David Lyons’s 1973 paper “On Formal Justice” begins unpromisingly:
A number of legal and political theorists have suggested that public officials who fail to act within the law that they administer act unjustly. This does not mean that injustice is always likely to be done merely because it often happens to be done when officials depart from the law. Some writers have held that injustice is done whenever an official fails to act within the law, regardless of the circumstances. I shall call this type of view “formal justice.”
This is odd, to say the least, as it implies that (formal) injustice is done when a judge breaks a speed limit but not when she decides two similar cases in wildly different ways. In point of fact, the focus of the paper as a whole is unclear; broad formulations such as ‘official deviation from the law’ are frequently used, implying a contrast between official obedience to the law and actual law-breaking, but so too are narrower formulations such as ‘adherence to existing legal rules’. This second concept – implying a rule-driven approach to applying the law – seems to be the focus here:
Such a view may be considered “formalistic” because it places value, in the name of justice, on adherence to existing legal rules without regard to “substantive” factors such as their contents, the consequences of obeying them, their defects or virtues, or any other circumstances of their application. The only condition imposed is that an official must by law follow the rule in his official capacity. Furthermore, those who attempt to account for this view believe that the requirements of formal justice rest directly on such notions as “proceeding by rule” or “treating like cases alike,” which are thought to be at the heart of our shared concept of justice. The basic requirements of formal justice are thus supposed to be exempt from the controversy over substantive principles of justice and their possible justification.
It is also the focus of the third section of the paper, on Hart. In The Concept of Law, Lyons argues, Hart offers three points in support of ideas of administrative justice. (As we have seen, all three of these had been made in less qualified and more moralised form in Hart’s 1958 paper.)
The first bases administrative justice on the precept “treat like cases alike”; the second grounds it on a notion with which the first is often confused, namely, following a rule; the third is rooted in the idea of impartially applying the law to particular cases.
Should we treat like cases alike? Perhaps so, but what does it have to do with justice? Lyons argues that treating like cases alike so has no intrinsic relationship with justice as an outcome – given that the same principle is involved in duties such as promise-keeping and helping those in need – but this is either irrelevant or an equivocation: the point is not that this principle is conducive to justice of outcome but that it is (perhaps) a form of procedural justice, which can be used equally well to measure how justly the law is applied and how justly we carry out our other social obligations. He also points out (as Hart had done) that no two cases are ‘alike’ or ‘unlike’ in and of themselves: two cases can only be judged to be alike by the application of a given set of criteria. In the case of the law, Hart had argued, the criteria to be applied are precisely the rules of the law. Hence:
From the premises that justice fundamentally requires a uniform treatment of cases and that the law prescribes one way of uniformly dealing with them, we are asked to conclude that justice in the administration of the law requires officials to follow the law. But this argument begs the question at issue, which is whether the pattern of treatment prescribed by law is identical (or even compatible) with the pattern required by justice. Once we realize that the justice of a law is not determined by the law, or in other words that the resemblances and differences between persons, acts, and circumstances which the law tells us to consider are not necessarily the ones that justice says we may consider, the error of the formalist becomes obvious.
Lyons continues: “Why should we suppose that the pattern of treatment prescribed by the law is the same as (or even compatible with) that prescribed by any principle of justice?” (emphasis in original). The only argument he offers in favour of doing so is the hypothetical argument that following the law might be the only way of achieving the “uniform treatment of cases” required by justice – but, of course, this is clearly not the case. Acquitting the defendants who offer a bribe and convicting everyone else is treating like cases alike.
Treating like cases alike – where likeness is given by the criteria of the law – is dealing with them uniformly. But, as the bribery example demonstrates, dealing with cases uniformly is not justice unless the criteria used to define like cases are themselves just – and in the case of the law, we have no necessary reason to assume this. Is this a valid argument? It’s certainly persuasive, but I suspect it rests on a blurring of different senses of the word ‘just’ – and an undervaluation of the procedural nature of the justice being analysed. Suppose that, as the model of formal justice requires, we bracket out the justice of the laws and of the outcome of legal decisions; to make it simpler, suppose that we’re dealing exclusively with unjust laws and decisions with bad and unjust outcomes. Let’s say that having a surname beginning with Q has been declared a capital crime. Alternatively, suppose that we’re dealing exclusively with correct and welcome decisions on just laws: acquittals of falsely-accused murder suspects following an inquiry into police corruption, say. In the second case, we know that the law is just and that a just decision is, at least, highly likely; in the first, we know that the criteria given by the law are not just, and that the possibility of a just outcome is vanishingly small. Can we still speak of injustice being done by a capriciously varied application of the law – perhaps, if the judge delays three days before passing sentence, rules on a second case in five minutes flat and reads the third sentence in a silly voice? This, surely, would be a violation of fair official treatment of which even the acquitted defendants could complain, and which would make the position of the defendants in the nightmare scenario still worse. (Should this experience of stress and uncertainty be considered part of the ‘outcome’ of the case? Surely not – this would collapse a clear distinction (between outcome and procedure) for no real gain.) There seems no reason not to think in terms of procedural justice as a criterion for the application of the law – or, perhaps more precisely, for the process of the application of the law.
Lyons’s second argument addresses the related concept of rule-following, which – as we have seen – is generally required in order to make sense of the idea of treating like cases alike. (Lyons notes that judges can “devis[e] a uniform treatment of cases even when no relevant rules exist, for example, by comparing current cases among themselves”, but this puts too much weight on the need for rules to be stated formally and explicitly.) On rule-following and justice, Lyons is, again, sceptical:
The argument turns entirely on the notion of applying a rule to particular cases; it contains no further restrictions. If the result were a principle of justice, then any deviation from any rule that one is supposed to apply would be, in itself, an unjust act. Nothing restricts this mode of argument to the conduct of public officials, or even to the law.
Insofar as official nonconformity to law is regarded merely as the failure to follow rules, it is implausible to regard it as a kind of injustice. Is there anything else essential to official noncompliance that would provide the required link? It must be something essential to this kind of rule breaking, that is, something independent of all circumstances. Otherwise, a formal justice claim cannot be supported, for formal justice maintains that official disobedience is always morally objectionable, regardless of the circumstances.
Following a rule faithfully cannot be classed as justice, because rules are followed in all walks of life and may have nothing to do with justice. However, the question of equating rule-following with justice only arises in the context of duties or functions with effects on multiple other people – activities in the context of which it makes sense to talk about justice and injustice (Lyons’s counter-example of following the rules of grammar is question-begging). The “something essential” is the topic which was originally in question, that of following the rules given by the law. Once this is granted, this argument reduces to the previous one – that there is no procedural justice in the uniform application of unjust criteria – and is equally false.
Lastly, Lyons considers the argument for impartiality as a component of formal justice. He quotes Hart:
To say that the law against murder is justly applied is to say that it is impartially applied to all those and only those who are alike in having done what the law forbids; no prejudice or interest has deflected the administrator from treating them “equally”.
Indeed, it might be said that to apply a law justly to different cases is simply to take seriously the assertion that what is to be applied to different cases is the same general rule, without prejudice, interest, or caprice.
Contra Hart, however, Lyons maintains that the idea of impartiality has no necessary relationship with that of proceeding by rule. Uniformity does not entail impartiality: “[a]lthough impartiality may require some kind of uniform behavior, merely to deal with cases in a uniform manner is not to be impartial.” Indeed, uniformity understood ‘mechanically’ allows no scope for judges to show either partiality or impartiality. Only if the judge has a choice can the choice be made impartially; but
[i]f the formalist also believes that the choice of lawful alternatives is subject to criticism in the name of justice, then he must qualify his formal justice claim accordingly, because the simple requirement that officials act within the law does not enable the formalist to differentiate between the lawful alternatives.
Here Lyons appears to be making rather heavy weather of the ‘formalist’ claim that rules (with all their grey areas) should be applied, but applied impartially – both-and, not either-or. (Rather confusingly, Lyons uses ‘formalist’ as shorthand for ‘believer in the inherent virtue of formal justice’. He appears himself to be a formalist, in the more conventional sense of ‘analyst of a phenomenon in terms of its forms rather than any inherent qualities’.) When he pursues the topic of impartiality further, his argument becomes confused.
The claim that administrative justice requires impartial application of the law to particular cases is not inherently formalistic. One might agree, for example, that the just way of applying the law is the impartial way, while believing that justice may sometimes require that officials not apply the law. The formalistic version of the claim maintains that impartial application of the existing rules of law fully embodies administrative justice, with the understanding that this claim fundamentally requires officials to act within the limits laid down by law.
Certainly justice may sometimes require that officials not apply the law, but procedural justice – justice in the process of the application of the law – cannot. To say that “the just way of applying the law is the impartial way” is to make a claim about procedural justice, which is not at all affected by claims about the justice of the law itself or of particular outcomes. What Lyons means by the formulation “fully embodies administrative justice” is unclear, but it seems to suggest some claim about the justice of the process as a whole – not the justice with which it is administered.
Let us assume that officials should, to do justice, be impartial; this does not imply adherence to any particular set of rules, such as the rules of law. Again, suppose that the only just way of applying the law is the impartial way; it does not follow that an official who fails to follow the law acts unjustly. Let us agree that an application of the law which is not impartial is unjust; it does not follow that all deviations by officials from the law are unjust. For not every such departure could be described as an application of the law that fails to be impartial. An official might deliberately refuse to follow the law; this is not the same as applying it in, for example, a biased or prejudiced manner. This distinction is important, for the official may refuse to follow the law on principled grounds, precisely in order to prevent an injustice of which he would be the instrument.
The first claim here is irrelevant; the topic at issue here is precisely impartiality in following the rules of the law. The second is a non sequitur, which seems to broaden the argument unhelpfully to official deviancy in general. The remainder of the paragraph is valid, but what it expresses is simply the distinction between procedural injustice and injustice of outcome. It’s open to Lyons to argue that there is no procedural injustice in cases where justice of outcome is unaffected, but – as I argued above – this is only tenable if we effectively define ‘outcome’ to include all the effects on an individual of being involved in a court case, which is surely far too broad.
Next, Lyons on Fuller; then back to Hart.