U.S. and European courts alike often are confronted by cases litigating conflicting fundamental rights. To resolve those conflicts, U.S. courts often resort to balancing competing interests, whereas European courts tend to invoke the general principle of proportionality. ¹ The approaches taken are similar, but not the same; sometimes the general principle of proportionality is referred to, I think wrongly, as "balancing." The result? Confusion.

Happily, the sources of that confusion can be fairly found and resolved. ² A parallel problem lurking in the cases, however, is a similar confusion as to the hierarchy of norms. What are fundamental rights, fundamental freedoms, constitutional rights, civil rights, human rights, and general principles of law? How are they hierarchized?

Contemporary legal thought recognizes a plethora of terms used to refer to varieties of deductive general principles - or rules, which are hierarchically superior to ordinary legislation such as: fundamental rights and freedoms, human rights, basic rights, general principles of law, civil rights, constitutional rights etc. ³ There, too, we see confusion due to a lack of basic*88 categorical definition. ⁴ Diligent sourcing and rational hierarchization, however, can resolve the confusion and ambiguity of rights discourse, enabling further transatlantic rapprochement of legal reasoning. The rise of fundamental rights, whether called basic freedoms, human rights, constitutional rights, civil rights, or general principles of law, is the constitutionalization of private law. ⁵ The fact that these general deductive rules merge into and map onto each other is one example of what I call norm convergence. ⁶ Such convergence is one element of the construction of a global ius commune featuring deductive general principles of law from the civilianist legal system alongside binding case law from the Anglo-Saxon common law system. This global ius commune justifies itself by seeking to attain fulfillment of fundamental rights, expressed as general principles of law and adjudicated via the general principle of proportionality. ⁷

This paper argues that fundamental rights, human rights, constitutional rights, and general principles of law can be properly assimilated into each other because these norm complexes are used deductively as hierarchically superior "meta-rules" for the elaboration of other rules subordinated to them. ⁸ This paper further argues that the remaining general deductive norm complexes such as civil rights and basic rights or freedoms (which to my view are hierarchically subordinate) can, likewise, be categorized and hierarchized by way of concepts such as substance versus procedure, natural versus positive, alienable versus inalienable, and by relating these hierarchically subordinate general categories to each other and to the abovementioned general deductive norm complexes.

One tendency in contemporary legal thought is the clash of general *89 deductive principles, or fundamental rights, in cases heard before courts--the constitutionalization of what were once mere political rights, and their juridification, i.e. judicial adjudication through national constitutional courts and transnational courts and court-like bodies. ⁹ A second tendency in contemporary legal thought is the relativization of constitutional rights against each other. Rather than seeing fundamental rights as absolute and inviolable (the approach in early modernity), fundamental rights today are generally regarded as contextualized against each other. ¹⁰ The courts, noting that each of the clashing rights is somehow fundamental, try to resolve the conflict between fundamental rules or principles either by "balancing" ¹¹ the competing interests against each other (multi-factor interest analysis) or by engaging in judicial review to check whether the act in question is a proportional means (normally, a rational means; exceptionally, a necessary means) to a permissible end (means-end rational review with strict scrutiny for suspect classes). ¹² The tests and their application are complex. The doctrinal nuances and distinctions of such basic normative conflicts between competing deductive general principles are best extracted by considering the methods used in the United States ("U.S.") and the European Union ("EU") separately in order to note the commonalities and then synthesize common practices. Accordingly, this article illustrates the clash of competing relativized fundamental rights and the resolution of that clash using the examples of the cases of Viking ¹³ and Laval. ¹⁴

Viking and Laval are examples of conflicts between deductive fundamental*90 general principles. We will use them to extract the general principle of proportionality used in the EU court system in order to compare that legal method to what U.S. jurists somewhat inaccurately refers to as "balancing", a polysemic term. One aspect of "balancing" is more accurately referred to in U.S. parlance as: means-end rational review with strict scrutiny for suspect classes. Both the EU general principle of proportionality and the different varieties of the U.S. interest "balancing" face the same problems: the legality of state action, which infringes on fundamental rights and the evaluation of competing interests. ¹⁵ By examining the cases, we can extract, compare, and critique the relevant rules.

In Viking, the E.U. faced the conflict of the right to strike by a labor union with the right of enterprises to establish themselves anywhere in the Union. ¹⁶ That case, as is typical of so many contemporary cases, represented the clash between two general fundamental rights: the right to strike and the freedom of establishment. ¹⁷ Viking was resolved first by assimilating the competing rights into each other ¹⁸ then by relativizing those rights against each other, ¹⁹ and finally by regulating their relativized interplay via the general principle of proportionality. ²⁰

Here, as a matter of method?not substance, as I agree with the result--I critique the Viking decision. Viking did not ask the question of whether the freedom of establishment was invaded by the right to strike or, instead, whether the right to strike was invaded by the freedom of establishment. The court's decision is also open to critique, as the European Court of Justice ("ECJ") could have plausibly decided the case either way, both as a matter of practical fact and as a matter of law.

Viking and Laval are typical example of what Ronald Dworkin refers to as "hard cases" ²¹--cases which are critically important, which the court could decide either way, and which are, in Connolly's terms, "essentially contested" i.e. polysemious and conflicted. ²² Because the ECJ could have justifiably ruled either way, it faced two problems: the problem of legal reasoning (what is "right reasoning? ²³ i.e., "how to") and the problem of *91 justification (why is the reasoning right? i.e., "why to"). These two problems are different from each other, but they are also related to each other. I use Viking to draw out the problems of legal reasoning and of justification in order to try to help the courts find solid ground--a workable taxonomy of fundamental rights--on which to adjudicate fundamental normative conflicts to improve the efforts at transatlantic comparative law, specifically with regard to what I call "norm convergence" ²⁴ and with respect to "rights discourse." ²⁵

U.S. constitutional review distinguishes judicial review of conflicting fundamental rights subject to the Equal Protection and Due Process clauses from judicial review of conflicts of fundamental rights in cases involving the Fourth Amendment ²⁶ and the Takings Clause. ²⁷ The Fourth Amendment*92 to the U.S. Constitution prohibits unreasonable, e.g. warrantless, government searches and seizures of persons and things. ²⁸ The Takings Clause, for its part, states that private property shall not be taken by the state for public use without just compensation. ²⁹ In the context of the Fourth Amendment ³⁰ and in the context of the Fifth Amendment's Takings Clause, ³¹ and, at times, in the context of procedural due process, the U.S. Supreme Court balances the competing interests of the plaintiff and defendant. The various interests of each party are taken into account and expressed as factors and the various factors are assigned a weight, ³² generally implicitly. ³³ The factors which favor the plaintiff are then compared against *93 the factors favoring the defendant. ³⁴ Where the factors are of equal weight, the decision goes against the party bearing the burden of proof ³⁵?ordinarily, the plaintiff, because of the general principle that the moving party ordinarily bears the burden of proving their position. ³⁶ Some factors, if satisfied, may be dispositive. I refer to this as "multi-factor interest balancing" . Multi-factor interest balancing occurs at times in the due process context as well, ³⁷ though it seems more clearly visible to me in the Fourth Amendment ³⁸ and the Takings Clause context.

Multi-factor interest balancing tests, ³⁹ however, can be easily manipulated. Which party should be considered as having relevant interests? ⁴⁰ Which interests of the parties should be considered in the balance? Finally, *94 what weight should be applied to the various interests to be balanced? The competing interests are generally not specified by legislation; the weights to be applied certainly are not. Multi-factor balancing tests allow judges to manipulate legal outcomes by a) the selection of persons with interests to be considered; b) the selection of interests of relevant parties to be considered; and c) the determination of the weight to be attached to the relevant interests of relevant persons. ⁴¹

The ability of judges to select which persons and interests are relevant opens balancing tests to criticism as being unpredictable, manipulable, subjective, and unjust. ⁴² These criticisms can only be answered by introducing economic and policy arguments into legal balancing tests. If some--or, better yet, all--of the relevant interests are valued monetarily, or if the political will of the legislator can be invoked, ⁴³ this critique is reduced, or disappears. The problem of attributing objective weights to competing interests partly explains the rise of law and economics as a solution to the problem of judicial voluntarism. Multi-factor interest balancing tests are most visible and least manipulable in economic contexts such as takings because, with respect to takings, the court is directly concerned with market value, i.e. quantifiable economic interests. ⁴⁴

Ad hoc multi-factor interest balancing is methodologically weak (Whose interests? Which factors? What weight?). ⁴⁵ A related method for resolving conflicting fundamental rights, sometimes also called balancing, arose in the due process and equal protection contexts: rational basis means-end review with strict scrutiny for suspect classes. ⁴⁶ This method is rarely termed "proportionality" in U.S. legal discourse, but it will be immediately obvious to a European lawyer that we are looking at that method: the law must be proportional; ⁴⁷ that is, it must be a rational and necessary *95 means to a permissible end, which burdens fundamental human rights as little as possible. ⁴⁸

Means-end review in U.S. law arises most evidently in the context of equal protection and due process. ⁴⁹ The typical case is an action by the government, which infringes on a constitutional right. ⁵⁰ The action is then challenged by a private person as unconstitutional. ⁵¹ If the law is not a rational means, or if the end sought by the law is impermissible (e.g., the end is ultra vires; or, likewise, the end extinguishes an inalienable fundamental constitutional right), then the law will be quashed as an unconstitutional exercise of state power. ⁵² Typical cases arise out of violations of the constitutional guarantee of due process or equal protection. In simplest terms, the law must be a reasonable means to a permissible end. ⁵³ If the law burdens a "suspect class" ⁵⁴ (racial minorities for example), or infringes a fundamental constitutional right, ⁵⁵ then the law must serve a compelling state interest and the means chosen must be the least restrictive of the fundamental right so infringed. The most coherent expressions of means-end review also require that, in cases of strict scrutiny, the law must not be a merely rational means, but also a necessary means to a permissible end. ⁵⁶ All of this, excepting*96 the concept of "suspect class", is strikingly familiar to E.U. lawyers. I have elsewhere addressed the origins and rise of the general principle of proportionality as a globalized and globalizing trait of contemporary legal thought. ⁵⁷ I regard that principle as an apotheosis of western legal thought: the rule of law. ⁵⁸

U.S. constitutional doctrine seems conflicted and confused in the use of multi-factor interest balancing and in the use of means-end review with strict scrutiny for suspect classes. I say that because in the due process context sometimes we see multi-factor interest balancing, ⁵⁹ but we also see means-end review. ⁶⁰ Meanwhile, there is the lurking issue of whether the Due Process Clause speaks only to procedural rights, or also has a substantive content: that issue seems currently settled but has not always been so - and times change. ⁶¹ There is also the issue of whether the Due Process *97 Clause of the U.S. Federal Constitution may be invoked by individuals against the U.S.federated states, which parallels the issue of direct effect in the E.U. Finally, there is also the issue whether the Constitution (or EU Treaty) may be invoked as creating rights and duties among private actors inter se which in the U.S. is addressed under the rubrics "state action" and "color of law" whereas in Europe it is addressed as the problem of "third party affect" (also known as "Drittwirkung"). How are these issues, which clearly parallel corresponding U.S. doctrines and issues, treated in E.U. law?

1. Proportionality

a. Elements of the Test

The general principle of proportionality requires that laws be rational and necessary means to a permissible end. ⁶² The case of Laval, which adjudicated the right of establishment against the right to strike, is a typical example of both the clash of fundamental rights and the use of proportionality to resolve that clash. ⁶³

In Laval, the court ruled that a restriction on fundamental freedoms such as the right of establishment "is warranted only if [1 it pursues a legitimate objective compatible with the Treaty [legitimate ends] and is justified by [2 overriding reasons of public interest; if that is the case, [3 it must be suitable [i.e. reasonable] for securing the attainment of the objective which it pursues and [4 not go beyond what is necessary in order to attain it" [necessity]. ⁶⁴ This test - proportionality - looks very much like the means-ends rational review test with strict scrutiny for suspect classes seen in U.S. law. The two tests likely share a common theoretical origin. ⁶⁵

The difficulty, however, is that we are comparing two competing fundamental rights. Thus, we could easily characterize either right as being invaded by the exercise of the other. ⁶⁶ Moreover, whichever right we characterize*98 as invaded by the exercise of the other right will have, at least as a practical matter, an inevitable procedural advantage over the invading right, due to the burden of proof as well as the presumption of the legality of legislation. ⁶⁷ For both of those reasons, the doubtful cases?and hard cases are always doubtful?will, as a matter of logic and procedure, tend to be resolved by default in favor of the invaded right because of the allocation of the burden of proof. ⁶⁸ If, and only if, we can distinguish fundamental rights from each other can we hope to prioritize whether the right to strike is hierarchically superior to the freedom of establishment ⁶⁹ in horizontal effects cases - and then structure the schema of general rule and specific exception correctly, allocating the burden of proof and any legal presumptions accordingly.

b. Rights or Freedoms?

The Viking case presented a hopeful start when it characterized the right to strike (more exactly the right to take collective industrial action) as a right, ⁷⁰ and distinguished that right from the freedom of establishment. ⁷¹ The careful reader might have noted the different terms and hoped to glean therefrom some definition of "rights" as opposed to "freedoms," which would enable us to correctly characterize the hierarchical relationship between fundamental freedoms--here, the freedom of establishment-- and fundamental rights--here, the right to strike. Alas, this hope for categorical clarity to establish parameters for presumptions and possibly also the burden of proof was dashed at the last minute. The court, in keeping with the contemporary trend, assimilated freedoms into rights and rights into general principles. ⁷² The court then relativized fundamental rights and freedoms *99 against each other, in keeping with trends in contemporary legal thought; these relativized rights were then regulated by the general principle of proportionality. ⁷³ The assimilation of fundamental freedoms and rights into each other -making "right" synonymous with "freedom"-- at first glance appears sensible because "rights" and "freedoms" both are deductive general principles. Equating fundamental rights and freedoms, however, destroys the possible categories of "right" as opposed to "freedom." A firm categorical distinction would allow us to differentiate "rights" from "freedoms" and would allow us to determine whether rights or freedoms are hierarchically superior, which in turn would have allowed us to allocate presumptions of legality and burdens of proof, resulting in a clearer determination of the case and a more plausible justification for the result. A freedom is a claim that one is not subject to state sanction, to state interference, to state power; the absence of constraint. I regard "freedom" as one particular form of right. ⁷⁴

c. Whose Right Was Invaded? The Coase Theorem.

Recall that we are trying to determine whether, by defining the nature of the legal interest as a right or a freedom, we could establish a hierarchization of rights with respect to freedoms such that we would then be able to say whether the right invaded the freedom or the freedom invaded the right. The ECJ in Viking did not take that tack. How else might we solve the problem? We ought not to try to solve the problem by looking to see which party did or did not exercise the right or freedom because, as Ronald Coase pointed out, in any conflict of rights, the court can cast either party's right as invading or invaded. ⁷⁵ The implication from Coase is that looking to "first in time" or "actor versus acted upon" is not a sensible path to try to hierarchize rights and freedoms. ⁷⁶ Coase essentially recast the principle that any omission may be re-characterized as an act, or vice-versa, through the idea that rights and duties are correlatives to reach his idea of reversibility of injury, in order to permit courts to retreat from enthymemes and presumptions*100 so that the court can consciously and dispassionately analyze what outcome is socially desirable. Coase's theorem implies that results-oriented, outcome determinative, purposive reasoning is legitimate. If Coase is right, then "first in time" is no more of a solution than "act" or "omission." ⁷⁷ That is, we cannot definitively determine that action is legitimate or not based either on who did or did not act, or based on whose right was invaded. Are there any other ways to try to hierarchize the right to strike versus the freedom of establishment?

One step the court in Viking takes is to compare the source of the competing rules. The right to strike is found in national, E.U., and international law. ⁷⁸ In contrast, the freedom of establishment is not found internationally (caveat lector: the WTO is working on it). Since E.U. law enjoys supremacy with respect to the law of its Member States, the court could readily dismiss the characterization of the right to strike by the Member State as largely irrelevant, except as an interpretive guide to the content of the same right at the E.U. level. Likewise, the court was free to essentially ignore the right as it is defined internationally, since international treaties are presumed not to grant rights or duties to physical persons, or even legal persons, and to grant rights and duties to states only, unless a different intent is proven. ⁷⁹ That is just what the court did in order to downplay the primacy of the right to strike with regard to the right of establishment. ⁸⁰ Thus, the court did not distinguish the right to strike?a norm found in the law of Member States, E.U. law, and in international law?from the freedom of establishment?a norm found only in the E.U. law. The court in Viking did not make or rely on those distinctions in its determination that the right to strike in that case disproportionally ⁸¹ infringed on the right of establishment. The court did not attribute hierarchical priority to the right to strike even though that right existed at the national and global levels. ⁸² In fact, the ECJ found that the competing freedom of establishment displaced the right to strike, ⁸³ which appears undemocratic. That in turn leads to the question whether and how the decision of the court is justifiable.

*101 d. Relativizing Rights

The puzzle we are facing--whether a given right can be categorized as somehow hierarchically superior to another right--is an intriguing one. It is also pervasive, yet often latent, in contemporary legal thought due to the constitutionalization ⁸⁴ of fundamental rights as the proper goal and legitimizer of state power ⁸⁵ in the post-Westphalian international system. ⁸⁶ While I have proposed a method for categorizing and then hierarchizing general deductive principles, i.e. fundamental constitutional rights, ⁸⁷ I leave that issue for the end of this paper in order to encourage the reader to think abstractly and creatively about how she or he might resolve the conflict of norms. Further, I wish to expose other ideas first, which I think will help to resolve the question of whether and how fundamental rights can be hierarchized as general principles of law and then adjudicated by the general principle of proportionality.

1. Direct Effect

Viking is interesting in the treatment given to the question of the direct effect of the EU Treaty, ⁸⁸ as well as to the horizontal or vertical nature of that effect, whether direct, or indirect. The court was confronted ⁸⁹ with these possibilities: (1) find either a vertical effect--between the E.U. and the Member State--or a horizontal effect ⁹⁰--(E.U. rights being affected to private parties inter-se); (2) find either a direct and binding) effect, or a merely indirect and persuasive effect of the treaty; or (3) find no effect at all. ⁹¹ The court could have generated a plausible, coherently structured argument to support reaching any of the five outcomes--horizontal direct, horizontal indirect, vertical direct, vertical indirect, hortatory norms, or no effect whatsoever, even as a persuasive evidence of the law. The court found that the treaty had a (1) direct effect and the direct effect was (2) horizontal, i.e. between private parties inter se. ⁹² This outcome was probably the most plausible, but we shall see it is not unproblematic.

*102 a. Third Party Beneficiaries

Ordinarily, treaties and constitutions only address public law persons; exceptionally, they may grant rights and duties to physical persons. ⁹³ When constitutive instruments grant rights and duties to physical persons, the issue arises as to whether those rights are directly enforceable by the persons so affected, or instead only indirectly enforceable by their State (in international treaty law, this is the question whether the treaty is self executing). ⁹⁴ Third party beneficiary doctrine from contract law has been used to analyze the problem of basic constitutional rights under the rubric of Drittwirkung (in English, "third party effect"). "Self execution," "state action," "color of law," "third party effect," and "direct effect" all refer to the same issue: the application of public law to private persons and especially to the relations of private persons inter se. ⁹⁵

In order to understand third party effect (Drittwirkung), ⁹⁶ we should remember that the direct effect doctrine is essentially the application of the contract idea of third-party beneficiaries to constitutional and treaty interpretation. ⁹⁷ It is one expression of the logic of an "economic constitution," that the State establishes framework economic rules or basic rights and that those basic rights in turn constitute the State. Consequently, constitutions and treaties are like contracts, and analogies to contract law can be made in the interpretation of treaties and constitutions. The (ordo-liberal) economic constitution consists of the idea of "markets without states," ⁹⁸ one form of a limited State. ⁹⁹ Though the economic constitution is positive, not natural, *103 as different states have different economic constitutions, the rights created thereunder are basic-- i.e., fundamental. ¹⁰⁰ This is the logical source of the distinction between natural, fundamental, constitutional rights and positive, alienable, fundamental economic freedoms.

Abstractly speaking, in a conflict between basic rights and fundamental inalienable rights, the inalienable fundamental right should prevail, leaving the question of whether the right to strike is inalienable--natural, inherent to the person--or found in all times and places, which it is not. Categorizing fundamental rights and freedoms may help us find their hierarchical relationships, which in turn may help us to resolve conflicts between them. In Viking, the court did not attempt to categorize and establish a hierarchy of competing and conflicting fundamental rights and freedoms. ¹⁰¹ Consequently, my comments here on that point are purely speculative.

b. Is the Party a State Actor?

To determine the effect, if any, of a constitutional or treaty provision, and whether that effect is direct or indirect, horizontal or vertical, we must first determine whether the person to whom a right or duty applies is a state actor. ¹⁰² In the case of Viking, this should not have been so conclusively answered in the negative. In the corporatist model of governance, which is consistent with an ordo-liberal vision of the economic constitution, ¹⁰³ labor unions act as quasi-state organs, establishing labor conditions such as minimum wages, maximum hours, vacation and pension rules, and workplace safety rules. ¹⁰⁴ They are quasi-state actors. The Viking court, therefore, could have said that the labor union was an instance of de facto state power; that the private law person, acting under the color of state authority (i.e. exercising sovereign functions --acto jure imperii) is in fact assimilated into the State for the application of the Treaty of Lisbon--though it did not. Such a finding would have meant that the EU Treaty had a direct effect and that the right of establishment applied, requiring demarcation with respect to the right to strike.

*104 c. If the Labor Union is Not a State Actor, Does the Treaty or Constitution Apply?

The ECJ did not characterize the labor union as a quasi-state actor. ¹⁰⁵ Nonetheless, it reached the same substantive outcome as it would have reached had it decided to treat the labor union as a quasi-state actor. The court found the treaty directly applicable both as to the right to strike ¹⁰⁶ and as to the freedom of establishment. ¹⁰⁷ Both the labor union and the enterprise were directly addressed by the treaty and bound in their interactions with each other to respect whatever rights were created or exposed by the treaty. ¹⁰⁸ By finding this to be a case of direct horizontal effect--of public law rights directly applicable by private law persons in their interactions--the court retained flexibility, but lost some credibility. The court can thus in later cases re-characterize the rights and duties of the labor union because the direct effect of one right does not imply a direct effect of others. Had the court held the labor union to be an instance of state power, all the actions of the union would be subject to rights and duties under the treaty.

In Viking, the court characterized the labor union as a private law person, but nonetheless directly addressed by the Treaty, ultimately leading us back to Van Gend & Loos, which held that private persons under the Treaty are addressees of basic rights and duties. ¹⁰⁹ Because the ECJ in Viking determined the treaty had direct horizontal effect, ¹¹⁰ the question whether the third-party effect could instead have been indirect became logically irrelevant and thus was not addressed. The court did not address the issue of possible indirect effects, which was prudently parsimonious. If the court had instead decided that there was no direct horizontal effect of private parties inter se, it would then have had to decide whether there was at least indirect effect or no effect at all. The decisions of the court on some questions foreclose the consideration of others. Parsimony is prudent adjudication, but is not a good heuristic for legal scholarship. Hence this article tries to go beyond Viking to see its implications.

*105 e. The Decline of the Public-Private Distinction

Viking presented the court with the choices of horizontal or vertical direct effect. ¹¹¹ Both of those options--the imputation of rights to private law persons by treaty (horizontal direct effect) and the treatment of the labor union as a quasi-state body (vertical direct effect)--are examples of what Professor Duncan Kennedy has long referred to as the "decline of the public-private distinction." ¹¹² That distinction is less important today than it has been in the past, and it has declined because it is seen as less and less tenable, leading to formally justified yet substantively unjust outcomes. ¹¹³ In E.U. law, the doctrines by which the distinction declines are the concept of "direct effect," "third party effect," ¹¹⁴ and "proportionality." ¹¹⁵ The public law/private law dichotomy declined even faster in the U.S. than in the EU, because branches of the common law, unlike civilian law, ¹¹⁶ rarely, if ever, have independent, i.e. autonomous, principles of interpretation.

All this leaves us facing omnipresent questions that the court did not address: (1) Did the right to strike invade the freedom of establishment or, instead, did the freedom of establishment invade the right to strike? (2) Would a hierarchy of fundamental rights and freedoms help to resolve their conflicts and, if so, how should they be placed in a hierarchy?

These questions are ultimately at issue in Viking and that the court tried to resolve the conflict at bar rather than develop and deploy a complex structure of rights and freedoms to determine whether the right invaded, or was invaded by, the freedom. This was prudent and parsimonious, but also somewhat dissatisfying, as much of the court's reasoning appears stunted and conclusory as a consequence.

What I have tried to hint at up to this point is the problematic nature of the Viking case and almost all such cases dealing with conflicts of norms. Conflicting fundamental norms are an essential and inescapable feature*106 of contemporary legal thought. Contemporary legal thought attributed to a plethora of rights to natural and legal persons--in response to the failure of the state-centered approach to properly protect human rights as shown so horrifically during the Second World War. After human rights were attributed to persons under national constitutions and international treaties, those private law persons were then empowered to directly enforce their public law rights and freedoms against each other and states. Permitting individuals to enforce their fundamental legal rights ¹¹⁷ (e.g., to life and moral integrity ¹¹⁸) against states was a reaction to the systematic abuse of human rights during the Second World War. The rise of fundamental rights discourse in the post-War era is one of the key features of the post-Westphalian system. ¹¹⁹ Human rights today are the goal and legitimizers of all state power as evidenced by the existence of non-derogable jus cogens ¹²⁰ fundamental rights (e.g., the right not to be tortured, enslaved, or exterminated).

The problem with this complex cacophony of conflicting claims--or is it an orchestra?--at which I have hinted and now clearly and succinctly state, is that the Judges could have constructed a plausible judicial edifice to reach any desired result. Judicial power in hard cases like Viking is arbitrary--not merely in the sense of arbitral or decisional, but also in the sense of capricious. Courts can almost always construct plausible inferential chains to reach any result they desire: outcomes can be manipulated. This arbitrary power, however, is a problem, for the arbitral and possibly arbitrary judicial power must be justified.

The manipulability of judicial outcomes is disconcerting because if any outcome could be plausibly reached, then no outcome is entirely foreseeable. Predictable outcomes are desirable for legal certainty and the procedural rule of law. Predictable outcomes are needed in order to structure expectations and transactions. If we were all governed by the unpredictability of the Mad Hatter, then we would be unable to undertake socially desirable economic activity.

Unpredictability is one problem which results from manipulability; manipulability also renders justification more important and more difficult. If judges can reason in formally valid ways to any outcome they desire, how is their exercise of what amounts to pure judicial power justifiable? How can one justify undemocratic and unpredictable judicial fiat? If any outcome could be plausibly reached, a justification will be needed for every outcome, yet such justification will be impossible, since the court could have reached any decision.

The problem then is not merely finding the correct legal reasoning; it is also, and more problematically, a problem of justification. So far, we have not found a resolution in the character or relationship of fundamental rights and freedoms. Perhaps thinking about the judiciary as an institution may help us to understand the problem better and even start to give us some hints as to the solution of the problem of justification.

Recall that the structural difference between the judiciary and the legislature is this: judges look at transactions with hindsight (ex post facto perspective); legislators look at transactions with foresight (ex ante perspective). ¹²¹ The judge makes decisions ex post facto. The legislator makes predictions ex ante. These are not the only salient institutional differences between judges and legislators, but the different temporal perspectives judges and legislators bring to the governed transaction is the most important institutional difference. The legislator provides general rules for application of a law. The judiciary provides specific answers to concrete cases where all facts have already transpired.

There are other institutional differences between the legislature and the judiciary. Judges are highly intelligent and technical experts in the law, especially in adjudication. Moreover, judges are not economically desperate-- they are just the opposite. Judges are neither struggling merely to survive at life, living hand-to-mouth the way the majority of people do, nor are they scrambling for fame, fortune, or public political power. Furthermore, judges are democratically insulated. Of course, the fact of democratic isolation contributes to the problem of judicial legitimacy. Paradoxically, it is the fact of political insulation that enables the assertion of judicial power in the first instance. Expertise and political isolation arguably give judges a better perspective and enable them to make socially beneficial, albeit unpopular decisions (e.g., ending racial segregation). Judicial power is really only*108 legitimated after the fact by the coherence of judicial argumentation, however (i.e., the act of justification and the results of judgments). If the judge gets it right, everyone is happy, even though the judicial decision would have been rejected by the people or a majority of voters. If the judge gets it wrong, there will be hell to pay: at the extreme, riots, assassinations, terrorism and repression. Moreover, the judicial activism might backfire: the politically unpopular minority might become more ruthlessly oppressed or exploited than ever due to backlash. However, despite that caveat, legitimacy can arise ex post facto through the effects of the judgment and also through the logical persuasive power of the accepted arguments. Well justified laws attract compliance: the governed are persuaded. Law is not merely a man with a gun, nor is the state but a mafia writ large.

To see the issue of legitimacy ex post facto clearly, consider the example of racial segregation. A majority of voters would not have voted to end racial segregation in 1960. U.S. courts ended segregation, not the voters or even their legislators. The court's decisions were extremely unpopular. The results of those unpopular decisions increased productivity and reduced violence. Thus, the success of the outcome legitimized the unpopular method after the fact. When the court misjudges the resulting social influences of their decisions as beneficial, however, conflicts result and judicial legitimacy is lost. This is why gay people are still second class citizens in the U.S. Institutional actors simply do not wish to risk standing up for that minority because they cannot imagine the benefits which will result from gay adoption, gay marriage, and complete integration of a controversial minority into society.

In Viking and Laval, the ECJ took the unpopular yet socially desirable decision to effectively open Western labor markets to posted workers and outsourcing into Eastern Europe. ¹²² This was the obvious decision, due to a shortage of youth in Western Europe and a capital deficit in Eastern Europe: demographics and capital structure explain why the outcome was desired as well as the desire to build a single market for goods, labor, capital, and services. ¹²³ The logic of the union and the law of supply and demand explain why the ECJ correctly took the politically unpopular but societally beneficial decision. Courts are loath to admit that they engage in such reasoning, yet it is a common occurrence. ¹²⁴ That then is our next topic: legal reasoning, and its relation to the problem of justification.

I use the term "effects-based reasoning" to refer to one variety of what Duncan Kennedy calls "purposive reasoning" or goal-oriented thinking. ¹²⁵ I am not referring to the direct effect of public law on private persons. I am referring by "effects-based reasoning" to a pragmatic, results-oriented mode of thought.

In effects-based reasoning, the court looks to the practical, real world effects of transactions and decisions. The court looks to the outcome that would result from its decision. For example, the ECJ characterized garbage as "goods" enjoying the right of free movement in Walloon Waste; ¹²⁶ in an even more extreme vein, the U.S. Supreme Court in Wickard v. Fillburn ¹²⁷ characterized feed, grown on a farm for consumption on that farm, as "interstate commerce" subject to federal regulation. ¹²⁸ The courts did not look to the common meaning of the term but instead imposed outlandish meanings to obtain a certain result. ¹²⁹

Courts are reluctant to admit they engage in effects-based reasoning. There is nothing per se illegitimate about effects-based reasoning. The problem of justification, however, already crucial in a close case, becomes even greater in a case involving conflicting fundamental rights and freedoms. Results-oriented thinking amplifies the problem of justification. When courts reason from an outcome to determine the legal conditions which must be satisfied for that outcome to obtain in order to determine which legal conditions in fact are fulfilled, the process of reasoning itself provides no justification. This lack of inherent justification, the absent sense that the objective preexisting law compelled the court's decision, explains why courts are hesitant to admit to thinking strategically about the premises which would impel and the results which would follow a given legal decision. Results-oriented thinking appears to subvert the judiciary's role as expositor after the fact of legislative will. That appearance is only illusory, however. Reasoning from possible outcomes to known facts (backward chaining of inferences) is legitimate because: (1) backward *110 chaining has heueristic value; (2) backward chaining generates formally valid chains of inferences; and (3) the question of justification is not equivalent to the question of valid forms of legal reasoning.

Backward chaining of inferences is both formally valid and an excellent heuristic. There is nothing inherently invalid about result oriented thinking. I argue that the questions of the formal validity of the court's reasoning, the material validity of the court's decision, and justification are separable. Those questions are related to each other yet are distinct.

Can results-oriented thinking be substantively justified? If so, how? We can best understand effects-based reasoning by understanding a form of inference referred to in artificial intelligence literature as "backward chaining" of inferences. By understanding the exact process of inferencing we can then separate the question of formal inferential validity from substantive justification. Once we see the formal validity and heuristic utility of backward chaining to obtain and compare competing results-oriented inferential chains we can then turn more confidently to the harder problem of justification.

In this section I wish to expose some procedurally valid modes of inferencing. I do not wish to argue that these inferential modalities are how courts do in fact reason, nor do I argue that courts ought to employ the inferential methods exposed in this section. All I wish is to show how formally logical inferential chains can be generated and point out that courts could and maybe do or possibly should use these modes of inferencing.

"Backward chaining" is an inferential method by which formally valid chains of inferences are derived from a given conclusion reasoning backward through whatever inferential steps would support reaching that conclusion. It is reasoning from the hypothesized outcome back to the set of known given premises. ¹³⁰ That is, backward chaining is reasoning from a given goal back through the possible causal inferences which would reach that goal to the known facts. ¹³¹ Litigators often "backward chain" because they want the court to reach the decisions that favor their clients.

"Forward chaining" is the opposite of backward chaining. It is the more usual method employed by judges. ¹³² In forward chaining, one proceeds from known facts to the inferences which they logically imply, pushing the chain of inferences toward whatever conclusion(s) the known *111 premises entail. ¹³³

Most courts regard backward chaining as illegitimate when undertaken by courts. ¹³⁴ This distrust of backward chaining is probably as a result of courts' self-perceived position as neutral, uninterested adjudicators merely applying the legislature's prognostications. Lawyers, however, backward chain frequently. ¹³⁵ It is considered good rhetoric, but supposedly bad adjudication. ¹³⁶

The problem with backward chaining is not formal validity. The problem with backward chaining is a problem of justification. Any creative thinker can backward chain from a desired outcome to the known facts in a procedurally valid way. The possibility to reach a variety of formally justified outcomes exists due to the plethora of valid inferential modes: deduction, induction, reduction, analogy, probabilistic reasoning and because of the variety of sources of law: custom, statute, constitution, treaties, and, internationally, general principles of law and scholarly works (la doctrine; Rechtslehre).

All formal demonstration is ultimately tautological. ¹³⁷ Thus, backward chaining is not useful as a method to justify decisions. While it might reveal a formally valid chain of inferences, that formally valid chain may well be materially invalid. On the other hand, suppose the court backward chained from every possible outcome and then picked the most plausible inferential chain? Suddenly backwards chaining no longer looks capricious, but appears formally and substantively valid. The closest we see to this form of conscious comparative backward chaining is in those decisions of the ECJ which are written as response. In a responsa style decision, in the manner of Aquinas, ¹³⁸ the ECJ raises the plaintiffs' arguments, the defendants' arguments and then its conclusion, issue by issue. ¹³⁹ Even there, the *112 court is forward chaining, but doing so by considering each argued inference presented by the parties sequentially.

Although backward chaining is rarely admitted to and is not useful for solving the problem of justification, it is extremely useful as a heuristic. A heuristic is a method used to discover knowledge (i.e. eu-reka! found it!). ¹⁴⁰ Essentially, we can use backward chaining to reveal enthymemes in our presumptions as well as possibly otherwise invisible legal issues.

We can see that backward chaining is in fact a formally valid method of inference when we recall the method of logical proof known as reductio ad absurdam. In a reductio argument, also known as proof by contradiction, we grant all premises of a given argument involving mutually exclusive alternatives (which we intend to disprove) and point out that the admitted premises necessarily lead to a self-contradiction. ¹⁴¹ The following is an example of the form of a proof by contradiction ¹⁴²:

Either A is true or B is true but not both.

If B is true then C is true.

C is not true.

Let us admit, arguendo, that B is true. But that cannot be because C is not true - therefore B is not true, therefore A is true.

In ascii logical syntax:

A/B; (exclusive or)

B=>C; (implication)

-C=>-B=>A. (negation, implication).

In law, reductio arguments are often mislabeled: arguing that a certain interpretation of a law leads to a ridiculous outcome is not reductio ad absurdam; it is reductio ad ridiculum ¹⁴³--the reduction to an impossibility or to the ludicrous (q.v.). Lawyers often misuse the term "reduction ad absurdam" confusing it with reduction ad ridiculum, but such a beginners' error should be avoided by scholars and judges.

As a mode of formal demonstration, reductio ad absurdam arguments *113 are valid. ¹⁴⁴ As a mode of practical reasoning, reductio ad ridiculum arguments state that the position argued for is inadmissible, for it would imply some ludicrous result. ¹⁴⁵ Recall that formal demonstration proves logically necessary implications--what must be the case, from the given premises--not possible implications, i.e. what might be. In contrast, practical reasoning is probabilistic and pragmatic, proving what is most likely the case.

When Aristotle wrote about intuitive reasoning, ¹⁴⁶ he meant this idea: out of the sum of all our experiences and capacities, in the face of any problem, we have an intuitive foreknowledge of the likeliest solution. ¹⁴⁷ Aristotle was insufficiently skeptical toward logical intuitionism. He thought that one's intuitions were just about always, if not always, correct. ¹⁴⁸ I would say that the intuitions we have indicate the possible solution of which we are most aware--not necessarily the best solution, nor even necessarily the only solution. Moreover, our intuitions--our holistic perception of the resolution to the problem--are, sometimes, in error. Intuitions are an excellent heuristic, however. By verifying what we believe to be the best solution, we determine whether in fact it is a solution, the solution, or no solution and may even discover other solutions.

Advances in neuroscience since Aristotle have shown that mental processes are the result of neural linking. ¹⁴⁹ We know that neural links are plastic, ¹⁵⁰ changeable, and that neurons are redundantly networked. ¹⁵¹ We *114 also know that different sections of the brain specialize in different tasks. ¹⁵² Cognitive science also has shown that the brain degrades gracefully: ¹⁵³ the destruction of, for example, a part of the brain which specializes in speech leads to the rerouting of the neurons in the remainder of the brain to best compensate for the lost part of the brain. ¹⁵⁴ Cognitive function is thereby degraded but not entirely destroyed. Neuroscience debates whether cognition occurs sub symbolically or symbolically. ¹⁵⁵ One could model Aristotle's intuitions ¹⁵⁶ as a sub symbolic communication between the various portions of his brain leading to the likeliest conclusion, - which he then empirically verified. All this is discussed so that judges can better understand the nature of their own innermost thoughts in the hope that they will become better at the various cognitive functions involved in adjudication. ¹⁵⁷

It seems that we are no closer to solving our problem of conflicts between fundamental general deductive norms. I wanted first to point out that courts can generate formally valid inferential chains to support just about any outcome in any case. I also illustrated the possible modes of inference (practical or formal; backward or forward chaining), and institutional pressures and limits to properly contextualize the problem of justification. The only way out is to devise a taxonomy of fundamental deductive legal rights and then to see whether that taxonomy helps us to hierarchize rights, or at least generate the warrants needed to determine whether the right to strike is dispositive or if, instead, the freedom of establishment is dispositive.

1. Hierarchizing Rights - A Typology ¹⁵⁸

If we are to solve the problem of conflicts between competing fundamental rights and freedoms, we need to develop a typology to understand what rules are and if, and how, they might be hierarchized.

*115 a. Rules are Conditionals.

Laws are conditional statements--if . . . then---not imperatives. ¹⁵⁹ A law states that "if" a certain condition is satisfied "then" a certain legal consequence follows.

b. Meta Rules are Rules for Making Other Rules.

The first type of rule is what I refer to as "meta rules"--rules to decide how to validly form other rules.

c. (Fundamental)(Constitutional)(Basic)(Human) Rights and Freedoms--Deductive General Principles

Legal systems are formed with and by a variety of meta rules. ¹⁶⁰ Those meta rules have in common that they are general deductive principles and/or that they are procedures for forming other legally valid rules. As a matter of formal logic, general principles of law, fundamental rights, basic rights, constitutional rights, civil rights, are all general deductive principles. I argue that the distinction between "right" and "freedom" is untenable; that "freedoms" are merely an executory right to undertake a certain action. I have argued elsewhere, ¹⁶¹ at length, for what I regard as the tenable distinction between executory and vested rights; between substantive and procedural rights. I regard constitutional rights, general principles of law, and fundamental (human) rights as logically equivalent to each other - they the most general of deductive principles and are hierarchically superior to ordinary laws. General principles, constitutional rights, and fundamental/human rights are further divided into subcategories. Civil rights, e.g., are one particular variety of constitutional right, they are the procedural rights which inhere in citizens to participate in their own governance. Basic rights are also a constitutional right. Basic rights are economic framework rights which constitute the market order of a polity; thus they are positive, not natural and alienable, yet nonetheless are hierarchically prior to other positive laws. ¹⁶²

Fundamental rights have also been characterized as economic rights, political rights, or social rights. ¹⁶³ The problem with distinguishing a right *116 as economic, political, or social, is that some rights fit more than one category; consequently those categories are not always useful for our task, which is to try to categorize rights.

We can distinguish and hierarchize general deductive norm complexes by characterizing the rights in question as natural or positive.

Some rights are natural, i.e. universal in time and space; most rights are positive. Natural rights are hierarchically superior to positive rights. Rights can also be characterized as substantive or procedural. Procedural rights are logically prior to substantive rights, in that procedural rights determine how substantive rights are formed. So those are some categorical ways to think about fundamental rights and freedoms. The hierarchy I propose here may help us determine which right invades which (a question the Viking court never really asked itself).

Where do we place the right to strike and the freedom of establishment in the taxonomy given? The right to strike is not a procedural right, since it does not necessarily determine how other rights are formed. Yet it is a political right: a right to a certain form of political expression. The problem with "political" rights is that often that conclusion is used to sweep away any legal implication of the existence of such a right. Yet, the right to vote, the right to free speech, are also examples of political rights: rights to certain procedures in the democratic state: those political rights are in fact legally enforceable. The political nature of human rights prior to the Second World War incidently meant their non-enforcement and the result was a human rights catastrophe and in the post war world human rights suddenly stopped being "political" . So I do no regard determining a right as "political" to be particularly useful: courts however have and do use that distinction. When we compare "political" rights to the right to vote or the right to public assembly or to speak one's own mind the distinction of a right as "political" and thus "unenforceable" reveals itself as an incoherent conclusion which is dangerous to democracy.

Rights are sometimes called "economic" "social" or "cultural": I do not regard those distinctions as particularly useful, but courts regularly make those distinctions and so we must work with them.

While we could call the right to strike a "political right" we could also call it a basic right, a part of the economic constitution. As we saw, the labor union most definitely contributes to constituting the market economy by acting de facto as an instance of state power for the determination of *117 working conditions. The right to strike is internationally recognized: it is currently a universal right, recognized, to varying degrees, in all places. ¹⁶⁴ For that reason it could be seen as a natural right, an expression, ultimately, of the right of self preservation (just like every other right). ¹⁶⁵ On the other hand, though all societies today may recognize some right of workers to organize collectively, that right has not been universally recognized historically, or even recognized at all. So we could see the right to strike either as inalienable, universal, and natural or instead as positive and political. Which illustrates a general problem: it is difficult to characterize fundamental rights correctly precisely because they are abstract general principles of law.

In contrast to the right to strike, the freedom of establishment is clearly a positive right, not a natural one. Not all societies currently recognize a right of foreign enterprises to establish themselves on their soil. The right of establishment is a basic right, constitutive of the economic order, but a positive right. It concerns alienable economic rights, not inalienable human rights. Both rights could be characterized as alienable and positive not natural, both rights seem to be constitutive of the State's economic order, but only the right to strike could be characterized as currently universal. One could make a plausible argument that the right to strike and the freedom of establishment are hierarchically equivalent as basic rights, aspects of the economic constitution as a matter of positive law. However, if one determines that political rights are superior to economic rights, and characterizes the right to strike as a political right, then the right to strike would enjoy the benefit of burdens of proof and of presumptions in the proportionality inquiry. If one instead determines political rights are nonjusticiable then instead the freedom of establishment prevails.

Because the right of workers to collectively organize is universally recognized I would argue that the right to strike is hierarchically superior to the right to establishment; in that case we should see the actions of the enterprise as infringing the rights of the workers, rather than in the way the court did, as interfering with the freedom of establishment.

Let's change our perspective now and think "intuitively," in a results *118 oriented fashion. There are two possible outcomes. In one outcome, the right to strike displaces the freedom of establishment. In the other outcome, the freedom of establishment displaces the right to strike. Where do those results take society? What do they lead to? I have argued elsewhere that the cases of Viking and Laval are in fact a perfect example of results-oriented thinking. The court decided as it did, in my opinion:

1) to open Western labor markets to Eastern European labor in order to

a) anchor Eastern Europe's fledgling democracies.

b) to allow the law of supply and demand to efficiently allocate surplus Eastern European labor and Western European capital.

2) to break national labor markets and rebuild them as a single European labor market. ¹⁶⁶

The courts' decision in Viking ultimately aims to contribute to the construction of a single integrated European labor market in order to guarantee workers' mobility and create economic interdependence. It seeks those goals, ultimately, in order to prevent wars for market share and resources - by creating prosperity through interdependence and mutual understanding.

Given that desired result--to find that the right of establishment was more important in this case than the right to strike--the court then would simply reason backwards through the variety of possible arguments to reach that conclusion, and then pick the most plausible inferential chain which would reach the desired conclusion. The court could have reached the result, e.g.:

1) by finding the right to strike does not exist in E.U. law (which is implausible).

2) by finding the right to strike is less important here than the right to establishment (which is plausible, but only justified by the desired outcome of a single integrated labor market which would be as productive as possible through free competition).

a) either by calling the labor union a quasi/state actor

b) or by finding direct horizontal effect c) or by finding indirect (persuasive, interpretive) effect

The court in Viking took 2b, though in fact an entire elaborate decision tree could be established for Viking with these nodes:

*119 POSSIBLE CHARACTERIZATIONS OF THE RIGHTS AT STAKE IN VIKING

THE RIGHT TO STRIKE
Horizontal Direct Effect (HDE)      [X]
Horizontal Indirect Effect (HIE)
Vertical Direct Effect (VDE)
Vertical Indirect Effect (VIE)
No Effect (NE)

                                  HDE  HIE  VDE  VIE  NE        (FREEDOM OF ESTABLISHMENT)  

[X] Marks the spot: Viking found the treaty directly effective both as to the right to strike and as to the right of establishment; the direct applicability of the treaty was a case of horizontal effect. ¹⁶⁷

As shown, a grid of 25 possible outcomes as to the relationship between the right to strike and the freedom of establishment; plus four steps of proportionality (rational means, permissible ends, necessary, least restrictive means). Meanwhile as to effect two key possibilities: the labor union and the enterprise alike might be seen as instances of state power or, more plausibly, as private law actors.

The decision in Viking clearly was plausible. However, other outcomes could have been plausibly reached. The decision is justified precisely by the results, but could have been better justified had the court consciously and conscientiously compared and characterized the right to strike and the freedom of establishment in some structured hierarchy of norms.

I have tried to show several interrelated problems:

* the distinctions between E.U. proportionality and U.S. interest balancing.

* the manipulability of the proportionality and balancing tests.

* the institutional strengths, weaknesses, and limits of the judiciary

* the problem of justification.

* the problem and possibility of hierarchizing general deductive principles as a way to determine which right is more important in a case of conflicting fundamental rights.

* the uses and limits of effects based reasoning, i.e. results oriented legal reasoning.

---

Notes:

Penn Central Transportation Co. contains one of the most complete discussions of the Takings Clause. The Court explained that resolving whether public action works a taking is ordinarily an ad hoc inquiry in which several factors are particularly significant-the economic impact of the regulation, the extent to which it interferes with investment-backed expectations, and the character of the governmental action (438 U.S., at 124). The opinion does not repudiate the rule that a permanent physical occupation is a government action of such a unique character that it is a taking without regard to other factors that a court might ordinarily examine." Loretto v. Teleprompter Manhattan CATV Corp. 458 U.S. 419, 432, (1982).

"First, the private interest that will be affected by the official action; second, the risk of an erroneous deprivation of such interest through the procedures used, and the probable value, if any, of additional or substitute procedural safeguards; and finally, the Government's interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedural requirement would entail.' Id., at 335, 96 S.Ct. 893; Wilkinson v. Austin, 545 U.S. 209, 224-25 (2005).