Do Judges have Discretion? Is there a Right Answer?

My review of the issues involved in the interpretation of law starts with a critique of the well-known thesis of the contemporary legal philosopher Ronald Dworkin. Dworkin’s position at its highest¹ is that there is always one best, right answer to questions of judicial interpretation.

From this alleged fact of a theoretically discoverable ‘right’ answer to any question of law it follows that judges have a duty to find that answer and that one of the parties has a right to a decision in her favour.⁵ And as the law always yields an answer there is no need for judges to legislate⁶ (z.e. to look to policy); their job is to ‘discover’ or ‘find’ this existing ‘right’ answer (i.e. by weighing principles).

How are Dworkinian judges to ‘find’ these ‘right’ answers? The task is massive. It involves “... constructing] a scheme of abstract and concrete principles that provides a coherent justification for all common law precedents and, so far as these are to be justified on principle, constitutional and statutory provisions as well.”⁷ T o undertake this task Dworkin imagines a superhuman judge he names Hercules.

I am interested in Dworkin’s interpretive theory predominantly on the prescriptive level. Is it an approach to be recommended to judges? Subsumed within that question lies an evaluation of the coherence of the notion of ‘one right answer’ in judicial interpretation. However as for Dworkin’s theory on the descriptive level, as a claim about how judges (in common law systems at any rate) in fact do interpret the cases that come before them, I merely make five comments. Firstly, it is not absolutely clear whether Dworkin actually does advocate his theory as the best description of actual judicial practice.¹¹ Secondly, if he does, then resolution of that claim is a purely empirical matter needing only testing against decisions in actual cases. The number of conflicting cases needed to disprove the Dworkinian descriptive claim would be a matter of dispute. Thirdly, as a descriptive theory only, the use of a mythical figure, a mythical figure of unparalleled abilities, as the only competent utilizer of the theory seems to undermine its practicality. The theory would be better with Hercules omitted. Fourthly, I note and adopt Mackie’s six comments¹² about Dworkin’s theory as a descriptive theory including Mackie’s observation that what judges say they are doing, what they think they are doing, and what an accurate, impartial description of what they are doing would be are all distinct and need not converge.

I want to return, though, to the question of whether judicial decisions ought to be based on Dworkin’s theory of interpretation. I can do this best in several stages. I begin by examining judicial discretion and the underpinnings required to make the ‘one right answer’ doctrine coherent. Having brought them into the open I then consider Dworkin’s responses to the challenge from scepticism. Finally I weigh the benefits and costs of adopting Dworkin’s theory of interpretation.

In my view the ‘one right answer’ thesis and its concomitant, that judges have no strong discretion in deciding cases, is only coherent in the framework of an acceptance of moral objectivism or moral realism. According to Dworkin, what the law is — at least in Hard Cases — depends on the judge’s evaluation of what larger theory of abstract and concrete principles best explains and justifies the settled law. This evaluation of what is best is explicitly not an empirical determination of established norms or majority opinion.¹⁴ Rather, it is an evaluation by the judge of what theoretical foundation she thinks provides the best ‘fit’ with the existing settled law. And the conclusion generated by application of this ‘best fit’ theory need not correspond with the judge’s own preferred outcome of the case.¹⁵ But the process of ‘finding’ a best explanation and justification does involve recourse to a moral theory, a conception of “... the political morality presupposed by the laws and institutions of the community.”¹⁶ Nor is there any escaping the inevitable fact that the judge must “rely on his own judgment as to what the principles of that morality are.”¹

It seems quite evident, on this view, that a Dworkinian judge’s evaluation of what the law is (at least in Hard Cases) depends on what that judge believes is the best political morality consistent with the settled standards.

This first query is analogous to enquiring whether Dworkin’s theory of interpretation requires the support of natural law doctrine — i.e. that there is a universal, immutable, ‘higher’ law, open to discovery by and accessible to human reason, which constitutes a law of nature¹⁹ and provides the content of an objective morality. Thus, does Dworkin really mean to say that only that which accords with, or at any rate does not infringe, some ‘higher’ or objectively ‘right’ criteria can ultimately be part of the ‘settled’ law of a land?

Focusing solely on this first queiy it is my view that Dworkin’s theory does not require a natural law or moral objectivist backing or support. Dworkin can evade the issue at this stage. He could, for example, quite well admit that all law is made by human beings — formally by statute and precedent or informally by evolving principles and conventional practices — and that the background ‘settled’ law from which Hercules works will vary from jurisdiction to jurisdiction and from time to time.

Unfortunately Dworkin does not unambiguously do this. There are continual hints, or at least suggestions open to alternative readings, that these reputedly contingent, historical, ‘settled’ standards are in fact open to evaluation against some higher, mind­independent, extra-best fit standards.²¹ Similarly, what else is one to think of a passage that apparently condones judicial lying when moral standards are felt strongly to conflict with settled law?

If the judge decides that the reasons supplied by background moral rights are so strong that he has a moral duty to do what he can to support these rights, then it may be that he must lie, because he cannot be of any help unless he is understood as saying, in his official role, that the legal rights are different from what he believes they are.²² What type of standards are these ‘background moral rights’ that they provide a basis for lying? Dworkin does not say, but the hint of moral objectivism is not far from the surface.

And of course Dworkin’s constant and repeated usage of the verbs ‘to find’ and ‘to discover’ must suggest to many, although admittedly they need not be meant or read this way, that there are right moral principles ‘out there’ whatever the particular system or settled laws. Indeed it is, in my opinion, an element in Dworkin’s appeal that he points in two directions at the same time. Consistently his thesis is formulated in a way that is open to contrary readings.²³

Nevertheless, it is not strictly relevant to the merits of Dworkin’s thesis that a certain ambiguity exists in the way he himself frames that theory. We can follow his advice and put his own theory in its best light. Of course the Hercules thesis would be much simpler were it annexed to some form of openly admitted natural law doctrine or moral realism.

On this assumption our Herculean judge operates within some system or other of historically contingent man-made rules and principles, where the ‘settled’ law of that system is to be accepted as a given and the ‘best’ explanatory and justificatory theory of it formulated and applied. My second query thus becomes whether Dworkin’s theory of interpretation is plausible in its denial of all elements of subjectivity when interpreting from within, and only from within, that particular well-developed legal system. Or phrased differently, can a theory of interpretation which demands recourse to moral evaluation (in framing the theory to be used to ‘find’ that society’s ‘right answer’) involve no strong judicial discretion?

Notice that where the answer in an individual case before a judge is said to depend (at least in Hard Cases) on the ‘best’ political morality, on the moral framework that ‘best’ accounts for the ‘settled’ laws, then there can be ‘one right answer’ only if the settled standards point in every case to one correct moral evaluation. This is true even though Dworkin, in advocating that any principle (legal or moral) may form part of a judge’s deliberations, is careful to caution his readers that the weight to be attached to any particular principle will depend on the level of institutional support which it enjoys. That is to say, although for Dworkin any principle may be relevant to a legal decision, the weight it is accorded depends on its institutional support which is, in turn, related to the content of the positive law. Dworkin’s ‘best fit’ political morality, in other words, is not simply a matter of the most attractive political morality; it is constrained (more or less) by the jurisdiction’s positive law. Might not this Dworkinian committment to the institutional nature of law imply that there can be (even in Hard Cases) ‘one right answer’ without necessarily pre-supposing one correct moral answer?

The answer is no. In some instances at least the decisive factors for Dworkin’s Herculean judge will be moral ones. Hercules will have to decide, say, that one moral principle is weightier than another though both have institutional support. And perhaps more crucially, evaluating degrees of institutional support will often depend on Hercules’ earlier formulation of the ‘best’ political morality and hence itself involve a moral determination — one background political morality accounting for the settled laws being ‘better’ than the others according to Dworkin. No, it is clear on any reading of Dworkin that there can be ‘one right answer’ in all cases only if the settled standards point in every case to one correct moral evaluation. So to elucidate the further assumptions needed to make that conclusion coherent and plausible is my second quest.

Dworkin’s thesis reduces to the bare proposition that because no judge will reach the stage where either the ‘settled’ rules themselves (Easy Cases) or the political morality drawn from the ‘settled’ law (Hard Cases) provide no further reasons for deciding one way rather than another, it must follow that one answer is right. At the risk of over-simplifying, Dworkin’s image is of a balance on to each side of which weighted reasons are piled. ⁷ Sooner or later one side will win. A tie, although theoretically possible, is less and less likely the more sophisticated the given legal system to the point where ties in a well- developed legal system (as in the United States) are in practice impossible.²⁸ Judicial decision-making, on this analogy, is implicitly a reason-based, ordering affair however multi-dimensional,²⁹ complex and massive the actual Dworkinian interpretive exercise.³⁰

The fundamental objection to which Dworkin must respond asserts that moral standards and principles are not as clear and settled as positive enactments and judicial precedents — in fact, that even from within a well-established system of settled laws moral evaluation has an ineluctably subjective element. Different judges, with different personalities and convictions — not to mention unique combinations of upbringing, preferences, biases,³¹ experiences, and values — will sometimes arrive at different conclusions in the same case. What else must be accepted in order to assert that all but one of these conclusions (or perhaps all of them) is wrong?

One might have thought that this objection could be side­stepped by making a distinction between a) the grounds on which individual judges base their reasoning and b) the question of whether or not a judicial decision is objectively justifiable by reference to legal rules and principles. The strategy here would be to assert that the admitted subjectivity and variability of a) does not filter across to b). Somehow, in other words, there can be ‘objectively’ right decisions (given the existing legal rules and principles) in a world in which individual judges reach different conclusions.

In my opinion, however, this strategy fails. As we shall see, all that can salvage Dworkin’s theory of interpretation at this second stage from within a well-established system of settled laws is a wholehearted embrace of moral objectivism or moral realism. I do not believe that he can evade this issue any longer. There is, I think, no escaping the fact that Dworkin conceives of moral³² evaluation in terms of the accumulating of reasons, albeit of weighted reasons. The fact that different judges reach different conclusions is attributed to varying and diverging calculations of the substantive “political theory”³³ that best “fits”³⁴ the settled laws concerned. This calculation is to be done on the basis of the weight of accumulated ‘reasons’. But whatever these ‘reasons’ are, and Dworkin is vague about this, they cannot include feelings, passions, sentiments, inclinations or preferences. None of these is rational. More importantly, all of them vary to some extent between individuals. They are relative and subjective. Dworkin cannot accept my areasonable,³⁵ sceptical moral theory — or indeed any version of moral subjectivism — without also jettisoning the theoretical coherence of his ‘one right answer’ doctrine. If at some fundamental level areasonable passions, sentiments and preferences are involved in moral evaluation then Dworkin’s claim that judges (who, he insists, are to look to morality) never have strong discretion must be wrong. So where at the first level of enquiry Dworkin could conceivably evade taking a stand, at this second level, from within an established legal system, he must opt for ‘objective’, ‘higher’, mind-independent moral values that can be reasoned to or some other way discovered or else his theory of judicial interpretation collapses.

b)