Dworkin’s ‘law as integrity’

Dworkin’S ‘Law as Integrity’

Following his exposition of the methodological claims of fit and best light, Dworkin moves to propose his own conception of law – the theory he believes best fits legal practice and puts it in its best light. This theory – ‘law as integrity’ – describes legal interpretation essentially as follows: the legal interpreter first arrives at a set of moral principles that fit the ‘institutional legal materials’ and put them in their best light; and then, using those principles, the interpreter establishes the correct legal rights and duties. Thus (confusingly enough for the reader, though perhaps unsurprisingly), Dworkin’s conception of law reintroduces the fit and best light requirements as criteria for the correct set of moral principles from which all legal determinations derive. Section I explains Dworkin’s theory and examines its merit (concluding that it presents an unintelligible description of what ‘integrity’ is), while section II addresses the relation between ‘law as integrity’ and the impartiality of legal interpretation. As we shall see, according to ‘law as integrity’ proper legal interpretation is always impartial.

‘Law as Integrity’: Presentation

As previously discussed, Dworkin maintains that a theory of legal practice must hypothesise a function or a purpose (however vague or abstract) that is served by the law.¹ Now according to Dworkin’s own theory, the purpose of the law is the justification of state coercion: the law is aimed at justifying the way in which the state exercises its coercive powers. The law fulfils this purpose, says Dworkin, by demanding that every legal requirement conform to certain moral principles. What moral principles are these? These are the moral principles which best ‘fit and justify’ the ‘institutional legal materials’. The ‘institutional legal materials’ are all the statutes and precedents and regulations of administrative agencies and other such materials from official legal sources. The legal interpreter, examining these legal materials, must arrive at a set of moral principles that best fits these materials and puts them in their best light. For example, the principle that a man may not benefit from a wrong he committed is presumably a principle that fits the legal materials (it underlies many statutes and many judicial doctrines), and it puts these materials in their best light (it makes these materials look good and desirable). Any legal requirement must then conform, in some way (we will examine how), to this principle. So the method with which legal interpreters arrive at the correct legal rights and duties is essentially this: they first arrive at a set of moral principles which is evidenced in the institutional legal materials (that is, that fits the legal materials) and which makes these legal materials appear good and desirable (that is, that puts these materials in their best light); and they then proceed to determine what the law requires by making sure that any legal requirement conforms to these moral principles.

So according to ‘law as integrity’ all legal requirements must conform to the set of moral principles that best fits legal practice and puts it in its best light. It is essential for Dworkin that all legal requirements conform to one and the same ‘set of moral principles’ or ‘scheme of justice’ (Dworkin uses these two expressions interchangeably, and so will I).² The law, says Dworkin, must ‘speak with one voice’: a proposition is legally valid only if it conforms to the same set of moral principles to which all other legal propositions conform. So long as all legal propositions are interrelated in this way, the law ‘speaks with one voice’, because any utterance it makes agrees with the set of moral principles with which all its other utterances agree.

Now what is meant by the idea of one ‘set of moral principles’ or one ‘scheme of justice’? Dworkin elaborates on these notions in the following way:

Thus ‘integrity’ – which is the name Dworkin gives to the demand that the law ‘speak with one voice’ – does not require that all our laws conform to a set of moral principles which never conflict among themselves in their practical counsels (this would be a rather absurd demand); instead, integrity demands that a certain chosen priority or ranking among moral principles be maintained – throughout the particular legal scheme, and throughout the legal system as a whole.

Dworkin argues for this understanding of law and its test of legal validity along the two dimensions he believes to constitute the criteria for the correct theory of law – fit and best light. (As we saw, the best light requirement is misconceived; but Dworkin’s elaboration of the claim that ‘law as integrity’ puts legal practice in its best light will help us better understand – or try to understand – what Dworkin means by ‘integrity’.) Dworkin’s principal argument in defence of ‘law as integrity’ appeals to our instinct: our instinct demonstrates that we value the idea of integrity, and that we expect our laws to respect it – even if we have never given ourselves a conscious account of this. ‘Law as integrity’ is a conception that fits the legal materials, as an investigation would presumably show, because we already use integrity in shaping our law; and ‘law as integrity’ puts legal practice in its best light because, as Dworkin will argue, integrity is an important value of political morality – which is why we appreciate and respect it to begin with. In other words, Dworkin’s theory purports to bring to light a hitherto unarticulated but existing understanding of law: ‘Astronomers’, says Dworkin, ‘postulated Neptune before they discovered it. They knew that only another planet, whose orbit lay beyond those already recognised, could explain the behaviour of the nearer planets. Our instinct … suggests another political idea standing beside justice and fairness. Integrity is our Neptune’.⁴

What is the ‘instinct’ of which Dworkin speaks, and which occupies such an important place in the argument for integrity? That instinct, says Dworkin, manifests itself in our rejection of legal ‘checkerboard solutions’. He writes:

The reason why we feel that it shouldn’t, says Dworkin, lies with our respect for integrity: ‘If there must be a compromise because people are divided about justice, then the compromise must be external, not internal; it must be a compromise about which scheme of justice to adopt rather than a compromised scheme of justice.’⁶ The law must ‘settle on some coherent principle’ and never ‘affirm for some people a principle it denies to others’.⁷ An example of a law settling on a coherent principle, says Dworkin, is a statute prohibiting abortion which contains an exception for pregnancies caused by rape; an example of a law failing to settle on a coherent principle is a statute permitting abortion only to ‘women born in one specified decade each century’: only the former constitutes a ‘principled’ statutory scheme.⁸

Principled V Unprincipled Statutory Schemes

Looking at Dworkin’s abortion statutes, it seems that the crucial difference between the ‘principled’ statute and the checkerboard statute (which permits abortion only for women born in the 1950s) is that in the former, but not in the latter, we identify a justification for the different treatment the statute accords the two categories. The question to which we demand an answer – and which only the former statute can supply – is ‘what is it in the class of cases that falls within and that which falls without the statute’s directive that accounts for this distinction between them?’. It is presumably the absence of such a justification which makes the checkerboard statute unacceptable to us. This idea, as Dworkin also notes, is expressed in the notion of ‘equality before the law’: the similarly situated are to be treated similarly. Equality before the law appears to demand that the different treatment accorded different categories be justified by the difference between these categories. In short, we expect the distinctions our statutes make to be principled distinctions, and neither of the distinctions made by Dworkin’s examples of checkerboard statutes is principled.

As we shall see, Dworkin sees things somewhat differently; but he does seem to recognise that the problem with checkerboard statutes is that they lack a reason for treating their classes differently: ‘Shall we just say’, he says, ‘that a checkerboard solution is unjust by definition because it treats different people differently for no good reason, and justice requires treating like cases alike? This suggestion seems in the right neighbourhood …’⁹ What does Dworkin mean here by ‘a good reason’? He says that so long as we can ‘recognise’ a reason for treating different people differently, a legal scheme is not a checkerboard solution – even if we think that reason weak or even plain wrong: ‘Suppose you think abortion is murder and that it makes no difference whether the pregnancy is the result of rape. Would you not think a statute prohibiting abortion except in the case of rape distinctly better than a statute prohibiting abortion except for women born in one specified decade each century? … You see the first of these statutes as a solution that gives effect to two recognisable principles of justice, ordered in a certain way, even though you reject one of the principles.’¹⁰ The former statute is not a checkerboard statute (and is therefore ‘distinctly better’) because the exception it contains is grounded in a recognised reason for treating different people differently (that reason being the woman’s complete lack of responsibility for the pregnancy, and perhaps the misery awaiting the unwanted child) – even if that reason (or principle) is one we reject. In an appended footnote Dworkin adds: ‘We can easily imagine other examples of compromises [that is, compromised statutes] we would accept as not being violations of integrity because they reflect principles of justice we recognise though we do not ourselves endorse them.’¹¹ It is enough that a statute’s drawn distinctions are supported by principles we can recognise – even if not endorse – to make it a principled (rather than a checkerboard) statute. In short, in contradistinction to statutes whose distinctions are grounded in recognised principles, checkerboard statutes draw distinctions which are arbitrary: there are underlying principles for treating pregnancies caused by rape differently from other pregnancies in regulating abortion; but there seem to be no underlying principles for treating women born in one decade differently from women born in another.

Does Integrity Fit the Legal Materials?

Recall that ‘integrity’ demands that all legal requirements conform to one ‘set of moral principles’ or one ‘scheme of justice’. Now given what underlies the distinction between checkerboard statutes and principled ones, between internal and external compromises, how do checkerboard statutes (and our instinct in regard to them) support the case for integrity? Dworkin’s answer is surprisingly straightforward. What underwrites our aversion to checkerboard statutes, he says, is our respect for integrity: ‘The most natural explanation of why we oppose checkerboard statutes appeals to [integrity]: we say that a state that adopts these internal compromises is acting in an unprincipled way … [I]t must endorse principles to justify part of what it has done that it must reject to justify the rest.’¹² Checkerboard statutes, in other words, do not conform to one scheme of justice: we expect our law to ‘speak with one voice’ (that is, to follow from the same set of principles throughout), and in checkerboard statutes we hear the sounds of various voices.

But far from being ‘natural’, this explanation is quite baffling: there is a sea of difference between the expectation that statutes draw distinctions which conform to some recognised principle (which is what is lacking in checkerboard statutes), and the expectation that statutes draw distinctions which conform to the very same set of principles to which all other legal rights and duties conform. Obviously, a statutory distinction may conform to a recognised principle which does not belong to the set of principles to which all other legal requirements conform. In that case that statute would not be a checkerboard statute; but it would certainly violate integrity (for, to repeat, integrity demands not only that legal requirements be supported by recognised principles, but that those principles be equally in force throughout the legal system). In fact, Dworkin’s very examples of ‘principled’ statutes – those statutes which do not offend our checkerboard instincts – may very well fail to respect integrity. It is true, of course, that the presence of a recognised principle is a prerequisite for the very possibility of Dworkin’s integrity: if laws habitually exhibited the unprincipled distinctions we see in checkerboard statutes, integrity could not have even been a possibility. There can be no unprincipled statutory scheme which respects integrity. But there can be principled statutory schemes which do not. So our checkerboard instincts fall far short of endorsing integrity’s holism: they can be appeased where integrity is lacking, and it therefore cannot be the case that what gives rise to our aversion to checkerboard statutes is our aspiration for integrity.

So in fact we are dealing here with two independent notions (on the one hand the idea that treating seemingly similar people differently need be justified by a recognised principle, on the other – integrity, the idea that treating seemingly similar people differently need be justified by the set of principles justifying all other legal rights and duties), and no quick inference can be made from the respect for the one notion to the respect for the other. Yet Dworkin collapses these two notions, simply believing them to be one (thinking that the lack of integrity is simply the problem with checkerboard broadly applied: lack of integrity is arbitrariness generalised from the case of specific statutory schemes to the law as a whole, and the aversion for checkerboard solutions is but a particular case of the general impulse for integrity).¹³ This, however, appears to be a mistake.

Dworkin repeatedly attempts to explain the problem with checkerboard statutes by appealing to the ideal of integrity: he claims, for example, that with checkerboard statutes: ‘one principle of justice is not outweighed or qualified by another in some way that expresses a ranking of the two. Only a single principle is involved; it is affirmed for one group and denied for another, and this is what our sense of propriety denounces.’¹⁴ (Lack of integrity is, presumably, precisely the manifestation of this latter problem: that a certain principle, or a set of principles, is denied in one case and affirmed in another.) But this identification of the two problems does not work (the two, as I said, are distinct). The distinction between checkerboard statutes and principled ones cannot be a distinction between legal schemes which ‘affirm’ a principle to one group and deny it to another, and legal schemes that do not: both checkerboard statutes and principled ones ‘affirm a principle for one group and deny it to another’. For instance, in the rape exception abortion statute (Dworkin’s example of a principled statute) the principle to which the rape exception conforms (presumably, the principle of individual autonomy) is affirmed for one group of women (those whose pregnancies are caused by rape) and denied to all others. To claim that the checkerboard/noncheckerboard distinction revolves around whether a statute ‘expresses a ranking of conflicting principles’ or ‘affirms conflicting ones’ is merely to play with words.

Here is another unsuccessful attempt to explain the problem with checkerboard statutes in terms of integrity. Principled statutes, says Dworkin, present the same ranking of principles throughout: ‘A scheme of inheritance taxes might recognise [two conflicting] principles in a certain relation by setting rates of tax that are less than confiscatory. But we insist that whatever relative weighting of the two principles the solution assumes must flow throughout the scheme …’¹⁵ This, once more, is precisely what integrity demands. Presumably, our checkerboard instincts are not offended by the scheme of inheritance taxes – despite the fact that it conforms to conflicting principles – because it manifests the same ranking of principles from beginning to end. Yet, the notion of a ‘ranked set of principles’ is always defined, by Dworkin, by reference to the legal scheme as a whole: the ‘ranked set of principles’ in the principled abortion statute which exempts rape-caused pregnancies includes the principles prohibiting abortion as well as the principles sanctioning them; and similarly here – the ‘ranked set of principles’ is given by the entire scheme of inheritance taxes. Consequently, it appears to be a logical impossibility to observe a legal scheme which does not manifest a consistent application of a set of principles throughout: the set of principles manifested in any legal scheme is necessarily the set constituted by that legal scheme in its entirety. The distinction between checkerboard and non-checkerboard statutes simply cannot be a distinction between a legal scheme which manifests the same ranking of principles from beginning to end and a legal scheme which does not. Indeed the example of a legal scheme which ranks conflicting principles but still violates integrity by failing to respect that ranking ‘throughout’ is one we badly need to see, but never get to.

There are more examples, but they are all symptoms of the same unhappy condition: Dworkin merges two independent ideas – the lack of a recognisable reason, or principle, for a drawn distinction (the checkerboard idea), and the lack of a recognisable reason, or principle, that is equally applicable all over the place (integrity). And even though there is some affinity between the two notions, it is certainly not the identity that Dworkin purports to see. Why does Dworkin think the two are identical? I think Dworkin believes that some notion of rationality underlies both the aversion towards unprincipled distinctions and the respect for integrity. As we saw, the problem with checkerboard statutes is ultimately their inability to justify the distinctions they draw between categories: they simply draw arbitrary distinctions. Now I think that Dworkin thinks – though he does not say so, nor does he attempt to defend this position – that the problem with the lack of integrity is that it, too, must ultimately evidence arbitrariness: just as checkerboard statutes are arbitrary because they cannot justify treating one class in one way and another in another, so is it arbitrary to act according to one ‘moral scheme’ (or a ‘set of principles’) on one occasion and a different ‘moral scheme’ on another.¹⁶

The obvious problem with this hypothesis (which merges the arbitrariness of checkerboard statutes with the alleged arbitrariness of a lack of integrity) is that unlike the arbitrariness evidenced in checkerboard statutes, which we perceive as morally unacceptable, the kind of arbitrariness the lack of integrity is supposed to manifest is both morally acceptable and widely practised.¹⁷