that the prize has not been awarded.” It continues to state: “even though it is true that the regulation of the ONCE, concretely article 14 of the Regulations of the Lottery (Royal Decree 1200/99 of the 9th of July (…) expressly disposes that the prize is awarded upon the opportune presentation of the winning lottery ticket, which may not be substituted by any other document or testimony, we must not forget that on July 13th, 1990 and November 11th, 1994, this sentencing hall, in a case regarding the National lottery, which has similar regimen, had already made this doctrine

8

9

In the unusual cases the Spanish doctrine openly recognizes the binding character of jurisprudence (by means of cassation appeals for unification of doctrine and cassation appeals in the interest of the law) cfr. V. Ferreres Comella, El principio de texatividad en materia penal y el valor normativo de la jurisprudencia (Cívitas: Madrid, 2002) 160. P. Chiassoni ‘Il precedente giudiziale: tre esercizi di disincanto’ (2004) Analisi e diritto 96.

Precedent as subject of interpretation (a civil law perspective)

109

more flexible with respect to the supposed limits in which the purchase and misplacing of the lottery ticket and the fact that the prize money had not been collected could be proven…”. One might state something similar about the manipulative rulings of the Constitutional Court; that is, those that proceed to transform the meaning of legal rules, involving more than an interpretation conforming with the Constitution; or also the modification and integration of the dispositions subjected to examination, in such a way that they escape off constitutional process with a normative scope and a content different from the original content. 10 Those two aspects are sufficiently important so that with respect to civil law, the theory of judicial interpretation stops being exclusively a theory about the interpretation of legal rules, extending itself to the interpretation of judgments. B. DIMENSIONS

OF

PRECEDENT

I think that it is necessary to dedicate a few lines to the dimensions or aspects of precedent before speaking about it as a subject of interpretation, which will help us delimit the question that we will tackle later. Any discourse about the precedent implies taking into account these four dimensions: the institutional, the objective, the structural, and the efficacy.11 The institutional dimension points out that the idea of precedent is tightly related to the institutional organization and the existing authority relationships in the court system. The vertical precedent presupposes a hierarchical organization within the legal organs in which inferiors are obligated to follow the precedents of their superiors. The horizontal precedent exists when said obligation exists with respect to the decisions of judges at the same level. Finally, the self precedent is produced when a judge is obligated to follow his or her own precedents. The objective dimension refers to the determination of that which attributes efficacy to the precedent; in other words, what the precedent is in a strict sense. As a first approximation, one might say that the efficacy of the precedent is limited by the ratio decidendi, leaving aside the obiter dicta. The structural dimension alludes to the conformity of the precedent, taking into account the plurality of the courts, which can intervene in establishing the precedent itself. With respect to this, we can differentiate between four hypotheses: a) that only one precedent exists; b) that a plurality of precedents exists all of which give the exact same solution to the question at hand; c) that a plurality of inconsistent precedents exists; and d) that a situation of “jurisprudential chaos” exists, that is to say that there are many decisions, some of them contradictory, with respect to each individual question. The efficacy dimension refers to the nature and intensity of the precedent in a decision made in a prior case. Here, the discrepancies about the notion of precedent become clear as they are differentiated between broad and strict precedents. In a 10 11

Cfr F. J. Diaz Revorio, Las sentencias interpretativas del Tribunal Constitucional, (Lex Nova: Valladolid, 2001) 134. M. Taruffo, ‘Dimensioni del precedente giudiziario’ (1994) Rivista Trimestrale di Diritto e Procedura Civile 415–416. For an overall view cfr. R. Siltala, A Theory of Precedent (Hart Publishing: Oxford-Portland, 2000).

110

Victoria Iturralde

strict sense, precedent includes only the obligatory precedent; in the event that we must precisely state that it could be a system of strict obligatory character with respect to the precedent, one in which this obligatory character has no exception, or a system of defeasible or outweighable precedent (excepting some more or less demanding reasons, but excluding the possibility of freely not following a precedent). In a broad sense, the concept of precedent includes both obligatory precedents and persuasive precedents: obligatory precedents are those that the judge must follow even if he or she finds them unjust, unreasonable, erroneous, etc. Persuasive precedents are those that the judge should or may take into account without any obligation to follow them. 12 These different dimensions of precedent are not all equally important. In this way, while the objective and efficacy dimensions have to do with a conceptual question (what precedent is), the structural dimension has to do with problems that may occur in determining the precedent, and the institutional dimension has to with the determination of what judicial decisions constitute precedents (which can only be determined in relation to each judicial order). Although this is not the central idea of this article, a strict concept of the precedent including only the (named) obligatory precedents seems more adequate to me. In this sense, Schauer states: “The bare skeleton of an appeal to precedent is easily stated: The previous treatment of occurrence A in manner x constitutes, solely because of historical pedigree, a reason for treating A in manner x if and when A occurs,” “If precedents matters, a prior decision now believed erroneous still affects the current decision simply because it is prior, and thus an argument from precedent operates substantially like an argument from rule.”13 C. PRECEDENT

AS SUBJECT OF INTERPRETATION

In this section, I will point out the difficulties surrounding the precedent as a subject of interpretation and therefore the discretional leeway that the courts have when applying one. To analyze this, we must differentiate between two models of reasoning from precedent:14 12

In this broad sense, it is said that it is necessary to take into account a scale composed of various degrees of efficacy: (a) strong binding (or absolute binding) without exceptions (absolutely obligatory precedent); (b) strong binding, but subject to exceptions and limitations determined by the order itself (conditionally obligatory precedent); (c) strong binding in which the judge has the possibility to set aside the precedent if he or she has relevant reasons to do so (conditionally obligatory precedent); (d) weak binding is when it would be normal for the judge to follow a precedent but he or she need not justify setting it aside if he or she chooses to (persuasive precedent); (e) no binding is when the judge has total discretion as to whether or not to follow the precedent (in this case, it is not necessary to speak of a precedent), M. Taruffo, ‘Per un analisi comparata del precedenti guidiziario’ (1996) 6 Ragion practica 55–64. In a similar sense A. Aarnio, ‘I precedenti e loro validita’ (1996) 6 Ragion practica 9–18. 13 F. Schauer Playing by the Rules (Clarendon Press: Oxford, 1991) 182. 14 Z. Bankowski, N. MacCormick, L. Morawski and A. Ruíz Miguel, ‘Rationales for precedent’, in Interpretingn Precedents, ed. N. MacCormick and R. Summers (Ashgate: Dartmouth, 1997) 497. Said authors add another model to the two above: (c) the principle-exemplifying model, where the precedent, in relation to its own factual context, can be seen as exhibiting, and living support

Precedent as subject of interpretation (a civil law perspective)

111

(I) The rule-stating model, where some rule (ratio decidendi) is ascribed to the precedent as one which it is appropriated for latter courts to apply (unless they can be distinguished or overruled). (II) The model of particular analogy, where each case is simply treated as an illuminating example of a correct decision given all its own facts and hence a useful guide for decision in similar cases. Here the issue is: which are similar cases? These two ways of reasoning with respect to precedent create problems that are in part the same and in part different. In the particular analogy model, the fundamental question consists of the “description” of the cases and the decision as to whether or not sufficient similarity between them exists as to apply the decision made in the previous case to the case at hand. The fundamental problem of precedent in the rule-stating model is the formulation of the ratio decidendi, due to the fact that the decisions that constitute precedents are not formulated as general rules, and that what binds the judge is not the final decision (el fallo) itself but the ratio decidendi, which can only be obtained after the dispositive of the ruling and its relation to the judicial structure, the facts, and the decision.15 Taking this into account, I will consider the main problems that the determination of the precedent presents in each of the models. (I) THE RULE-STATING MODEL In this model, the principal issue is related to the determination of the ratio decidendi of the previous case and in the application of this case to the one at hand. 1. Ratio Decidendi The problems that the ratio decidendi brings up are the following: (a) the concept itself; (b) the difficulty in separating the ratio from the obiter dicta, (c) the method used to determine the ratio, and (d) the possibility that, with respect to the same judgment, different interpreters can find different ratios. If one takes into account the multiple existing definitions of “ratio decidendi,” one concludes that there is not uniformity in establishing said concept. As has been said, the different definitions of ratio decidendi are, in some cases, irreconcilable and present a large margin of vagueness. 16 According to Twinning and Miers 17, five uses of the term ratio decidendi have been dominant in the literature: to, some legal principle o principles that may be relevant for deciding future cases, and perhaps contributing to new legal developments. In the instant case, one looks to the precedents to see whether they contain or represent principles used for justifying the currently preferred decision.” 15 L. Vacca, ‘I Prededenti e i responsi dei giuristi’ in Lo stile delle sentenze e l´utilizazione dei precedenti, ed. L. Vacca (Giappichelli: Torino, 2000), 43. 16 P. Chiassoni, ‘Il precedente giudiziale: tre esercizi di disincanto’ (2004) Analisi e diritto, 75–101. More definitions in W. Twining and D. Miers, How to Do Things with Rules (Weidenfeld and Nicolson: London, 1991), 311–312. 17 W. Twining and D. Miers, How to Do Things with Rules (Weidenfeld and Nicholson: London, 1991), 311–312.

112

Victoria Iturralde

(a) The rule(s) of law explicitly stated by the judge as the basis for the decision, that is, the explicit answer(s) to the question(s) of law in the case; (b) The reason(s) explicitly given by the judge for the decision, that is, the explicit justification for the answer(s) given to the question(s) in the case; (c) The rule(s) of law implicit in the reasoning of the judge in justifying the decision, that is, the implicit answer(s) to the question(s) of law in the case; (d) The reason(s) implicitly given by the judge for the decision, that is, the implicit justification for the answer(s) given to the question(s) in the case; and (e) The rule(s) of law for which the case is made to stand or is cited as authority by a subsequent interpreter, that is, the imputed answer(s) to the question(s) of law in the case. Chiassoni organizes the different definitions to the following three meanings:18 (rd1)In the first sense, ratio decidendi is understood as the rule (the criteria, principle, or premise) upon which a case is decided. This rule can be the necessary and sufficient basis, the sufficient but not necessary basis, or the necessary but not sufficient basis of reasoning that leads to a judicial decision. (rd2)In the second sense, ratio decidendi is the rule the judge uses to justify his or her decision regarding a case including both the arguments that sustain it as well as the description of the facts to which it has been applied. (rd3)In the third sense, ratio decidendi is understood as any essential element of the argumentation that is necessary to arrive at a decision in a case. One of the difficulties in establishing the ratio decidendi of a decision comes from the fact that it is not easy to draw the dividing line between the ratio decidendi and the obiter dicta, especially because it is not generally expressly indicated by the courts. The distinction between ratio decidendi and obiter dicta is, in theory, clear: the ratio decidendi is a necessary rule for making the decision, such that unnecessary considerations made in the judgment are mere obiter dicta. That being said, the question is to determine the opinions which are “necessary” to make the decision.19 We can differentiate two different cases of obiter. 20 On the one hand, in a strict sense the obiter are those considerations that have no relation to the logical implication with respect to the supposed factual subject matter of the case. In this sense, the obiter dicta are totally irrelevant. On the other hand, there are manifestations that, even when they form part of the obiter, are the antecedent or the indispensible consequence of the judge’s interpretive election in a concrete case, in such a way that they are connected to the ratio decidendi by a prejudicial nexus or subordination. In this case, the connection between ratio and obiter is so narrow that one can consider both to be “precedent”. 21

18

P. Chiassoni, ‘Il precedente giudiziale: tre esercizi di disincanto, (n. 16),75–101 and P. Chiassoni ‘Il fascino discreto della common law. Appunti sulla “rilevanza dei precedenti giudiziali”’, in I metodi della giustizia civile, ed. M. Bessone, E Silvestri and M. Taruffo (CEDAM: Milano, 2000), 7. See also P. Chiassoni, La giurisprudenza civile (Giuffrè: Milano, 1999), 149. 19 J. Hardisty, ‘Reflections on Stare Decisis’ (1979) 55 Indiana Law Review, 58. 20 Cfr. F. Galgano, ‘L’interpretazione del precedente giudiziale’ (1985) Contratto e impresa, 701/ 21 S. Chiarloni, ‘Eficacia del predecente giudiziario e tipologia dei contrasti di giurisprudenza’, in La giurisprudenza per máxime e il valore del precedente, ed. G. Visintini (CEDAM, Padova 1988), 71.

Precedent as subject of interpretation (a civil law perspective)

113

In addition to the previously discussed questions, we add another: how can the method of determining the ratio decidendi of a case be established? One of the classic criteria in this area is Goodhart’s thesis, according to which the method for finding the ratio decidendi consists of determining the material facts as seen by the judge and his conclusion based upon them. We can compare this theory with other relevant opinions (Wanbaugh, Allen, Gray, Montrose), which claim that the ratio decidendi may be found in “the reasons given for the decisions by judges, or on the rules on law enunciated by the court.” Finally, there may be different opinions as to what the ratio decidendi of a given case is; that is to say, different interpreters may deduce different ratios for the same decision. The paradigmatic example of this is the case of Donoghue vs. Stevenson in which we can obtain different ratios based on the same decision. 22 On the one hand, this occurs because there tends to be discrepancies over the solution to the three problems we mentioned before; and, on the other hand, because the courts rarely expressly establish upon what ratio the judge resolves the case. “It but rarely if ever occurs – as Stone points out – that the previous decision has explicitly stated: “Out of the facts (a)–(j) appearing in this case the material facts as I see them and upon which I am solely basing my conclusion are A, B, and C”. Yet unless it does this there is no logically valid method whereby a later court, looking only at that single case, can decide with certainty that the facts A, B and C are “material”, while facts D–J are not. The question – What single principle does a particular case establish? has for these reasons been said (and correctly so) to be ‘strictly nonsensical, that is, inherently incapable of being answered.” 23 2. Plurality of reasons in a decision Judges can give two or more reasons as the basis of their decisions. The question is whether all of them together constitute the ratio decidendi or if it is only one (or several) of them. This may occur when a single person makes the decision as well as if a decision is made through collegiate appellation; but this question becomes more important in this second case, as the precedents are normally a creation of this type of organs. Among the special difficulties which may attend the doctrine of stare decisis in the case of an appellate tribunal of multiple members, four stand out: 24 - First, those which may arise when the court is composed of an even number of judges and these are equally divided. - Secondly, those which arise where the court (or its majority) reaches the same conclusion for different and inconsistent reasons. - Thirdly, those which arise when the court (or its majority) bases its conclusion on two or more reasons which are consistent and cumulative, and the question 22

W. Twining and D. Miers, How to Do Things with Rules (Weidenfeld and Nicholson: London, 1991), 316; J. Stone, Precedent and Law (Butterworths: Sidney, 1985) 86–87, 231–232. 23 J. Stone, ‘The Ratio of the Ratio Decidendi’ (1959) 22 The Modern Law Review 618. 24 Lord Asquith, ‘Some Aspects of the Work of the Court of Appeal’ (1950) 1 Society of Public Teachers of Law, 357 and 357–358; R. Cross and J. W. Harris, Precedent in English Law (Clarendon Press: Oxford, 1991), 91–96.

114

-

Victoria Iturralde

is where some one or more of these reasons, and if so, which are part of the ratio decidendi, or whether some, and which, are obiter dicta. And fourthly, those which arise where a court bound by its own previous decisions finds these irreconcilable either with each other or with the decision of a superior court. 25

3. Plurality of precedents Due to the dynamic character of the law, in certain cases, in order that a precedent be considered as such, a decision about a case is not sufficient; but, beyond that, certain stability with respect to that decision is required, which can only be obtained through application of said ratio by the courts during a more or less extended period of time. The following hypotheses can occur when there are successive decisions about the same question: Firstly, that the distinct decisions provide the same solution; this is called “constant jurisprudence”. We must, however, take into account that “constant jurisprudence” elaborated during a brief period of time may be the beginning of actual “constant jurisprudence”, but also may represent an ephemeral tendency destined to be abandoned. On the contrary, a dilated “constant jurisprudence” may express a perfectly consolidated orientation, but may also indicate an orientation destined to be modified. Secondly, there may be several conflicting precedents. This situation could have to do with two decisions or with two conflicting jurisprudential lines which could have been declared by the same tribunal or by one or more different ones. In these cases, a legal rule requiring a tribunal to intervene in order to assure the uniformity of jurisprudence may or may not exist, and for this reason the prevalent jurisprudential orientation may not be clear. In either case, the judge should make this decision as he or she is the one who must justify it by giving his or her reasons for following one precedent and ignoring another. Thirdly, a situation of “jurisprudential chaos” may arise; that is to say, much dispersion and diverse and incoherent tendencies exist on the part of the same tribunal. A jurisprudentially chaotic situation propitiates that the successive decisions make reference to these precedent decisions that, paradoxically, become more comfortable since it is theoretically possible to find a precedent for any decision (ignoring the diverging or conflicting precedents). This represents an incorrect use of precedent. 26

25 26

Cfr. examples in V. Iturralde, El precedente en el Common Law (Cívitas: Madrid, 1995), 94–100. M. Taruffo, ‘Dimensioni del precedente giudiziario’ (1994) Rivista Trimestrale di Diritto e Procedura Civile, 415–416.

Precedent as subject of interpretation (a civil law perspective)

115

4. Application of precedent or distinction With the ratio decidendi established, another problem arises upon its application in subsequent cases. This is a problem because, just as occurs with the interpretation of legal rules, whether or not the new case forms part of said ratio. (If it does not, then it should be distinct from the case, prima facie being applicable). The technique for making distinctions may be correctly employed to determine if precedent X is not applicable to the case at hand since, as occurs with the application of a rule, the assumption made does not fall under the application of X. In this case, nothing more than an interpretation of the ratio decidendi in relation to the facts of the case occurs. But the technique for making distinctions can also be used to avoid the application of a precedent that is in fact applicable to a case if the solution is considered unjust, incorrect, etc. As Zander 27 declares, sometimes the court distinguished the undistinguishable as the way to escape from the clutches of an unwelcome precedent which would otherwise be binding. Strictly speaking, we must say that if the technique for making distinctions is correctly employed, it serves as a reason not to apply a ratio to a case to which it is not applicable (in the same way that on many occasions a rule is not applied to an individual case because it does not fall in rule’s realm of application). 28 5. Decisions without justification Evidently, the motivation for judicial decisions is the key to determining the ratio decidendi. The motivation is, apart from a legal obligation in many legal orders, an inherent requirement for the function of jurisprudence and its practice in democratic states. Now, beyond mentioning the generic obligation to explain rulings, we must establish when a judgment has been justified. According to Igartua29, a judicial 27

M. Zander, The Law-Making Process (Weidenfeld and Nicholson: London, 1989), 257. Cfr. K. Llewellyn, The Common Law Tradition (Little, Brown: Boston, 1960) who ready list of sixty-four techniques of following and avoiding precedents decisions. 28 Apart from the technique for making distinctions, there are two other ways to ignore a precedent: refusing to follow it and repealing it. 29 In this classification I follow J. Igartua, La motivación de las sentencias: imperativo constitucional (Centro de Estudios Políticos y Constitucionales: Madrid, 2003), 202–207. The omission of motivation includes two possibilities: formal omission and substantial omission. Formal omission occurs when the decision consists only of the verdict without any sign of motivation. This is the biggest and most frequent defect. Substantial omission includes those cases in which there is motivation but it is only apparent. It includes three possibilities: (i) partial motivation, (ii) implicit motivation, and (iii) motivation per relationem. Motivation is partial when some decisions that prepare the final decision; in this way, while it is habitual to account for questions of “law,” it is not so habitual to do so with questions of fact and with the determination of the judicial consequence(s). Motivation is implicit when the judge does not state the reasons for his or her decisions. In this case, one assumes that the judge has made the decision based on his or her own prior decisions. Motivation by remission occurs when the judge, upon making a decision about some controversial issue, does not elaborate an autonomous justification ad hoc but remits the reasons contained in another decision. The cases of motivation by remission are many. There are cases called motivation “matrooksa” in which the decision yields the motivation of another decision, which in turn may be remitted to that of a third ruling, and so on. Others

116

Victoria Iturralde

decision lacks motivation in the following cases: a) when total omission of motivation occurs, b) when the motivation is insufficient, and c) when the motivation is contradictory. Against this point of view, the Spanish Constitutional Court’s allows “motivation by remission” and “implicit motivation”; this makes it more difficult to determine the ratio and gives a great deal of discretion to successive courts when applying precedent.30 The same takes place when the Constitutional Court delivers “interpretative decisions not reflected in the final decision” (sentencias interpretativas sin reflejo en el fallo) 31 that is to say, those decisions that lack certainty in the decisum of the interpretive operation realized on the basis of the legal justification. This would not cause a major problem if in the verdict the judge clearly indicated what the declared constitutional interpretation was, as opposed to the declared unconstitutional interpretation. In a dissenting opinion for STC 59/20008 on the 14th of May regarding the constitutionality with respect to Article 153.1 of the Penal Code, judge Javier Delgado Barrio states: “My discrepancy has to do fundamentally with the final decision, …, given that the decision does not expressly declare its interpretative character, a simple reading of the non favourable pronouncement at hand can bring one to the conclusion that the cited precept, with only the content that they expressly obtain, results in accordance with the Constitution. And in the same judgment, judge Rodríguez Zapata states that “…the decision does not wholly comply with the function of an interpretive decision as it does not clearly and precisely delimit the interpretation compatible with the Constitution, nor does it expound on the reasons for which the decision was made, nor is it reflected in the verdict, and it is not even stated if all or some of the constitutional precepts invoked are those that prohibit the interpretation so apodictically stigmatized. These characteristics, along with the use of the dubitative expression ‘in effect, would lead towards its constitutionality’ create an unacceptable ambiguity, as the subject used for an interpretive decision should, precisely, clear up the doubts as the unconstitutionality, forbidding thereafter the application of the rule that results incompatible with the Constitution for successive situations within the judicial organs.”

exist in which the object of the decision to which the judge yields is not the true ratio decidendi of the decision invoked, but an affirmation given in case of need that is not strictly pertinent to the object of the trial (clearly obiter dicta). And finally, there are cases in which the judge yields directly to the reason for the decision whose impugnation constitutes precisely the object of the appeal. In this sense: - Motivation is insufficient when the judge does not offer the necessary reasons for an election among two or more alternatives; - Motivation is contradictory in the following situations: (ci) there is a contradiction between the verdict and the motivation, (cii) there is a lack of connection between the verdict and the arguments contained in the motivation, and (ciii) the motivation contains in itself contradictory arguments. 30 Cfr C. Zoco Zabala, Igualdad en la aplicación de las normas y motivación de las sentencias (artículos 14 y 24.1 CE) (Bosch: Barcelona, 2003), 106–107; A. J. Gómez Montoro, ‘El derecho a una resolución motivada y congruente en la jurisprudencia del Tribunal Constitucional’, in La Constitución y la práctica del derecho, T. I (Aranzadi: Pamplona, 1998), 509–516; 496. 31 A. Garrorena Morales, ‘Opacidad y desestimación de la inconstitucionalidad en el fallo de las sentencias interpretativas’ (2000–2001) 12 Anuario de Derecho Constitucional y Parlamentario 151.

Precedent as subject of interpretation (a civil law perspective)

117

6. Interpretation of the concepts employed in the ratio decidendi Once the ratio decidendi is established, we must determine what the meaning of its concepts is since said concepts do not constitute exact notions but terms accompanied by a degree of vagueness. This is the same as the problem with interpretation of the legal rules (despite scepticism on the part of some authors)32. (II) THE MODEL OF PARTICULAR ANALOGY. 1. The construction of the facts of the case This is the main problem in determining the precedent. Reasoning based on precedent is different from that which is based on rules because in the case of precedent, to form the generalization (that already exists in the case of rules) is necessary. When the description of the facts becomes part of what the precedent constitutes, the argument based on precedent is made similar to that based on rules. 33 The problem that arises when attempting to determine what the facts are is that there is not one unique way of describing the facts of a case, because the world is not divided into intrinsic, immutable categories; it is delimited by artificial categories created by people and variables of time and space. In the words of Schauer, they are like glaciers, in constant movement in spite their apparent immobility. 34 One of fundamental questions with respect to this is determining which among all of the facts are fundamental and which are not. As to this, Goodhard establishes as a criteria to abide by what has been established by the court, and if they do not do not make this distinction, this criteria indicates that all of them should be considered fundamental.35 2. Sufficient similarity in cases As no two events are exactly alike, the point is to lay down the likeness or the rule of relevance. Because of the first (precedent) case contains no generalization, the decision-maker in the second case creates the generalization at the time, in light of the circumstances of the instant case as well as those of the precedent case. The “rule of relevance” is not a matter of fact, but a determination of meaning with respect to the similarity (or lack thereof) between the cases and the relevance or irrelevance of the similarity. Obviously, that which is implicit in the relevance trial is the judge’s opinion as to the justice appropriate for treating the two cases the same way.

32

M. B. N. Sinclair, ‘The Semantics of the Common Law Predicates’, (1986) Indiana Law Journal 373–399. 33 F. Schauer, Playing by the Rules (Clarendon: Oxford, 1991), 245, 248. 34 F. Schauer, ‘Precedent’ (1987) Standford Law Review 585. 35 Cfr A. L. Goodhart Essays in Jurisprudence and the Common Law (Cambridge University Press: Cambridge, 1937), 25–26, 9–22.

118

Victoria Iturralde

3. Plurality of the previous cases Here, we can repeat what we pointed out in section A) 3 in the sense that the judge may find himself/herself with various previous cases that are similar to the one at hand but have different solutions. In this case, he or she must choose one of them.

ZHANG QI ON THE METHOD OF SEARCHING OF TRIAL EXPERIENCE

FOR

GUIDING CASES

ON THE

BASIS

On October 26, 2005, the Supreme People’s Court of the People’s Republic of China released “the Second Five-Year Reform Outline for the People’s Courts”, proposing to regulate and improve the system of ‘guiding cases’. Since then the practice of guiding cases have been developed significantly in the Chinese court system, and the method of using guiding cases is attracting more and more attention of Chinese judges. Obviously, the precondition for implementing this institution is finding the appropriate guiding cases, which is not an easy job. We always need to make great effort to find the guiding cases we need.1 The questions follow: how can we find the guiding cases we need? Which cases are the appropriate guiding cases? If we agree that similar cases should be decided similarly, what criteria can we rely on to judge (or evaluate) the similarity between cases? How can a judge in practice judge the similarity between a pending case and a guiding case, deciding whether to refer to the guiding case? During the August 2007, Lang Gui-Mei, Huang Bin, both members of the Institute of Applied Jurisprudence of Supreme People’s Court, conducted a special research with me. We investigated three high courts, three intermediate court and five district courts. Based on these field investigations, this article will focus on the questions proposed above. I will use the descriptive approach in this article, studying the experience of judicial practice as objectively as possible. In the meantime, I will also use some normative approaches, trying to propose some theoretical thinking and suggestions. On the one hand, this article aims to study and summarize the practical experience of Chinese court on this issue. On the other hand, this article will analyze and discuss the most appropriate legal method of implementing guiding cases system, hoping to benefit the legal practice of China. A. FINDING GUIDING CASES

THROUGH THE

SIMILARITY

BETWEEN

CASES

1. THEORETICAL FOUNDATION OF JUDGING THE SIMILARITY BETWEEN CASES The searching for and finding of appropriate guiding cases is a process of rational thinking. If we have the proper method, we will get twice the result with half the effort. Obviously the proper method requires an appropriate theoretical founda-

1

I shall give a preliminary definition of the guiding cases here to focus the legal method of them. All cases that are instructive or referential to the trial of cases are defined as guiding cases. I have argued that the guiding cases are instructive and accessory in nature and the definition here is consistent with my argument. The intension and extension of guiding cases have significant relationships with the search for guiding cases, which are discussed in my article “On the ‘Guidance’ of Guiding Cases” published on Law and Social development, vol. 6, 2007.

120

Zhang Qi

tion. In the matter of judging the similarity between cases, I suggest this theoretical foundation to be analogical reasoning. Analogical reasoning means applying legal rules prescribed for constitutive element A to the similar constitutive element B, which is not explicitly prescribed by law; it can also mean applying binding rules or jurisprudence of constitutive element A, which are prescribed by the guiding case, to the similar constitutive element B (of the pending). Since the pending case and the guiding case have similar constitutive elements, the law or jurisprudence applied to the guiding case can also be applied to the pending case (generally). The foundation of analogy here is the similarity between the constitutive elements of two cases. The similarity between the two cases leads to the similar judgment.2 What is the similarity between constitutive elements between two cases? How can it be judged? These are the key questions. To answer these questions, we should first of all make clear the meaning of constitutive element. By the virtue of the German jurist Karl Larenz’s study, the constitutive element here refers to an important view relating to law’s special evaluation of certain issues.3 Larenz gave such an example: the law does not explicitly prescribe that the legal liability of a seller if he lies about the merit of the subject matter which does not exist, but the law does prescribe the legal liability of a seller if he does not inform the buyer about the flaw purposely. How can we determine the legal liability in the former case? We can compare these two cases and then decide whether we can apply the rules of the latter case to the former case. The two cases are similar to a certain degree. The sellers in the two cases both realize that the buyer’s mistake the quality of the subject matter and use this false recognition to negotiate a contract. Here, purposely using the false recognition of the buyers is the key of the law’s evaluations, which is also the important view relating to law’s appraisal of certain issues and which can be called the constitutive element. The law pays little attention to the fact that the buyer does not inform the seller intentionally, or that the seller does not make necessary instructions to the buyer, or that the seller lied to the buyer. The law’s evaluations of these circumstances do not have fundamental differences, although the latter one is certainly more serious. According to the principle of equal disposal, the legal rules set down for the latter case should also be applicable to the former case. A constitutive element consists of special case facts, thus, the analysis of case facts and legal relationships which are closely relate to it, and are of significant importance in comprehending constitutive elements. The comparable case facts that judges focus on are the facts which have legal bearings and which can become the object of legal evaluation, especially those key points or controversial issues by which law can determine the nature of case facts.4 This is what we call the constitu2 3 4

See Karl Larenz, Legal Method, Trans. by Cheng Ai’e, The Commercial Press, 2004, p.258. Ibid., p.258. For example, is it an infringement of intellectual property by providing links to download MP3 at will? How can we determine the compensation fees if it is an infringement? A judge expressed similar meaning by proposing “fundamental facts”. The “fundamental facts” is similar to or the same with the key point of case facts mentioned above to a certain extent. They think the fundamental facts should be determined to identify characteristics of cases when considering them, only by identifying the characteristics of cases can we determine the similarity between cases. The process of identification is a process of case searching and the process of analyzing

On the Method of Searching for Guiding Cases on the Basis of Trial Experience

121

tive elements. On the one hand, we have to be sure about the similarity between the elements of the facts of the pending case, and of those of a decided guiding case. On the other hand, we have to be sure that the differences of the two cases do not lead to the different legal evaluation.5 In my opinion, the following factors are of great help when determining the constitutive elements, and furthermore, judging the similarity between cases. 2. SEVERAL FACTORS WHICH HAVE GREAT IMPORTANCE WHILE DETERMINING CONSTITUTIVE ELEMENTS

THE

1

5

Determine constitutive elements with the help of the object of litigation. The object of litigation is to resolve disputes between parties, and legal evaluation is part of the constitutive elements mentioned above. Therefore, the accurate understanding of the object of litigation and the comparison or analogy between the object of litigation of the pending case, and the guiding case can help us evaluate the similarity between cases. The following two cases show a detailed method of analogy based on the object of litigation. - Two administrative cases concerning marriage registration which were disposed alike. In one case, the wife sued against the marriage registration office, claiming for the repeal of their marriage registration. The cause of the action was the absence of the parties during the marriage registration. The statutory law did not have explicit rules for such a case. The judge found a similar case on the internet, which also claimed for the repeal of the marriage registration because of the absence of the parties during the marriage registration. In that case, the court in Wuhan judged the marriage registration to be legal and maintained the original marriage, considering the absence of parties to be an administrative flaw. The judge in Chengdu showed and explained the case to the parties, and subsequently, quashed the indictment. In these two cases, the parties and places were different, but they were both administrative cases claiming for the repeal of the registration of the marriage of parties who were absent at the time of that registration. In the meantime, the two cases had same main facts and same legal relationships. Therefore, the constitutive elements of the two cases were the same and could be disposed alike, the legal outcome applied to the former case could also be applied to the latter. - Two administrative cases concerning divorce registration which were disposed differently. In a pending case, a man did not have full civil capacity and his mother represented him to appeal for the repeal of his divorce registration. Again the law did not have explicit provisions for such an indictment. The judge searched for a case in which the court agreed to hear the indictment for the repeal of the divorce registration. In the pending case the pleadings of the parties were reconsidering the reallocation of the property, and the parties fundamental facts can not take place without the help of studying and interpreting legal rules. Legal principles will help judges to analyze and determine case facts. Ibid., Karl Larenz, Legal Method, pp. 258–259.

122

Zhang Qi

did not have any dispute over the divorce issue. In the preceding case, however, the court agreed to hear the indictment, maintained the divorce registration, and reversed the verdict of property allocation. The judge who handled the pending case thought these two cases were different because the dispute of the pending case was the personal relationship and the former case was it property relationship.6 In my opinion, the differences of the dispute led to different constitutive elements in each of the cases. Although both the cases consisted of a repeal of the divorce registration, the divorce registration related to personal relationship was not in dispute in the second case. What was in dispute, and demanded legal disposal, was property allocation. Hence the constitutive elements were the property allocation caused by the divorce. By contrast, the dispute in the pending case was the relationship of marriage and the constitutive element was the verdict of legality of divorce registration. Because of the different case facts and different legal relationships of the object of litigation between two cases, the constitutive elements were different. The legal outcome applied to the former case was not applicable to the pending case. 2

The influence of damage and fault on constitutive element. In a case claiming compensation because of the birth of a seriously disabled baby, the plaintiff sued the hospital for not being able to detect the congenital missing of hands during the prenatal ultrasonic examination, which ultimately led to the birth of a severily disabled baby. The judge searched the relevant cases in China and found a case in which the hospital paid compensation for not being able to detect the defect of the foetus during the prenatal examination.7 However, in that case, the newborn baby’s left forearm and right crus were completely missing, its right hand had only three fingers and it’s left foot had only two toes, and the hospital actually missed it when doing the b-ultrasound examination. Based on the relevant legal provisions and literatures, the judge decided that, first of all, the hospital bore fault in the former case because they should have detected the abnormallity of the foetus and the b-ultrasound examination is enough. However, in the first case, the hospital did not violate any law or regulations of doctors and did not have any fault. Moreover, the two cases had different ‘case facts’ and did not have enough similarities with each other, since in one of them the hands where missing while in the other what was missing was (almost) the entire arm. Personally I agree with the judge who handled the case, since here the difference with regards to the extension of the abnormality had led to the difference of quality between the case facts. Also, according to relevant legal provisions or regulations, whether the doctor was negligent was only part of the case facts. In conclusion, the two cases did not have sufficient similarity and the legal outcome applied to the former case was not applicable to the pending case.

3

The influence of declaration of wills of the parties on constitutive elements. In some cases, the declaration of wills of the parties has significant influence on the constitu-

6 7

The guiding cases is from People’s Judicature (Case Version), vol.4, 2007, p.40. China Trial, vol.8, 2006, pp.29–30

On the Method of Searching for Guiding Cases on the Basis of Trial Experience

123

tive elements. For example, in a dispute concerning house buying contract, the appellant was a property developer and the respondand of the appeal was a house buyer. The appellant asked the respondand to pay the residual fees of the house. The respondant, namely the house buyer, thought the unpaid fees were related to the payment of the public area to be shared. He thought that he needn’t pay these fees according to a preceding case which involved the public area to be shared. In that preceding case, the house buying contract incorporated the public area to the comprising construction area and the buyer didn’t know this when he signed the contract. Only when it came to the payment of residual money did the buyer realize it and refused to pay the fees of the public area. The court thought the contract was a form contract, one that should be interpreted in the interests of consumer, especially when the provisions were not explicit, or when the buyer was not meant to pay the fees of public area. The appellee thought this pending case was the same with the preceding case and the unpaid fees were for the public area, thus the case should be decided liked the preceding case. The appellate court decided that this case was different from the preceding case, because the buyer in the pending case had acknowledged the residual fees which consisted of fees of public area. Since he wrote an IOU which acknowledged the fees of public area, the pending case became a case concerning relationship of debtor and creditor and could not be decided like the preceding case. In my opinion, these two cases certainly dissimilar to each other, especially from the perspective of the constitutive element. The constitutive element of the preceding case was the disputable fees of the public area, but the constitutive element of the pending case was also the dispute about the creditor’s rights debt. The key difference between the cases was the IOU written by the house buyer and the acknowledgement of the fees of the public area. This IOU showed no dispute over the existence of the fees of the public area between the parties. Since the IOU was reasonable and had legal validity, the dispute that followed was a problem of delivering the goods and refunding. Therefore, the preceding case and the pending case had different constitutive elements and were not sufficiently alike. The preceding case could not become the guiding case or referential case to the pending case. 4

The influence of the purpose of law, and the intention of the parties’ action on determining the constitutive element. Objects of litigation, damages, faults of the parties and intentions of the parties are all factors that can assist in determining the constitutive element. However, sometimes these factors will still not be able to help us determine it. Under these circumstances we have to understand law’s critical evaluation of certain important issues, or return back to the purpose of legal regulation and legal principle, drawing a conclusion based on weighing and comparison of the parties’ action. In the meantime, the evaluation of the parties’ action is not a pure mechanic description; the exterior facts are not enough and we have to inquire into the parties’ intentions in order to analyze the nature of the action. For example, to compare two trademark tort cases concerning the use of geographic place name, we have to understand the purpose of legal provisions and the inten-

124

Zhang Qi

tions of the parties’ certain actions. We have to be clear that the purpose of legal rules is the protection of intellectual property. But what is the purpose of using a special geological landmark? Is it used to signify a geographic location, or confuse people of the object related to intellectual property? We can find the different purpose of using place name: in one case, an actor used a place name for the purpose of signifying a building’s geological location;8 in another case, an actor used a place name for the purpose of misleading the consumers to another wine trademark which also contained a place name and enjoyed widespread fame.9 Because of different purposes of using place name, the constitutive elements of the two cases had significant differences and the legal evaluation of these two cases also had great differences. The former case did not infringe the intellectual property while the latter did. Moreover, sometimes we can estimate similarity between cases with the help of the cause of action. Case facts are objective and legal relationships are somehow subjective. When facing the absolutely objective case facts, people always judge their legal relationships according to certain purposes. When we judge the similarity between cases according to legal relationships, we can judge that more swiftly and conveniently if we classify legal relationships further. The cause of action is the name of the case, and it reflects the nature of legal relationships. It is also the court’s conclusion of the legal relationships of a case.10 Once the cause of action is determined, the legal relationships of the case are determined to a certain degree. For example, in the hearings of administrative cases, an important characteristic of these cases is that most of them involve certain administrative authorities, and these cases have special administrative legal relationships, such as legal relationships of industry and commerce administration or legal relationships of finance administration. Judges could find guidance in similar preceding cases based on the cause of action. With the help of cause of action, the facts of cases and legal relationships can be more easily compared, thus the constitutive elements and similarity between cases can be determined. On April 1, 2008, the Supreme People’s Court put the “Regulations on the Cause of Civil Action” into force, which would help judges determine the issue of litigation and the similarity between cases. However, when we try to determine the constitutive elements with the help of the cause of action, we have to remind ourselves that the cause of action depends on different factors, and that specific cases under the same cause of action could feature various constitutive elements. Therefore, in most cases the case of action has only limited functions and we would better consider it to be guidance for the determination of constitutive elements, and not relying on it too much. B. THE FUNCTION

OF

VALUE JUDGMENTS

IN THE JUDGING OF

CASE SIMILARITY

The use of guiding cases presumes that we cannot get the outcome of a case by deductive reasoning of a provision of statutory law or by inductive reasoning. There8 9 10

Gazette of the Supreme People s Court of the People s Republic of China, vol.10, 2005, pp.45–46. Gazette of the Supreme People s Court of the People s Republic of China, vol.2, 2007, pp.46–48. See Notice of the Supreme People’s Court on Regulations on the Cause of Civil Action, People’s Court Daily, March 3, 2008, page 3. Some judges said that they found similarity by the cause of case and used the whole case for reference.

On the Method of Searching for Guiding Cases on the Basis of Trial Experience

125

fore we have to select similar cases among many preceding cases to guide the pending case based on analogical reasoning. As a type of the dialectic reasoning, analogical reasoning emphasizes the evaluation of the essential content of legal rules and case facts. Thus the value judgment has critical importance in judging similarity between different cases. Under some circumstances we can determine the constitutive elements of a case and judge the similarity between cases with the help of studying the actors’ declaration of will, the intentions of actions, damages and faults of actors. Under these circumstances, the function of value judgment is not obvious. But under other circumstances, value judgment has a distinctive function in judging the similarity between cases. The following circumstances are several conditions under which value judgments influence the judging of case similarity. 1. “DIRECTIONAL SELECTION” OF SIMILARITY BASED ON VALUE JUDGMENT There are two conditions when ‘directional selection’ of similarity is based on value judgment. Firstly, according to value judgment, although there are different cases with different outcomes, the judge selects a case, the value judgment of which is in accordance with his own values. For example, a district court in Nanjing heard a case of injury in the park. The key point of the issue in this case was whether the park, which charged fees, had the obligation of security protection. If the park had the obligation of protecting the safety of tourists, then the park should also compensate for the injury. The judge searched and found several similar cases in China, which had different verdicts by different courts. For example, in one case a man sought compensation from the May Flower Restaurant in Guangzhou. He had been dining with his family and was badly injured when a wine bottle on the next table exploded. The Guangdong High People’s Court made a judgment partly in favor of the plaintiff and the restaurant was asked to award damages to the plaintiff. In another case, a tourist was murdered in the Imperial Palace but the court in Beijing denied the request for compensation. The judge who handled the case in Nanjing thought the park should award damages to the injured from the perspective of justice, because the park had the obligation of protecting the safety of tourists. Under the guidance of such value judgment, he thought the case in Guangdong was similar to the pending case and awarded full compensation to the plaintiff. Secondly, according to value judgment, the preceding cases which fulfill legal loopholes are considered to be instructive and similar. The determination of similarity between cases and the instructions of the guiding cases show a kind of causality. That is, because of the similarity between the pending case and the preceding case, the disposal of specific issue is instructive to the pending case. However, these two points tangle with each other and sometimes may ‘put the cart before the horse’. A case is not instructive because of similarity, on the contrary, a case is ‘made to be’ similar because of its guidance. The value judgment is of great importance in this process. The judge who handles a particular case is inclined towards a certain legal outcome – an outcome which is found in the preceding case – and then proves the similarity between the preceding case and the pending case. In this way the pending case would be judged under the guidance of the preceding case. When the judge

126

Zhang Qi

searches for a similar case, he will first of all search and judge the specific nature of preceding cases. As for the specific nature of preceding cases, they are the fulfillment and development of the loopholes of law. In other words, it is the study and determination of how the preceding cases implement the law and whether the decisions in preceding cases fill legal loopholes or amend legal rules. If the preceding cases fulfill legal loopholes or amend current legal rules, such cases may be considered as cases with instructive functions. France does not acknowledge precedent to be a kind of legal resource, but judges could implement precedents to develop or change legal provisions.11 In China there are also practices of developing statutory law with the help of legal precedents, which may help judges in selecting guiding cases. This is especially true in administrative litigation. Since the legal provisions cannot cover every issue and there are some legal loopholes, judges who handle administrative litigation still have to make great effort to look into high-level law and apply low-level law according to legislative purpose of high-level law. For example, when applying ‘Regulation on Work-Related Injury Insurances’ and ‘Measures for the Ascertainment of Work-Related Injuries’ to work-related injury cases, the disposal of work time, causation and place exceeds the prescription in specific legal rules to a certain extent, but it accords to the legislative purpose of law and administrative regulations concerning the disposal of work-related injuries, which aimed at protecting workers. Such cases do a great help to judges who would hear similar cases. For example, a district court in Nanjing heard a case of new type in 2007. A worker was killed in a train crash. The “Regulation on Work-Related Injury Insurances” did not mention whether the death on the way to or from the job could be recognized as work-related injury. The labor office judged it not to be a traffic accident and could not be compensated by work-related injury standard, because a train is not an automobile according to the “Road Traffic Safety Law”. People’s Court Daily once published a similar case of Jingniu district court in Chengdu, Sichuan province. According to the legislative purpose of the “Regulation on Work-Related Injury Insurances” and in the interests of protecting workers’ legal rights, the court made a liberal interpretation of automobile and judged the train to be a kind of car although the train is not the automobile listed in the “Road Traffic Safety Law”. Therefore, the judge maintained the verdict of compensating by work-related injury standard made by the labour office. The judge in Nanjing took into consideration the fact that the workers, as a disadvantaged group, should be protected specifically, so the scopes of application of the “Road Traffic Safety Law” and the “Regulation on Work-Related Injury Insurances” were different. Thus he referred to the case in Sichuan and awarded compensation to the death’s family by the standard of work-related injury. Here we can see the method of disposing preceding cases by Chinese judges implementing the value of law. We can go deeper into this question by thinking the questions proposed by the American jurist Edward Levi. The question is “under

11

For example, under the condition of remaining the statutory law, the legal disposal of automobile accident can shift from the standard of fault liability to the standard of strict liability by implementing precedent. See Martin Shapiro, Courts: A Comparative and Political Analysis, Trans. by Zhang Sheng, Li Tong, China University of Political Science and Law Press, 2005, pp.199, 201.

On the Method of Searching for Guiding Cases on the Basis of Trial Experience

127

what circumstances is it justifiable to see the unlike cases alike?”12 Fundamentally speaking, the evaluation of similarity between cases depends on people’s thinking and conclusions of the justification of the problem, which is based on value judgement. It is necessary to mention that sometimes preceding cases may not be considered to fulfill legal loopholes when they judge issues which are not prescribed in explicit provisions. We need to analyze and determine whether the preceding cases reflect legislative purposes and legislative principles correctly. Only those cases which correctly reflect legislative spirit, legislative purpose and legislative principles are considered to have fulfilled legal loopholes. This work is a challenge. Some judges have proposed the following factors to be considered when facing these conditions: causes of action; facts, intentions of litigations hidden behind cases, grounds of judgment, analyses of facts, advanced academic theories and relevant legal provisions, etc. 2. THE CONSTITUTIVE ELEMENTS WITH VARIOUS FORMS – SIMILARITY BETWEEN APPARENTLY DIFFERENT CASES As for those guiding cases which fulfill legal loopholes, it is the cases of constitutive elements that have caused the problem of ‘legal fulfillment’ of loopholes. Such a kind of constitutive element has many differences from those of cases mentioned above. Because sometimes it enables the guiding cases to have influences on apparently different kinds of cases. For example, in the “Jia Guoyu v. Longkou Kitchenware product” case which was published in the Gazette of the Supreme People’s Court, the court awarded the damages for mental suffering because of a product’s defect, for the first time in the Chinese judicial history.13 This guiding case fulfilled the legal loopholes of compensation for mental suffering because of a product’s defect. It is instructive not only to those seeking compensation for mental suffering caused by product’ defects, but also to those caused by other kinds of torts. In 2001, Yunnan High Court heard a case concerning medical accident. In this case, the applicant was studying cosmetology at a beauty parlour and her appearance was harmed by a beautician because of the defendant’s intervention, the victim claimed for compensation for mental losses distinct from medical fees and repair fees, which were all supported by the trial court. The defendant appealed and the re-trial court referred to Jia Guoyu Case to demonstrate the justness of compensation for mental suffering in the ‘notice of dismissal’, thus stopped further appeals. In this case we can see various forms of constitutive elements, which have played a key role in judging the similarity between cases. Sometimes constitutive elements emerged as fundamental facts of cases, which are also disputes between parties and objects of legal evaluation, such as: declarations of wills of the parties, actions of the parties’ torts, damages caused by actions, faults of the parties, objective matters of parties’ actions, litigious claim, etc. At other times, the constitutive element emerged 12

Edward Levy, An Introduction to Legal Reasoning, Trans. by Zhuang Zhong, China University of Political Science and Law Press, 2002, p.6. 13 Gazette of the Supreme People’s Court of the People’s Republic of China, vol.7, 1997.

128

Zhang Qi

as an issue in a case settled by law. For example, in the case mentioned above, Chinese law and regulations did not have consistent, comprehensive and explicit rules about mental suffering caused by different kinds of torts, but in the real world such kinds of damage do exist, and compensation for such damages was required by justice. If there was a guiding case which made explicit decisions on this issue, such as in the Jia Guoyu case, then we could consider the constitutive elements to be similar between the preceding case and the pending case, and that the preceding case is instructive because the two cases were the same on the point of causing mental suffering, even though other aspects of the cases were different. Although the pending case and the guiding case had apparently different case briefs, the judge could still find similarity between the two cases, and also, direction from the guiding case. During this process, the value judgment of the judge showed the direction. 3. SELECTING GUIDING CASES BY CONSIDERING CHIEF JUDGE’S IDEA Sometimes, the selection of guiding cases is a process of searching actively for affirmative cases which justify the judges’ legal opinion after the preliminary understanding of the legal nature of the pending case. Under these circumstances, the judge of the pending case and the judge of the guiding case have a similar idea and a similar legal interpretation, exerting great influence on the searching, selecting and determining of the similar case.14 This is a process of active facts searching, case comparing and point concluding. To judge the similarity between a preceding case and a pending case, and to decide the guidance of preceding case, judges need to be “convinced by the essential holdings which support the verdict” under this kind of circumstances.15 In my personal experience, this is especially true in the hearings of intellectual property cases. Many intellectual property cases’ plaintiffs initiate lawsuits in more than one place. Usually a case is still on the docket while the same cases in other courts have already been decided. The plaintiff may come to a court carrying the decision in his favour and then initiates a similar lawsuit to another defendant. Under these conditions, when a judge was considering whether he should follow or refer to the decisions in other courts, he was not considering the similarity between cases, but the trial idea of the judge who handled the preceding 14

In a certain sense, it is a process of searching what one needs. As a judge’s reply to my question has noted, (even) the case facts were different, they would accept the case to be guiding case if it had similar ideas. In an administrative case heard by Kunming Intermediate Court, the administrative counterpart was punished for making a mobile phone call when driving a car, however, he refused to accept the punishment and sued the police office. The focus of the issue is whether he made phone calls when he was driving. In the hearing process, it was hard to provide evidence and the fine in this case was little. In the first trail, the verdict was in favor of the plaintiff. The defendant provided the appeal court with similar cases trailed by courts in Beijing while the appellate judge also found the case provided by the defendant on the internet. Based on the consideration of public interest, the appeal court decided to refer to the case in Beijing and rescinded the original judgment. The value judgment here led the judge to think the guidance of the Beijing’s case to the pending case. 15 P. S. Atiyah & R. S. Summers, Form and Substance in Anglo-American Law: A Comparative Study of Legal Reasoning, Trans. by Jin Min, Chen Linlin, Wang Xiaohong, China University of Political Science and Law Press, 2005, pp.96, 97.

On the Method of Searching for Guiding Cases on the Basis of Trial Experience

129

case. In other words, they tend to consider whether the preceding cases’, that is, the “trial idea of the judges,” can convince themselves, although the outcomes may not be the same.16 It may seem too arbitrary and too personal in a certain degree when judges who handled cases select guiding cases by considering the “trial idea” of judges in the preceding cases. There are reasons for Chinese judges to use this kind of method to judge the similarity between preceding cases, and the pending case. The guiding cases in China are different to case law in common law countries which adopt the principle of stare decisis. The Chinese guiding cases are not strictly binding on future cases, their actual influence is mainly persuasion and the parts that are convincing are the legal argumentation in the preceding case. 17 If a judge thinks the law application (the fulfillment of legal loopholes) and legal argumentation in a preceding case to be convincing, he will consider the case to be instructive and will then go on to study the issue of similarity. However, this kind of practice may cause the indeterminacy of the validity and effect of guiding cases to a certain degree and may not be good for the unification of the application of guiding cases in the whole nation. There are tensions between the substantial rationality of the application of guiding cases and the determinacy and unification of the application procedures and outcomes. We should take measures to solve these problems as much as possible. C. CONCLUSION: SELECTION

AMONG

SEVERAL GUIDING CASES

1. THE PROCESS OF JUDGING SIMILARITY If we have to conclude the search of guiding cases or referential cases, namely the methods of judging the similarity between cases, the following steps are necessary:

16

The judge mentioned above thought that judges had to consider the consistency of trial idea when looking for guiding cases. If the trial idea of the preceding case was not consistent with the pending case, judges did not have to follow the preceding judge because of different case brief. Even if there were seemly similar cases, we still did not have to consider the outcome. A judge in Jiangsu expressed similar attitude of a case of patent infringement heard by Nantong Intermediate Court. An obligee sued a patent infringement case to a court in Zhejiang and the court held that the infringement was established. However, judges in a court in Jiangsu did not think they should follow the preceding case in Zhejiang because they had “different opinion on this” despite of the same case facts. “The decision in (Zhejiang) court is a personal opinion”. The judges in Jiangsu even thought that the adjudication should be submitted as references, because in this way the judge would not have to respond and explain it in the verdict. 17 A judge in Nanjing pointed out that, when disposing cases concerning real estate and labor dispute, the guiding case which had same case facts and holdings and different claims with the pending case could be considered as similar case and were instructive to the pending case. Anyway, it’s hard to find two cases which were completely the same. We can find the point of instructiveness between the seemly different pending case and guiding cases. For example, in the disposal of house contract concerning product defects of houses, there maybe many plaintiffs sued to the same house developer in the same residence community. The claims could also be different; but the outcomes may be identical. In these conditions, the judge of the pending case should pay attention to the holdings of the preceding case and value preference and consideration of interest, trying to comprehend and abstract the point of legal application.

130

Zhang Qi

Firstly, find out all case facts and understand case briefs comprehensively. This is the precondition of knowing case facts correctly. The case facts and case briefs mentioned here consist of case facts of the preceding case and the pending case. It should be noted that the description of the judgment of case facts is not a completely objective process. When the judge who handles the case analyzes and judges the case facts, he wears ‘colored spectacles’, that is, he looks into facts by the instruction of law. What has happened and how dit it happen? What is the result of the event? The ‘facts problem’ and the ‘legal problem’ cannot be dichotomized completely. As the German jurist Larenz pointed out, a judge should consider the possible (legal) meaning of individual facts before the case facts come into being. “The root of difficulty is: before asking whether something happened, we have to describe ‘something’ in some way. It could be described by a general diction or legal diction. Once we use the legal diction, there already exists the influence of legal judgment. Many ways of expressions are shared by legal diction and everyday expression, however, such a kind of expression in legal diction has accurate meaning only in few “critical cases”.18

Secondly, judge the legal relationships and nature of preceding case and pending case respectively. Sometimes, it is hard to determine the nature of some complex, hard or new cases, the nature of which has many possibilities. Judges need to comprehend the details of case facts. Only in this way can judges determine the legal relationships the cases could possibly belong to or relate to. The details of case facts mention here refer to: declarations of wills, actions of the parties, purposes of actions, results of actions, subjective faults of the parties, contract objects and performance of contracts in civil litigation, etc. Thirdly, determine the constitutive elements of the pending case, and the preceding case through their case briefs, legal relationships, concrete case facts and objects or petitions of litigations, and then compare the constitutive elements of the two cases. Fourthly, determine the applicable legal provisions, guiding cases or instructive cases according to the outcomes of constitutive elements comparison. 2. DETERMINE THE KEY OF SIMILARITY BETWEEN CASES AND THE SUBSTANCE OF CASES’ CONSTITUTIVE ELEMENTS An important process of judging similarity is the determination of constitutive elements. I have discussed several factors which would help determine them in part one of this article. In part two, I discussed the detailed influence of value judgment when judging the similarity and determining the constitutive elements. The constitutive elements mentioned here is consistent with comparing the points mentioned by the German Jurist Arthur Kaufmann when comparing cases by analogical reasoning.19 The further questions are: what is decisive on our determination of comparing points or constitutive elements? What made one thing but not others to be the constitutive elements or comparing points? This is the substance of constitutive elements and also the key of deciding case similarity. 18 19

Ibid., Karl Larenz, Legal Method, pp. 160, 187. Arthur Kaufmann, Philosophy of Law, Trans. by Liu Xingyi, Law Press, 2004, p.116.

On the Method of Searching for Guiding Cases on the Basis of Trial Experience

131

The Harvard law professor Cass Sunstein once discussed the following five steps of analogical reasoning: “(1) Some fact pattern A (of the preceding case) has a certain characteristic which we may call X, Y, and Z; (2) Fact pattern B (of the pending case) has characteristic X, Y and A, or X, Y, Z and A; (3) The law treats A in a certain way; (4) When analyzing A, B and their relationships the principle of why A should be treated that way is found or set up;(5) Because B shares certain characteristics with A, the law should treat B the same way”.

Finding the principle as to why the preceding case (which is equivalent to the guiding cases in this article) should be treated that way, Sunstein argued the fourth step to be critical.20 Sunstein’s opinion is identical with that of Joseph Raz, an Oxford University professor. However, it seems the analysis of Raz goes much deeper. Raz pointed out the need for determining important similar points, different points and unimportant ones. Raz’s opinion is “the answer depends on the rationale of rules in precedent,21 the purpose of legal rules.”22 The rationale (rationality or substantial reasons) refers to purpose and value that can justify the legitimacy of rules in the precedent. “The substance of analogical reasoning is, if a certain reason is a fine justification of a rule, then it could also justify another similar rule.”23 Judicial procedures settle cases and convince the parties by providing justifications. The reason for using guiding cases to instruct the pending cases is the lack of explicit provisions for a special issue (constitutive elements) and the available legal outcomes provided by guiding cases. The special issue could be a certain object of litigation, a certain declaration of will, or a combination of certain damages and faults of actors, which were discussed in the first section of this article. It could also be a certain unresolved issue discussed in the second section, such as the obligation of safe protection in the public place which charged fees, the compensation for mental losses caused by torts, etc. They have common beliefs in the following issues: people believe the rationality or legitimacy of legal solution (certain rule’s or jurisprudence) in guiding cases, or in other words, people acknowledge the purpose or value which justifies the rules or jurisprudence in guiding cases. Because of the similarity between the special issue in the pending case and the special issue in the decided case, the justification of purpose or value of the legitimacy of rules or jurisprudence in the guiding case will also justify the application of such outcome to the similar pending case. Therefore, the key of judging the similarity between cases is the determination of the rationality or essential reasons of legal outcomes in the guiding case. In other words, it is the determination of purpose or value that justifies the rule or jurisprudence in the guiding case. However, the determination of the substance of constitutive elements or case similarity is not a judgment that is abstract, universal or fundamental, it is an “incompletely theorized judgments; and principles operating at a low or intermediate

20

Cass Sunstein, Legal Reasoning and Political Conflict, Trans. by Jin Chaowu, Hu Aiping, Gao Jianxun, Law Press, 2004, pp.77, 78. 21 The rationale mentioned here is identical to the “essential holdings” mentioned in Form and Substance in Anglo-American Law written by American jurists Atiyah and Summers. 22 Joseph Raz, The Authority of Law: Essays of Law and Morality, Trans. by Zhu Feng, 2005, p.176. 23 Ibid., p.177.

132

Zhang Qi

level of abstraction.”24 The determining process is based on judgment and casuistry to a great extent and depends on the exertion of power.25 Therefore, we need to use certain procedural and method to standardize the exertion of power. 3. PROCEDURAL AND METHOD – THE SEARCH FOR BALANCE BETWEEN FORM AND SUBSTANCE In a sense, law is a system of formal rules which try to avoid subjective, arbitrary contingency and indeterminacy and unpredictable in daily lives. When discussing the relationship between the form and substance of law, the American jurist P. S. Atiyah and Robert Summers correctly point out that the rule of law cannot take place if there is not enough authority or formally coercive law in a legal system. Actually, the legal system may not be available under that circumstance. Therefore, the credibility, determinacy, predictability and unification of application of guiding cases are important factors of making full use of guiding cases. From the perspective of the practice of guiding cases in China, since guiding cases are not binding, the procedure of searching and deciding guiding cases is influenced by value judgment to a great extent. The positive side of this practice is the avoidance of applying guiding cases stiffly, which may cause substantial injustice. However, this could also cause indeterminacy, unpredictability and inconsistency of the application of guiding cases.26 Therefore, for the health development of guiding cases and the practice of the rule of law principle, it is necessary to form some formal principles, procedures or methods of searching and deciding guiding cases. In Chinese judicial practice, most Chinese judges are willing to solve problems with formalistic procedures or methods. The reason is that Chinese judges should answer to their superior leader such as chief judges, presidents of people’s courts and the principals of political-legal committees and adjudication committees when applying guiding cases. If the application of guiding cases is much more of a formalistic and procedural process, then it is comparatively easy to handle or even prevent interventions from all around. As Atiyah and Summers’ have stated, “the function of formalistic warrant is like a barrier, in one way or another it insulates the deciding process from the substantial warrant which have not been integrated to the rules.”27 Drawing lessons from France which grants different authority to different judicial precedent, I would like to propose several preliminary suggestions to the selection among different cases: (1). The authority of guiding cases can be judged according to the office which released the guiding cases.28 The guiding cases released by the Supreme People’s Court have higher authority than those released by people’s high courts.

24 25 26

Ibid., Cass Sunstein, Legal Reasoning and Political Conflict, p.80. Ibid., Arthur Kaufmann, Philosophy of Law, p.116. Ibid., P. S. Atiyah & R. S. Summers, Form and Substance in Anglo-American Law: A Comparative Study of Legal Reasoning, pp.20, 21, 22, 23. 27 Ibid., p. 2. 28 Jacques Ghestin & Gilles Goubeaux, Traite’ de droit civil, Introduction generale, Trans. by Chen Peng, Zhang Lijuan, Shi Jiayou, Law Press, p. 427.

On the Method of Searching for Guiding Cases on the Basis of Trial Experience

133

(2). The authority of guiding cases can be judged by the date when the cases were released. Among the cases which released by the same (level) courts, the latest one has highest authority because it may express the preference of the court mostly. (3). The authority of guiding cases can be judged by the special field of guiding cases. Cases released by some professional courts, such as courts specialized at intellectual property, have higher authority. For example, the Beijing High People’s Court hears intellectual cases of which sue against the State Intellectual Property Office, thus the guiding cases of intellectual property released by the Beijing High People’s Court have higher authority. (4). The authority of guiding cases can be judged by the key issue of the pending case. If different professional courts released different guiding cases on a same kind of issues, then the guiding case released by the court which charged the key facts of litigation would have higher authority. In conclusion, there are various ways that Chinese judges find guiding cases. The fundamental method of selecting guiding cases is based on analogical reasoning and comparison of case similarity. At the same time, value judgment has great influence on judging case similarity. Also, the fulfilment or development of legal loopholes by the guiding cases or instructive cases, the consideration of trial idea of judges in the preceding case, all have significant influence on the searching for and selecting of guiding cases. In the process of searching and determining cases, there are tension between the determinacy, unification and substantial rationality, or the tension between formalistic rationality and substantial rationality. The determination of certain procedure and methods of selecting among different guiding cases will ease such tensions and avail the application of guiding cases.

EWOUD HONDIUS PRECEDENT REVISITED A. INTRODUCTION In 2006, the International Academy of Comparative Law discussed the topic ‘Precedent and the law’ at its XVIIth conference in Utrecht. This paper will highlight some of the findings of that conference and supplement these with a number of new developments. The national reports for the Utrecht conference, together with the general report, have been published in a single volume.1 This volume has been favourably reviewed by Gabriel García Cantero, one of the national reporters.2 Some of the reports have also been published separately, often in a more elaborate format.3 B. WHAT’S

IN A NAME

What is ‘precedent’? Any discussion of a legal topic usually begins with an – often lengthy – analysis of the precise subject-matter. In this paper I just want to take up one issue. Maurice Adams has argued that the term precedential force is mostly used as a kind of catch-all phrase covering two – in his opinion – conceptually separate notions, i.e. court decisions as ‘precedents as such’ on the one hand, and what he proposes to call the ‘gravitational force’ of court decisions on the other hand.4 Although, as we shall see below, I agree with Adams that there is a wide gulf between common law and civil law where precedent is concerned, this paper prefers to use the term ‘precedent’ to cover both situations, chiefly because the notion of precedent is so universally accepted, whereas ‘gravitational force’ reminds one more of Newton’s laws of physics.

1

2 3 4

Ewoud Hondius (ed), Precedent and the law (Bruylant: Brussels, 2007), 517 p., with national reports from Australia (Michael Kirby), Belgium (Maurice Adams), Brazil (Thomas Bustamante), Canada (Catherine Valcke), Colombia (Carlos Bernal Pulido), the Czech Republic (Zdenek Kühn), Denmark (Joseph Lookofsky), England (Simon Whittaker), Finland (Jaakko Husa), France (Philippe Malaurie), Greece (Paul Sourlas), Hungary (Béla Pokol), Israel (Daniel More and Anat Carmy Wiechman), Italy (Michele Taruffo), Japan (Emi Matsumoto), Korea (Sang Yong Kim), Lebanon (Joseph Chaoul), Macao (Ignacio Castelluci), the Netherlands (Olav Haazen), Poland (Malgorzata Krol), Serbia (Dusan Nikolic), Spain (Gabriel García Cantero), the United States (Mortimer Sellers) and including a report on international tribunals (Michael Likosky and David Sugarman). Gabriel García Cantero, ‘Civil law “versus”common law …, o viceversa?’ (2009) Revista jurídica del notariado, 411–440. For instance Michael Kirby, ‘Precedent law, practice and trends in Australia’ (2007) 28 Australian Bar Review, 243–253. Adams, in: Precedent and the law (n. 1), p. 150.

136

Ewoud Hondius

C. SETTING

THE SCENE

Precedents do not come out of the blue. In assessing their importance, the whole organisation of the judiciary and perhaps even the legal system is of the essence. These matters cannot be dealt with at length in this paper, but they should at least be alluded to. First, it does make a difference whether we are dealing with a decision of the United Kingdom’s House of Lords – recently replaced by a Supreme Court – which in 2002 handed down no more than 72 decisions (the number has not grown from 2002–2012), or with a decision coming from countries such as France, Italy and Spain, which in that year handed down 32 296, 63 534 and 28 444 decisions (the number has not decreased from 2002–2012) respectively.5 In such jurisdictions, precedent value is sometimes only given to a case, when its dictum has been repeated at least twice.6 The large numbers have the consequence that the courts need special procedures to ensure uniformity (assemblée plénière, Großer Senat). Other jurisdictions have special rules for creating precedents.7 Special rules have in some jurisdiction also been adopted to overrule earlier precedents.8 Large numbers of decisions are also produced in the United States, where the special situation is that all 50 states have their own legal system, the decisions of which do not formally bind courts of other states. Second, the style of judgments is also widely diverging. It ranges from the infamous single sentence decisions in France to the lengthy debates with academic writing and other case-law in Germany and bringing up policy arguments especially in common law jurisdictions (as against the legalistic and deductive style prevailing in civil law countries). Then, case reporting differs substantially from one legal system to the other. D. COMMON

AND CIVIL LAW DIVIDE

There are nowadays many comparative law projects which point to the fact that René David’s division of the world’s legal systems into five or seven distinct legal families no longer is relevant. And indeed some publications, such as those emanating from the common core of European private law project, indicate that this division is no longer of interest. As far as precedent is concerned, the dividing line between common and civil law, however, is still very important. At first glance, precedent and common law on the one hand and civil law and non-precedent seem to be natural allies.

5 6 7 8

Louis Vogel (ed), Cours suprêmes: comment le filtrage des recours révèle le pouvoir des juges (LGDJ : Paris, 2005), at 107–109. Bernal Pulido, in Precedent and the law (n 1.), p. 319. Sang and Castellucci, in Precedent and the law (n. 1.), p. 339 and 347. Castellucci, in Precedent and the law (n. 1), p. 347.

Precedent Revisited

E. HAS

137

THIS ALWAYS BEEN THE CASE?

It may come as a surprise to some readers, but historically there has not been a natural link between common law and precedent, nor has civil law always been without precedent. In England, the binding force of precedent was only accepted in the late nineteenth century. According to one author, ‘prior to the mid- or late nineteenth century, judges in England did not regard themselves bound by earlier decisions (…).’9 According to the thinking of the time of Bracton in 1256, ‘the law was not to be found in individual cases; rather the case decisions in their totality were a reflection of the law.’10 It was – among other things – the rise of the printed text and the setting in place of a hierarchy of courts, which in the sixteenth century combined to shift the nature of legal reasoning in the common law.11 Other elements which contributed to the rise of stare decisis were the growing custom of giving reasons for decisions and the improvement in the quality of law reporting. Only in 1898 was the binding force of precedents accepted,12 In continental Europe, the development has been completely the opposite. From a historical perspective, the era of ius commune was one in which there was a system of precedents on the European continent.13 With the French Revolution14 and the entry of the codification principle on the continent, the system of precedents was all but abolished. The Prussian Code of 1794 held that ‘auf altere Aussprüche der Richter bei künftigen Entscheidungen keine Rücksicht genommen werden (soll)’ (in future cases, old decisions shall not be taken into consideration).15 In France and the countries with a system of cassation, courts only had to take into account another court’s decision after a renvoi by the Court of cassation.16 Precedents were something of the past, at least in theory. Civil codes did not mention case-law;17 only constitutions might do so.18 In practice, case-law always remained of interest, especially after wellknown academics started to write annotations.19 In many jurisdictions, a change towards a growing use of precedents came because of external influences, such as the introduction of the European Convention on Human Rights.20 9 10

11

12 13 14 15 16 17 18 19 20

J. David Murphy, Robert Rueter, Stare Decisis in Commonwealth Appellate Courts (Butterworths: Toronto, 1981), at3. Thomas Lundmark, ‘“Soft” stare decisis: the common law doctrine retooled for Europe’, Richterrecht und Rechtsfortbildung in der Europäischen Rechtsgemeinschaft, ed. Reiner Schulze, Ulrike Seif (Mohr: Tübingen, 2003), 161–168. David J. Ibbetson, ‘Case-law and doctrine: a historical perspective on the English common law’, , in‘“Soft” stare decisis: the common law doctrine retooled for Europe’, Richterrecht und Rechtsfortbildung in der Europäischen Rechtsgemeinschaft, ed. Reiner Schulze, Ulrike Seif (Mohr: Tübingen, 2003), 27, 33. London Tramways Co v London County Council, [1898] AC 375. J. Drion, ‘Stare decisis/Het gezag van precedenten, in: Verzamelde geschriften van J. Drion’ (Kluwer: Deventer, 1968), 142–170. Malaurie in Precedent and the law (n. 1), 139. Verzamelde geschriften van J. Drion (n. 13), 152. Ibid., 147 (this is in fact slightly more nuanced). Sang in Precedent and the law (n. 1), 337. Adams in Precedent and the law (n. 1), 149. In the Netherlands Eduard Maurits Meijers and Paul Scholten, who both became annotators for the Nederlandse Jurisprudentie, were of major importance in this regard. Husa in Precedent and the law (n. 1), 267.

138 F. THE

Ewoud Hondius COMMON LAW: TOWARDS MORE FREEDOM?

What does the future hold in stock for us? Are the common law and the civil law converging and is it simply a question of time before Central and Eastern Europe’s backlog is overcome? It seems as if in England the very rigid adherence to stare decisis has slightly declined over the past half century. In recent years, there seems to be more room for divergence. Based on a number of cases which attracted much popular attention, Harris has concluded that precedents should be easier to depart from: ‘The presumption to date has been that stare decisis values should prevail over the overruling of “merely wrong” precedents. I think that the presumption should be reversed in favour of the overruling of wrong precedents unless their retention can be justified in the circumstances by overriding stare decisis values.’21 G. WESTERN

CIVIL LAW: TOWARDS STARE DECISIS?

Many writers in civil law countries seem to be of the opinion that the future lies with a system in which precedent plays a larger role, as compared with legislation. Adams for example argues for a culture of attributing binding effect to judicial precedents.22 In the general report for the Utrecht conference, I have defended the view that this is not a desirable development. It seems to me that the notion of the primacy of precedents of one single supreme court in a legal system no longer is adequate.23 There are two arguments which I gave. First, the central position of courts – and the rather humble role of legal writing – is changing. In most European countries, Constitutional courts or councils have gained importance. If one reads recent overviews of Belgian or German case-law, the importance of the decisions of Belgium’s Constitutional Court and Germany’s Bundesverfassungsgericht is overwhelming. But the Belgian Cour de cassation and the German Bundesgerichtshof will also have to follow the decisions of the two European courts: the Luxembourg Court of Justice – including its Court of first instance – of the European Union and the Court of Human Rights in Strasbourg.24 The Marckx case is an example of a European decision of the latter court which has exerted considerable influence in Belgium. 25

21

B. V. Harris, ‘Final appellate courts overruling their own ‘wrong’ precedents: the ongoing search for principle’ (2002) 118 Law Quarterly Review 408, 427. On p. 422 ff., the author develops a number of ‘considerations relevant to deciding whether to defer to or overrule precedent’. 22 Maurice Adams, ‘Stare decisis et quieta non movere/Rechtszekerheid en de gebondenheid aan precedenten in het Engelse rechtssysteem. Lessen voor ons?’ (2002) Tijdschrift voor Privaatrecht, 13, 50. 23 The same argument can be made with regard to public international law – see Nathan Miller, ‘An international jurisprudence? the operation of “precedent” across international tribunals’ (2002) 15 Leiden Journal of International Law 483, 498, who however argues that the various international tribunals have taken to referring to the case-law of one another. 24 This point has been elaborated by Meinhard Schröder in his PhD thesis submitted to the University of Munich on Gesetzesbindung des Richters und Rechtsweggarantie im Mehrebenensystem (Mohr: Tübingen, 2010), 347 p. 25 Marckx v Belgium, European Court of Human Rights 13 June 1979, No. 6833/74.

139

Precedent Revisited

Apart from this formal argument, there also is a substantive argument. The changes in norms and values, in theories as to finding the law, nowadays are so rapid, that case-law – which often takes a long time to be submitted to a legal system’s highest court – simply cannot cope with them. Legal writers will have to fill the gap. They should formulate general principles and fit in the various bodies of case-law developed by the traditional highest civil courts, the newly formed constitutional courts and the European courts. Attributing stare decisis to the decisions of one of these courts no longer fits into this system. This will mean that established precedents, which have not yet been challenged in the courts, may well lose their value, because they are at variance with the newly developed general principles, such as that of non-discrimination,26 equality27 or proportionality. So far, I have addressed purely national questions of precedent. But in Europe, there is a growing body of community law. It does not make sense that this law is being interpreted in different ways and by way of different systems of reasoning in the various jurisdictions. In this regard, civil law jurisdictions have much to learn from common law courts, where the use of precedents from other common law jurisdictions, if only as persuasive precedents, is by no means exceptional.28 Recently, English courts have even accepted continental cases as – persuasive – precedents.29 Civil law courts rarely take into account what is happening over the border. I would suggest that not only the courts are to blame. Legal academics have a task in preparing the use of comparative law. A most important part of this is the collection of cases, assembling these into a data-base and providing translations into easily accessible languages.30 H. CENTRAL

AND

EAST EUROPEAN

CIVIL LAW: TOWARDS MORE EMPHASIS

ON PRECEDENT

Has the recent adhesion of a large number of Central and Eastern Europe led to the maintenance of the current civil law and common law divide? I suggest that although this is something for the countries themselves to decide, for the ability of their legal system to react quickly to new developments it is essential that they follow their Western civil law counterparts. How to do so? The general report for the Utrecht conference comes up with a number of suggestions. First, legal doctrine is important in this matter. It is essential that in national books on legal methodology, the importance of case law is stressed. Second, cases should be reported far more often than is now done. Until recently, the number of law reviews or other publica26

See for instance Nicola Jägers, Corporate Human Rights Obligations: in Search of Accountability, Ph.D. thesis Utrecht, Antwerpen: Intersentia, 2002, 309 p. 27 Denis Berthiau, Le principe d’égalité et le droit civil des contrats (LGDJ : Paris, 1999), 475 p. 28 A famous example is the case of Donoghue v. Stevenson, where a New Jersey case was used. See Kirby in Precedent and the law (n. 1), p. 75. 29 See cases such as White v. Jones, Fairchild, and Greatorex v. Greatorex, [2000] 1 Weekly Law Reports 1970, quoted by Basil Markesinis, Comparative Law in the Courtroom and Classroom/The Story of the Last Thirty-Five Years (Hart: Oxford, 2003). 30 Some examples are the several collections of cases relating to the Vienna Sales Convention and the one on the EU directive on product liability, set out by Thomas Lundmark, “Soft” stare decisis: the common law doctrine retooled for Europe’ Richterrecht und Rechtsfortbildung in der Europäischen Rechtsgemeinschaft, ed. Reiner Schulze, Ulrike Seif (Mohr: Tübingen, 2003), 161, 168.

140

Ewoud Hondius

tions devoted to reporting of cases has been very limited. Also, the idea was often that reporting a case gives it more weight and therefore such cases must be carefully selected, preferably by justices in the Supreme Court. Since the process of selection takes a considerable time, cases are reported with considerable delay. I do not suggest that this selection process be abolished, but rather that it should be supplemented by law reports for specific areas which are based on voluntary submissions by judges and attorneys. The argument that this will be very expensive, in this era of electronic law reports is no longer valid. Third, in order for the reported cases to have any impact they should be annotated. This is the lesson to be learnt from Western Europe. It was the annotations of the two grandmasters of Dutch civil law, Meijers and Scholten, which revolutionised the place of case law in the Netherlands. In France, it is still virtually impossible for an outsider to fathom the relevance of a case handed down by the Cour de cassation without case notes. This is easier said than done. Courts may at first not like their awards being criticised and this may reflect upon the annotators. This in turn may deter possible annotators who do not wish to jeopardise their career. Fourth, legal education should take into consideration the importance of case-law as source. The impression one gets from legal education in Central and Eastern Europe, is that precedent is not taught at all. Exchange students, upon arrival in Western Europe, in their classroom papers do not report cases, unless by way of example. Fifth, if we wait for the present younger generation of law students to get into the key positions in society, it will take another 20–30 years before Central and Eastern European countries have caught up. We should therefore also provide continuing legal education to the present holders of key positions. It is not in all legal systems with a low degree of precedent that one can find a movement towards a greater respect for case-law. As Castelluci has observed, China has a system – which is gradually infiltrating Hong Kong and Macao – where a separation of powers is not practiced and where the courts are still considered the bouches de la loi.31 I. LEGITIMACY A central question as to precedent is that of legitimacy. Why should courts establish precedents for the future? Why should their successors adhere to these decisions? The topic comes up at most conferences of the International Academy of Comparative Law.32 There has existed a group of academics, called the Bielefelder Kreis, which seeks to advance understanding of fundamentals of law and its methodology through systematic comparative and theoretical analysis. This group published a well received volume on precedent.33 In the end, the principle of equality or of nondiscrimination will play a major role. There is no consensus in civil law countries as 31 32

Castelluci in Precedent and the law (n. 1), 349. See for instance Whitmore Gray, To what extent are judicial decisions and legal writings sources of law?, in: Rapports généraux au IXe congrès international de droit compare Téhéran 27 septembre – 4 octobre 1974 (Bruylant: Brussels, 1977), p. 31–53. 33 D. Neil MacCormick and Robert S. Summers (eds), Interpreting precedents – A comparative study (Ashgate: Aldershot, 1997).

Precedent Revisited

141

to the question of how the exception to legislative supremacy can be explained, however. J. ARBITRATION Arbitral awards do often show a reluctance to profess to establish and use precedents. Andrés Rigo Sureda argues in a recent essay that arbitrators should get rid of this reluctance.34 That is not the opinion of Charles Brower, Michael Ottolenghi and Peter Prows, who in their essay ‘The saga of CMS: res judicata, precedent, and the legitimacy of ICSID arbitration’ point to the recent opposition to ICSID arbitration on this count in Latin America, which has led to revocation of the ICSID Tearty or parts thereof by Bolivia, Ecuador and possibly Nicaragua and Venezuela.35 K. PRECEDENT

AND CORRUPTION

The general report for the Utrecht conference mentions a – possible – fifth group of jurisdictions: those where Western law has been exported beyond its host jurisdictions and where in the words of Patrick Glenn the problem of corruption has assumed massive proportions36. It would have had been interesting to report on them, but because no national reports from such jurisdictions had been submitted they had to be left out. Recently some literature has become available on corruption in the Indonesian judiciary. In an Australian book, Simon Butt discusses the ‘surat sakti ‘ (magic memos). These are letters issues by the Chief Justice of the Indonesian Supreme Court declaring to a lower court that a judicial decision cannot be enforced. Butt argues convincingly that these are often motivated by corruption (in Indonesian: ‘KKN: korupsi, kolusi and nepotisme’).37 It may be surmised that in countries where judicial corruption is endemic, the precedent value of judicial decisions is small. L. CONCLUSIONS What may we learn from the developments sketched in this paper? There seem to be conflicting tendencies in civil and common law. On the European continent, 34

Andrés Rigo Sureda, ‘Precedent in investment treaty arbitration’, in: International investment law for the 21st century/Essays in honour of Christoph Schreuer, ed. Christina Binder et al. (Oxford University Press: Oxford, 2009), 830–842. 35 Charles Brower, Michael Ottolenghi and Peter Prows, ‘The saga of CMS: res judicata, precedent, and the legitimacy of ICSID arbitration’, in: International investment law for the 21st century/Essays in honour of Christoph Schreuer, ed. Christina Binder et al. (Oxford University Press: Oxford, 2009),843–864. 36 H. Patrick Glenn, Legal traditions of the world/sustainable diversity in law (Oxford University Press: Oxford, 2000), 247. 37 Simon Butt, ‘Surat sakti: the decline of the authority of judicial decisions in Indonesia’, Indonesia: law and society, ed. Timothy Lindsey (Federation Press: Sydney, 2008), 346–362, including a detailed analysis of cases such as the PLN case and the Newmont case.

142

Ewoud Hondius

there was a system of precedents before the codification in the XIXth century. Because of the primacy which at the turn of the XVIIIth century was attributed to legislation, precedent lost its importance. But because of idleness of the legislature in the area of private law, the courts gradually regained some of their former positions. In this, they were helped by legal writing, which in its annotations provided case-law with terms of reference and allowed practitioners to see its importance. Legal theory comes to the conclusion that there now is once again a light version of the system of precedent, although with exceptions and ways out. Not all civil law systems are in the same state of development. Especially lagging behind are the systems of Central and Eastern Europe, which still suffer from their socialist past. In the common law systems, the development has been the reverse. Until the XIXth century, there was no system of precedent. Only recently has it been accepted that the courts should also follow ‘wrong’ precedents. In this sense there is a major difference with the civil law, where courts are not bound to follow wrong precedents.38 But the difference is diminishing, witness the recent call in the common law by B. V. Harris for reversal of the main rule.39 Most authors predict that the future lies in the common law, which will gradually extend its theory of precedent first to Western civil law nations and then to Central and Eastern Europe and other socialist or formerly socialist states.40 I do not share this view. I do agree that equality and predictability justify adherence to at least a limited theory of stare decisis.41 I also agree that legal theory should take into account that the law does not evolve within itself. However, I oppose the introduction in civil law jurisdictions to a strict system of precedent as may still be found in common law systems. Instead, a system based on persuasive authority of cases as contextualised by legal doctrine seems preferable. Support for this may be found in the national reports from most civil law countries for the Utrecht conference.42 As Malaurie has observed: ‘Le rôle du précédent est donc différent dans le sytème juridique français et la Common Law; ces deux systèmes se sont légèrement rapprochés, mais de manière superficielle. Cette différence est une des richesses de l’héritage culturel de l’Europe’.43

38

39 40 41 42 43

That continental courts, unlike their common law counterparts, do not consider themselves bound by ‘wrong cases’ is apparent from cases such as Hoge Raad 7 March 1980, Nederlandse Jurisprudentie 1980, 353, in which the Dutch Supreme Court accepted that its view had for a long time not been observed by lower courts. B. V. Harris, Final appellate courts overruling their own ‘wrong’precedents: the ongoing search for principle (n. 21), 427. See for instance Lookofsky in Precedent and the law (n. 1), p. 253, who argues that a better awareness of their role in creating jurisprudence, may make Danish court decisions more transparent. Lundmark, op. cit. (n. 10), at 166. In Precedent and the law (n. 1), Malaurie, Adams, Sourlas and Haazen concede that the differences between the two systems have been exaggerated. Malaurie, in Precedent and the law (n. 1), 139.

THE AUTHORS 1. Editors (and co-authors): Carlos Bernal Pulido Senior Lecturer, Macquarie University Macquarie Law School Building W3A Macquarie University NSW 2109 AUSTRALIA E-mail: [email protected] Thomas Bustamante Tenured Adjunct Professor, Federal University of Minas Gerais Faculdade de Direito da Universidade Federal de Minas Gerais Rua João Pinheiro, 100 Centro Belo Horizonte, MG 30.130–180 BRAZIL E-mail: [email protected] 2. Contributing Authors: Larry Alexander Warren Distinguished Professor of Law, University of San Diego University of San Diego School of Law 5998 Alcala Road San Diego, CA 92110 UNITED STATES OF AMERICA E-mail: [email protected] Pierluigi Chiassoni Professor, University of Genova Università degli studi di Genova Facoltà di Giurisprudenza Via Balbi, 30–17 Genova 16126 ITALY E-mail: [email protected]

144 Marina Gascón Abellán Full Professor, University of Castilla-La Mancha Universidad de Castilla-La Mancha. Campus de Albacete. Plaza de la Universidad, 1 02071 – Albacete SPAIN E-mail: [email protected] Ewoud Hondius Professor, University of Utrecht Janskerkhof 12 Kamer 3512BL Utrecht NETHERLANDS E-mail: [email protected] Victoria Iturralde Professor, University of the Basque Country Facultad de Derecho Campus Universitario de Gipuzkoa Manuel Lardizabal 2 20018 Donostia-San Sebastián SPAIN E-mail: [email protected] Patrícia Perrone Campos Mello Barrister at the Brazilian Bar Avenida Prefeito Mendes de Moraes, 990/203 São Conrado Rio de Janeiro, RJ 22.610-095 BRAZIL E-mail: [email protected] Zhang Qi Professor, Peking University Peking University Law School Law School Peking University 5 Yiheyuan Lu, Haidian District Beijing 100871 CHINA E-mail: [email protected]

The Authors

In this volume, the 3rd issue of the Proceedings of the 24th IVR World Congress, the reader will find a selection of papers on the general theme of ´The Philosophy of Precedentµ. The foundations, legal nature, structure, strength and uses of the case law were vividly debated at two special workshops dedicated to the study of legal precedent by a large number of scholars from around the world. The attention that the theme received at the congress indicates a trend of growing interest on the topic of precedent among legal philosophers, legal theorists

and practitioners from all of the legal families and traditions, including those of the so-called Civil Law Systems. Taking into consideration this movement towards the universality of precedent-based reasoning in legal discourses, the essays comprised in this volume provide an account of legal precedent which explains the connections between theoretical issues on the nature of precedent and practical queries over its foundations, structure, strength and uses in contemporary legal systems.

www.steiner-verlag.de Franz Steiner Verlag

ISBN 978-3-515-10150-9