The “Rule of Law” is an ideal that features prominently in contemporary political discourse. Departures from the Rule of Law are regarded as flaws or failings in a legal system, or at best regrettable necessities that carry a significant burden of justification. The Rule of Law is often considered to be the distinctive virtue of legal systems, and a prerequisite for the existence of a good or decent system of law. But while all (or almost all) agree that the Rule of Law is an indispensable aspect of a worthwhile legal system, there is less agreement on the content and scope of the ideal. It is normally thought to include such elements as official conformity to the law; the independence of the judiciary; laws being clear, prospective and public; and the ability to challenge the applicability of a law in a fair hearing (see Jowell 2007). Nonetheless, lawyers and theorists debate the merits of “formal” or “thin” accounts of the Rule of Law over “substantive” or “thick” accounts, without agreeing among themselves about the details of these accounts (see Tamanaha 2004, chs. 7–8). Formal and thin accounts of the Rule of Law focus on various formal and procedural features of the law, whereas substantive and thick accounts add some requirements of substantive justice, or human rights, or democratic processes.
In this article I will defend a restricted conception of the Rule of Law that is not limited to formal or procedural features, but does not embrace the general protection of human rights or substantive justice or democracy. So the account is neither thick nor thin, neither substantive nor formal, but something in between. The account is grounded in the fundamental idea that the Rule of Law serves to promote and protect governance by law under non-ideal conditions. I will begin by considering Lon Fuller’s and Joseph Raz’s highly influential analyses of the Rule of Law, and the limitations of those accounts. I will go on to propose an analysis of the Rule of Law that is broader than theirs and yet still narrower than “thick” or “substantive” accounts. Finally, I will argue that the value of the Rule of Law ultimately depends upon the overall moral merits of a legal system—not because the Rule of Law has only instrumental value, but because it is an inherently mixed-value good.
Fuller and Raz on the Rule of Law
There are two particularly influential philosophical accounts of the Rule of Law: those provided by Fuller (1969) and Raz (2009: 210–29). Both take a narrow view of the Rule of Law whilst regarding it as an important legal ideal. What is distinctive about the two accounts is that they suggest two (very different) ways in which the requirements of the Rule of Law might be unified. There is a great deal to be learned from both the accounts themselves and from their limitations.
Fuller’s account of legality is best known for the claim that it constitutes an “internal morality” (or “inner morality”) of law (1969, ch. 2). He points out, plausibly enough, that a legal system could not exist if all of its laws lacked any one of the following eight features (or “desiderata”), which could be described as: (1) generality; (2) publicity; (3) prospectiveness; (4) clarity; (5) non-contradiction with other laws; (6) the possibility of conformity to the law; (7) constancy through time; and, finally, (8) congruence between the announced laws and their administration. These represent the principles of legality that a legal system should aspire to satisfy (1969: 41–44. For two illuminating discussions of the desiderata see Marmor 2007 and Kramer 2007, ch. 2). What makes these principles an “internal morality” is a more elusive matter. Fuller’s key thought seems to be that law is the enterprise of subjecting human conduct to the governance of rules (1969: 46, 49, 162). In order to succeed in this enterprise the eight principles must be respected. In particular, principle (8)—faithful application of the existing law—is necessary if there is to be any point in citizens obeying the law (1969: 209–10). In effect the state is undertaking (a) that it will be possible for the citizen to know and obey the law and (b) that if the citizen obeys the law then the state will abide by it too (1969: 39–40). Consequently, it is an affront to the citizen’s dignity as a responsible agent to depart from the principles (1969: 162).
On the other hand, Fuller recognizes that the principles have only a pro tanto force, i.e., that they can be outweighed by other considerations (including conflicts within the principles themselves, 1969: 45). Thus it can be acceptable to use retroactive legislation to correct a failure of publicity (1969: 53–54). Other illustrations of the pro tanto force of the principles might be the permissibility of some laws bestowing a welcome status on named individuals (e.g., the original membership of a newly formed court) and some cases of selective enforcement of the criminal law. It is also widely thought that the objective standard of care in negligence law is acceptable even though it renders some “short-comers” (i.e., those incapable of satisfying the standard) liable (Honoré 1999). So the importance of Fuller’s principles of legality lies in the value of (a) being able to conform to the law and (b) knowing the consequences of conformity and nonconformity. One limitation of this analysis lies in the fact that these values are always susceptible to being outweighed by some competing value that can be better achieved by departures from the principles. The Rule of Law, on Fuller’s approach, no longer seems to impose quite as strong a normative constraint on the law as it is often thought to do.
The more fundamental limitation of Fuller’s account, however, is simply that it omits so much that is central to our ordinary conception of the Rule of Law, such as the independence of the judiciary, the effective ability of citizens to challenge the legality of state actions in the courts and the observance of due process in civil litigation and criminal prosecutions. Fuller’s principles are quite simply those that are necessary for subjecting human conduct to the governance of rules. But they say nothing about the other conditions that must prevail if the law is to govern a community. (Indeed it may be that Fuller did not envisage his principles of legality as being coextensive with the “Rule of Law,” a term he first uses in the “Reply to Critics” in the revised edition of his work (1969: 187–242).)
So let us turn to Raz’s alternative analysis of the Rule of Law. Raz rejects the idea that the Rule of Law represents an “inner morality” of law (2009: 223–24). He sees it instead as premised on the assumption that people should obey the law and be ruled by it (2009: 212). The “basic intuition” of the Rule of Law is that if the law is to be obeyed it must be capable of guiding the behavior of those to whom it applies (2009: 214). The Rule of Law itself is constituted by those features of the law that are necessary for it to be capable of guiding behavior. This rationale is applicable to Fuller’s desiderata of the Rule of Law, though Raz himself places more emphasis on institutional features that are important for ensuring consistent application of the law and supervising conformity to the Rule of Law itself. He suggests the following (non-exhaustive) list of principles (2009: 214–19):
(1) all laws should be prospective, open and clear;
(2) laws should be relatively stable;
(3) the making of particular laws (particular legal orders) should be guided by open, stable, clear and general rules;
(4) the independence of the judiciary must be guaranteed;
(5) the principles of natural justice must be observed;
(6) the courts should have review powers over the implementation of the other principles;
(7) the courts should be easily accessible; and
(8) the discretion of the crime-preventing agencies should not be allowed to pervert the law.
Raz ultimately sees the Rule of Law as an inherent virtue of the law, because a legal system that better conforms to the Rule of Law is better able to achieve the function of guiding human behavior. He emphasizes, however, that this is not a moral virtue, but rather an instrumental virtue, i.e., the virtue of efficiency (2009: 226). In his well-known image, the specific function of knives is to cut, and thus the sharpness of a knife is one of its inherent virtues, but a knife can be put to both morally desirable and morally deplorable uses. So too with the uses to which the law can be put. A legal system that conforms to the Rule of Law, is a “good” one, but only in the sense of being well-suited to the specific function of law (2009: 223–26).
On the other hand, Raz does attribute other (moral) value to the Rule of Law, such as (a) reducing the scope for arbitrary power, (b) facilitating individual planning and (c) serving the value of human dignity by reducing uncertainty over the future and reducing frustrated or disappointed expectations (2009: 219–23). But he sees the Rule of Law as a purely negative virtue, because its value lies in lowering the risks that are created by the law’s existence in the first place. It is like the duty not to deceive in communicating with others. Absent the ability to communicate we could not deceive. But the value of communication does not lie in failing to deceive others—it lies in the positive goods that can be achieved through communication (2009: 224, 228).
Raz’s account is forceful, but it has three important limitations. The first concerns the way in which it frames the conception of the Rule of Law in terms of the guidance functions of the law. The Rule of Law is normally thought to be particularly relevant to how the state acts, directing it to govern through (and be governed by) law. The reason for this is obvious enough: the state has many possible means at its disposal to alter the behavior of its subjects and bring about various ends. The state can “guide” conduct by (a) intentionally bringing about that conduct, or, more specifically, (b) intentionally bringing about that conduct by laying down standards to be followed. The behavior of the general public can be altered in many ways other than by laying down standards, e.g., by persuasion (health campaigns, propaganda), or by providing services and resources (roads, hospitals, art galleries, public parks), or by changing the circumstances of choice (e.g., by expanding or contracting the money supply, buying and selling assets, deploying more police on the beat). In the case of these sorts of measures, there is no reason to think that they will be more effective in guiding behavior if they are legally authorized than if the state or its officials simply act without authorization.
The Rule of Law, however, does require the state to have legal authorization for the use of its resources and personnel. It is not simply that the state and its officials should obey the general law: there should also be legal controls over how it acts within the general law. This aspect of the Rule of Law is ignored by Raz because he sees the matter in terms of the relationship of the law to the government (2009: 212’13). For Raz, to say that the government should be subject to the law is tautologous, since an action unauthorized by law cannot be an action by the government as a government (2009: 212). The only alternative Raz sees to this “legalistic” conception of government is government in a “political sense,” i.e., government in the sense of where “real power” is located in a community (e.g., with big business, or trade unions, etc.). But it is more common to think that the problem of the Rule of Law is a particularly pressing one for the state, and the “state” in the political sense of a system of institutions which have characteristic social functions and which make characteristic claims about the justification for those functions. The state is not necessarily the location of “real power” within a community: it is instead an entity with a special legal and political status within society (and, very often, considerable social power). It would be misleading to think of unauthorized actions of the state as being of no greater significance than the unlawful actions of other social actors. The state is intimately connected to the law, and there is a heightened concern that it be subject to the law.
Secondly, even on its own terms the instrumental view of the Rule of Law contains a significant inherent limitation, as Raz himself recognizes (2009: 225). Even where the legal standards are used to guide behavior, it is not the case that the ends pursued by the state will be more effectively pursued through conformity to the Rule of Law. There is an important reason for this. The state often pursues some end indirectly through the law: it makes doing X legally required in order to promote or bring about Y (Raz 2009: 224–25; see also 167–68). So the state may require vehicles to be inspected and registered with a central authority. But ordinarily the aim of such measures will not be inspection or registration for its own sake. Inspection and registration are normally desirable because they promote the goal of road safety by ensuring that more vehicles are roadworthy than they would be without these measures. Now whether that indirect goal will be better served by, e.g., a very clear and precise set of criteria (conforming to the desiderata of clarity), or by an extremely vague standard, seems to be a contingent question. Perhaps more car-users will err on the side of higher quality if the standard is extremely vague, leading to greater road safety. Otherwise they will settle for nothing more than barely passing the precise criteria. Equally, perhaps the most effective way of dealing with the phenomenon of “designer drugs” (where the molecular structure of a chemical is modified to escape the definitional scope of current prohibitions, while preserving the effect of the substance) would be to criminalize and punish them retroactively, thereby deterring others but violating the Rule of Law requirement of prospectiveness. Similarly, the widespread official toleration of certain unlawful methods of evidence-gathering (such as maintaining an unlawful database of people’s DNA, or using intrusive surveillance) might be more effective in reducing some types of criminal wrongdoing than the scrupulous use of legal methods.
These examples indicate that it is only the performance of the very conduct specified in the standard that is made easier by conformity to the desiderata of the Rule of Law. The underlying aim of the standard, i.e., the goal that the introduction of the standard was designed to promote, might be better achieved by departures from the Rule of Law. So it is misleading to think that adherence to the Rule of Law is straightforwardly underwritten by considerations of instrumental effectiveness. The use of clear, or prospective, or public, standards is not an inherently superior method of achieving some ends than the use of standards that lack these features, or to the use of some other means altogether. It is only inherently conducive to enabling those to whom the standards are directed to do the very thing that the standard requires: i.e., it is only inherently conducive to enabling the standard itself to be followed. So it all depends on the nature of the goal in question whether the most effective method of pursuing it is via a standard that conforms to the Rule of Law.