The Rule of Law and Law and Order: Between Rechtsstaat and Gesetzesstaat
by Michael Neumann.
I seek a concept of the rule of law between that embodied in Rechtsstaat and that sometimes expressed by the phrase "die Herrschaft des Gesetzes", or the Gesetzesstaat.(1) The former, in its most fulsome, least positivistic incarnation, I will take to represent rule by a law infused with elements of what Hobbes called 'equity', of what North Americans might call 'natural justice', of what political theorists refer to as natural law, and of what Kant called practical reason. The latter I will take to be rule by law in a narrower sense, that which is contained in the codes and statutes of a sovereign state, whether or not that state has any moral or political legitimacy.
Rechtsstaat embodies what might be called a moralized conception of the rule of law, and that English phrase has taken on a moral weight it can scarcely bear. The ringing declarations of the United Nations and the International Commission of Jurists would have us believe that, to achieve the rule of law, we must have not only procedural fairness, not only a formal respect for what are called human rights, but even a social fabric that maintains such rights at their fullest valuation, defended by a government based on free, democratic consent. Such a Rechtsstaat will be found, of course, only in utopian writings.
On the other hand, those who oppose to the Rechtsstaat a Gesetzesstaat almost revel in its amorality, and hasten to sacrifice all on the altar of Realpolitik. They wish to purge the rule of law not only of all moral presuppositions, but of any normative content whatever. For them the law, no matter how evil, is the law. To moralize it, they feel, is charlatanism.
Though both these interpretations of the rule of law are defensible, neither is fruitful. It is not that the notion of a Rechtsstaat necessarily introduces a 'false' neutrality into the notion of the rule of law: one might well suppose, for instance, that genuine equality before the law requires the abolition of economic privilege. But the concept of the Rechtsstaat discovers in the rule of law, if I may parody Kant's first critique, only the morality we ourselves introduce. What is gained by constructing a moralized ideal of the state? Wouldn't it be clearer, and no less powerful, simply to hold that the state should follow certain rules of justice, and to declare that its laws command no respect unless they at least do no outrage to fundamental moral principles?
On the other hand, some positivistic conceptions of the rule of law are, as Rousseau already observed, undermined by their own realism. If, to borrow Austin's famous phrase, the rule of law is mere conformity to the 'command of the Sovereign', what distinguishes it from a sort of codified bullying? When the rule of law is reduced to rule according to whatever happens to be written in lawbooks - however vile, however senseless - then we have a situation in which legal right, at least, is might alone. This, Rousseau says, is vacuous: if the powers that be have might, they don't need that kind of right, and if they lack might, they also lack the right it confers.(2) So this positivistic version of the rule of law boils down to saying that those who can enforce their will, can enforce the written expression of their will - true, but not very illuminating. Achieving a positivistic rule of law is a trivial matter of ensuring that our laws have the formal character of state commands.
I think we can do better than this: we can find a rule of law neither dressed up in numerous layers of moral petticoats, nor stripped bare of anything but power. This project calls for a process of moral reduction, somewhat like what Husserl called phenomenological reduction. Our point of departure will be a moralized, contemporary Anglo-Saxon version of the rule of law, one which prescribes not only basic fairness and justice, but respect for human rights. I will then purge this conception of its moral content, much as the positivists would have us do. But this reduction will leave us with more than the positivist conception of the rule of law. It will lack moral content, yet retain normative content, and the concept will be substantial rather than merely formal. Laws, within this framework need not be good to be laws, but they must exclude certain kinds of evil. Any rule according to any legal codes that fails to do so, will not count as the rule of law. From a philosophical standpoint, this is an interesting development, because it promises constraints on action that accomplish moral purposes without themselves being moral. They are instead, as we shall see, essentially epistemic. We will arrive at, not an epistemic foundation for morality, but at least at a fruitful alliance between morals and epistemology.
As a tool for reducing the rule of law to its barest elements, I will take it as axiomatic that the rule of law must be rule according to law, and that law must at least be what the positivists say it is - the codified statutes of a sovereign executive. This is not at all to adopt the positivist position, because the assumption is not that the formal requirements are sufficient for law, but only that they are necessary. Good rules which are not government statutes never count as laws, but nothing is a law just because it is a government statute. To be a law, it must meet other requirements as well.
What are they? What requirements must the 'commands of the Sovereign' meet in order to provide a basis for rule according to law? They cannot be the moral requirements of the modern Rechtsstaat or the U.N.'s Universal Declaration of Human Rights. Though we shall see that some violations of human rights are indeed incompatible with the rule of law, this is certainly not always the case. For example, any country in Europe in, say, 1910, had numerous statutes which would today count as violations of basic justice and human rights: laws against the poor, laws which deprived women of full citizenship, laws which suppressed basic civil liberties. Yet we would not say that, in these states, there was no rule according to law, or that the rule of law did not prevail. The rule of law is not the rule of good law.
Suppose, then, that at least the more obvious moral content is purged from our concept of the rule of law. We are left with the concepts prevalent in modern Anglo-American political theory as exemplified in the works of John Rawls and a lesser-know but more lucid legal philosopher, Jonathan Raz. Their notion of the rule of law is not entirely formal, but certainly far removed from natural law traditions. They specify that, for the rule of law to prevail, the following sorts of conditions must be met: The laws must be public, general, and non-retroactive; the courts shall observe procedural justice; and the judicial process shall remain insulated from corruption and government interference. As for the laws themselves, they can violate all notions of justice and humanity as long as they do not themselves undermine the judicial process. Racial laws, for instance, would be entirely permissible, though ex post facto laws would not. The only moral content attached to the rule of law concerns what goes on in the courtroom. The defendant can be prosecuted under the most evil statutes, so long as he gets a fair hearing. I will call this the Anglo-American conception of the rule of law.
The Anglo-American conception may seem disappointing in its virtual indifference to substantial injustice, but I will argue that it does not go far enough in that direction: it should be indifferent to procedural injustice as well. We will find that, even after we have stripped the rule of law of this remnant of morality, we have not, after all, reduced it to something purely formal. On the contrary, it has substantial content of its own, and this content brings it closer to a Rechtsstaat than one might expect.
What I am proposing is that the requirement of fair procedure be replaced by an entirely amoral insistence on state control: a certain proportion of those guilty of an offence must be apprehended, tried, and convicted. The proportion must be high enough to assure public order. Must the innocent also go free? Not necessarily. Convicting some of the innocent must be compatible with the rule of law; otherwise it once again becomes an unattainable ideal. On the other hand, too large a number of such convctions means that things are not, by and large, going according to law, and the avoidability condition is violated if I cannot avoid penalties for what I didn't do. But if it is clear that certain sorts of suspicious behaviour will get me convicted, convicting the innocent in such cases is compatible with the rule of law, because I can avoid such behaviour. This holds even if conviction on suspicion is manifestly unfair. The rule of law, as I conceive it, requires only (a) that rates of apprehension and conviction be maintained through the police and courts, and (b) that innocent people can usually avoid conviction. It does not require that any of this happens through fair procedure.
It might be supposed that fair procedure is part of the notion of rule according to law. This is a comforting illusion. Rule according to law means, if it means anything, that things go as the law says they go. If the law says that murderers are to be executed, rule according to law requires that they actually be executed.. This has nothing to do with fair procedure, but with the point of rule according to law, which is that people can in some measure plan their lives by seeing in the penal codes what the State demands of them. Of course, if there are laws against unfair procedure, bribery, perjury, and so on, they must be respected, but institutionalized unfairness is quite compatible with rule according to law. If I know that I will be convicted unless I pay off the judge, if I know that a poor man's testimony counts half as much as a rich man's, I can plan my life accordingly. Indeed, poor people have been doing this for many centuries. On the other hand, fair procedure does not guarantee that I can plan my life. In theory at least, a perfectly fair system of justice could end up, for instance through a series of mistaken beliefs about misleading evidence, convicting only the innocent and letting all the guilty go free. Yet an unfair but generally accurate system which convicts some of the innocent merely serves notice that I must avoid even the suspicion of murdering someone. Unfair but consistent and predictable procedures need not undermine rules like "murder means execution", and are quite compatible with things going according to law.
We have now reduced the rule of law to the bare notion of rule according to law. But rule according to law goes beyond mere enforcement of whatever commands happen to exist in some country's statutes, and it goes beyond mere public order. It contrasts, not only with anarchy, but also with direct compulsion.
Suppose, for example, a ruler did not use laws at all. Suppose his subjects were fitted with wireless receivers through which they were stimulated to perform certain actions, and, perhaps, to think certain thoughts. Then we would not have rule according to law, even if there were statutes about what these people should do. Instead of being given commands or instructions whose fulfilment depends on the understanding of the ruled, people would simply have good or evil inflicted on them without even the opportunity to follow rules. People would be treated, not even like trained animals, but rather like laboratory rats, and the statutes would be, not genuine laws, but something more like a laboratory manual.
Rule according to law is therefore not rule according to whatever happens to be in the lawbooks, but rule according to laws that the ruled can understand as a guide to action. This places two fundamental constraints on legislation: the laws must imposed only avoidable penalties, and they must be formulated in publicly observable terms. The first requirement, avoidability, means that obedience to the law enables me to avoid the penalties of disobedience. "If you murder, you will be imprisoned" meets the avoidability condition because I can avoid being imprisoned by not murdering. "If you are male, you will be imprisoned" violates the avoidability condition because I cannot avoid being male, and therefore cannot avoid being imprisoned.
The avoidability condition follows from the idea of rule according to law. "If you are male, you will be imprisoned" violated the avoidability condition. It doesn't tell me to do anything. It does little more than remind the government of its own intentions, and announce them to others. No attempt is made to guide my behaviour, and the idea of such guidance is the essence of rule according to law.
When we move from laws to the constraints they impose, the avoidability condition changes. The law against murder constrains all citizens from murdering, and the constraint is unavoidable: I encounter it from the moment of my birth, and never escape it. Such unavoidable constraints are entirely compatible with rule according to law as long as they are universal. But laws may also impose non-universal constraints, and rule according to law requires that such non-universal constraints be avoidable. "If you are a butcher, you must clean your shop daily" is an non-universal constraint that meets the avoidability condition, because I can avoid being a butcher. "If you are male, you must clean your shop daily" violates the avoidability condition, because I cannot avoid being male.
This last rule does, of course, guide my behaviour, and does so with respect to any penalties the law may impose. But it does not guide my behaviour with respect to the constraint it imposes. In that sense I am just being manipulated, not instructed; no understanding, no modification of my behaviour can avoid the constraint. This matters partly because there is no real distinction between a constraint and a penalty: what if the law read, "If you are male, do not move your arms and legs"? If we are ruled according to law, our understanding of the law should enable us to avoid penalties masquerading as constraints. Moreover, the guidance provided by the law should be comprehensive in the sense that the weight of government never simply falls on us out of the sky. Special constraints should come only as a consequence of disobeying some rule: otherwise they are just like the restraints with which we secure laboratory animals to test benches. In a society under the rule of law, law rules in the sense that it enables us to avoid, through our understanding of what sort of behaviour is expected of us, any burden not borne by all.
The emphasis on avoidability invites the objection that, for all we know, nothing we do is avoidable, since everything is determined by forces outside our control. To set limits on the avoidability condition, the public observability condition is required. Something is publicly observable only if we can all stand around and look at it: being struck is publicly observable and feeling intimidated is not. The avoidability condition is restricted by the public observability condition in the sense that I am indeed liable for some unavoidable actions: those which are unavoidable due to factors which are not publicly observable. Thus we can be punished for murder because, whatever a murderer's internal compulsions, we can't stand around and look at them. But we cannot be penalized for failing to jump to the moon, because publicly observable features of the physical world make our failure to do so unavoidable.
Public observability does not simply keep the avoidability condition within manageable bounds. It also helps to ensure that people are genuinely ruled according to law. Laws which refer to people's internal states are intrinsically unclear, because we cannot know the effects of our actions on the inner states of those affected.. For example, we cannot be sure when we run afoul of statutes which forbid 'creating a threatening environment' for someone, because we often cannot and do not know when someone will feel threatened. The public observability condition ensures that the laws by which we are ruled will be understandable to everyone, so that our behaviour can be guided by those rules. This is essential to rule according to law.
One might think that the public observability requirement is far too strong, because crucial judicial notions like intent apparently refer to what is unobservable. We do not, however, have to get along without these notions, or any others that refer to human psychology - emotional distress, for instance - as long as we in effect take these terms in a publicly observable sense. "In effect" means that the words do not actually have to be defined in publicly observable terms, so long as judicial decisions can be understood as taking the words in that way. As long as, for instance, attribution of intent depends entirely on the publicly observable actions of the accused, or emotional distress is measured in the actions and the physical symptoms of the victim, the public observability requirement is met.
These are the broad outlines of the conception of the rule of law that I propose. It corresponds closely to what is ordinarily termed 'law and order'. Its affinities with the Gesetzesstaat will be apparent: the emphasis on states with the power to enforce the laws, and the indifference to such moral niceties as political legitimacy and natural rights. And even the Gesetzestaat, which at least admits procedural fairness, may seem closer to the Rechtstaat than this idea of the rule of law as mere law and order.
Despite appearances, the proposed conception genuinely occupies a middle ground between the Rechtsstaat and the Gesetzesstaat. Its non-moral notion of law differs sharply from the natural-law survivals found in the Rechtsstaat. But it also differs sharply from the formal, positivistic conceptions associated with the Gesetzestaat. Though it lacks moral content, it has morally important implications, and provides more protection against certain moral outrages than does morality itself.
To see this, note that the avoidability condition itself is substantial, not formal, and therefore not like the formal 'generality' requirement found in many accounts of the rule of law. "Everyone who grows over two meters shall be shot" has as much formal generality as "Everyone who murders shall be imprisoned", but the avoidability requirement (as constrained by the public observability requirement) forbids the former while permitting the latter. The avoidability requirement looks at the content of a law to see whether citizens or subjects really can avoid the penalty, not at its logical form.
One might add that modifying the generality requirement is not a promising strategy for extracting something of normative value from the notion of the rule of law. It is not just that logical generality is far too permissive to guarantee anything. It is also that attempts to find guarantees of freedom or equality in some appeal to 'substantial' generality will stretch that notion past its breaking-point. The idea of substantial generality only tells us that some aspects of our real position in society should not put us under some disadvantage vis-à-vis the legal system. It does not tell us which factors count. There is nothing to indicate whether the law should weigh equally on us with respect to need, to merit, to beauty, to physical strength, to achievement, or, for that matter, with respect to whether or not we have committed murder. A system that equalizes political and economic power relations between factory owners and workers has no more or less substantial generality than one which does so for children and adults, supermodels and ugly mathematicians, or, for that matter, murderers and non-murderers. Which of these substantially general systems is preferable is a moral decision that has nothing to do with the notion of substantial generality itself, nor with any notion of substantial rationality underlying it.
The substantial dimension of the avoidability condition does, however, have normative implications: it constrains the sorts of things people may do through legislation. Not only does it do so in ways that converge with the requirements of morality; it does so better than the principles of contemporary moral systems. It does not rely, like the natural law morality of the Rechtsstaat, on rights which appear out of nowhere, nor does it make a risky appeal to consequences, like utilitarianism or many theories of justice. Consider, for instance, a racist law, like "Blacks cannot use public beaches". To say that this violates rights is both intellectually suspect and politically dubious: the rights are of mysterious origin, and they invite claims of 'conflicting rights' which can take many decades to 'weigh'. To forbid such laws because of their bad consequences is downright dangerous: In the long run and all things considered, even the very worst laws might have good consequences. Who is to say that Naziism itself, by alerting Western democracies to the horrors of genocide, mightn't eventually turn out to have good consequences overall?
The rule of law takes a simpler and more reliable approach: such legislation is always forbidden because it is directed at an unavoidable class of persons, i.e., at people who can't avoid being what they are and what draws on them the wrath of the state. Racial laws, whether they impose penalties or constraints, are forbidden because you can't avoid being of the race you are. The same will be true of laws which aim at sex, age, or any other unavoidable characteristics.(3) These sorts of discriminatory statutes appropriately incur special condemnation under the rule of law. They contrast with the denial of civil liberties - the suppression of free speech, the practice of religion, and so on - which the rule of law does not forbid.
Though the avoidability condition has morally important implications, they are not, strictly speaking, moral implications: they do not imply that the forbidden legislation is morally wrong. Racial 'laws' are illegitimate, not because they contravene some right or moral principle, but because they do not mesh with rule according to law. If I constrain you according to your race, it is not like when I constrain you because you have committed a crime: there is no instruction you could have followed to avoid the constraint. You are, in effect, being pushed around, not given orders: any so-called law constraining only members of a certain race is merely an unavoidable punishment masquerading as an order. "If you are black, do not use this beach" is like "If you are male, do not use your arms", not like "if you are a butcher, wash your hands before handling meat." Not only does the rule of law support the intuitions of morality, it provides that support from an unexpected quarter. Racial 'laws' are outlawed simply because they are not the sort of rules that can play a role in rule according to law.(4)
If the avoidability condition provides unexpected support for morality, the public observability condition provides unexpected constraints. The rule of law forbids legislation which might be morally right, but which involves the imposition of penalties or constraints for acts whose relevant characteristics are not publicly observable. This has interesting implications for issues such as the censorship of obscene material. The rule of law appears to be quite comfortable with such censorship, but requires a definition of obscenity so narrow as to be virtually useless. It allows us to forbid the use of a certain specified words, or the depiction of certain body parts, but it forbids us to do what so many 'reasonable' censorship laws attempt: outlawing material which 'tends to corrupt minors', or 'offends community standards', or, in other contexts, 'creates a hostile environment'. We can use such language if we like - once again, the rule of law does not indulge in linguistic formalism. But if we use such language, we must define our terms so that everyone can observe exactly when those terms apply to purportedly obscene material. This means, once again, that we will almost certainly have to resort of lists of words or catalogues of images, or at least judge cases just as if we had such a list. In other words, the rule of law all but cripples attempts at censorship by placing the most severe constraints on how censorship laws are to be framed.
Here again the source of the constraints is more interesting than the constraints themselves. Rights of free expression, or any other rights, have nothing to do with the matter. The constraints follow entirely from the requirement that laws be interpreted so that obedience and disobedience can be publicly observed. And this in turn follows from the idea of rule according to law, which requires that the rules be understood by everyone.
The effect of these restrictions is to limit legislation to the public sphere. Clearly morality might command otherwise: there is no a priori ground for supposing that only legislation interpreted in publicly observable terms can do good. Nevertheless constraints on the legislation of 'private' morality may be welcome even from a moral standpoint. When the enforcement of legislation depends on the correct assessment of such inner states as feeling threatened, feeling offended, or feeling frightened, how can we tell when we will find ourselves accused? Certainly such legislation is often repressive precisely because its application is almost necessarily capricious: think, for instance, of Argentinian laws about 'insulting the honour of the navy'. This arbitrariness is incompatible with the very presuppositions of moralizing legislation, namely that it is desirable to plan or engineer what happens in the private sphere. Here again, the rule of law shows itself to be substantial yet amoral: it has consequences, perhaps even good consequences, for matters of moral importance, but its constraints are not generated from any moral assumptions.
We are now in a position to compare the above conception of the rule of law with the Gesetzesstaat and the Rechtsstaat, and to see the advantages of an alternative to these traditional approaches. The relevant points are sketched below.
In the first place, contemporary views of the rule of law lurch between extremes of idealism and cynicism. The Gesetzesstaat poses no ideals at all. Inexplicably, it requires morality only of its own procedures. The Rechtsstaat poses wonderful ideals, but they are unattainable: no one seriously expects a state to exhibit perfect respect for human rights. Nor do idealistic conceptions offer any theoretical muscle to compensate for their lack of practical force. Why should we respect the ideals so authoritatively proclaimed? What happens when they conflict? Idealist conceptions of the rule of law would be welcome were they well-founded and coherent, but in most cases they are little more than an impotent expression of good wishes. Their popularity rests primarily on the implausible expectation that having a very moral concept of something will somehow encourage the instantiation of that concept in the world The rule of law as law and order is an attractive compromise between the Gesetzesstaat and the Rechtsstaat because it neither abandons norms altogether, nor sets them beyond our reach.
Second, the rule of law as law and order supports morality without pulling moral obligations out of a hat. It does not hide in the formalism of the Gesetzesstaat. Its normative implications derive from a substantive conception of what it is to live according to rules, not from attempts to breathe moral life into very abstract conceptions like generality. Unlike the Rechtsstaat, it is indeed compatible with horrendous moral outrages. But it is not as if the Rechtsstaat is any more capable of preventing these outrages than any other ideal or conception. The rule of law as law and order at least finds support for morality somewhere else than in morality itself: that is, in the epistemic standards that must be met if people are to regulate their behaviour through their understanding of the state's commands.
Finally, the rule of law as law and order embodies the non-moral yet important value of civilization. If civilized behaviour is behaviour that follows certain rules, and operates through understanding rather than brute force, the rule of law is civilized in a manner that has little to do with 'high culture'. Instead it exemplifies the preference for treating people, if not as Rational Beings, at least as reasoning ones.
The value of the rule of law is that it articulates a practicable ideal - something desirable and readily attainable. No more can be claimed for it. It protects no one; it does nothing to advance social justice; it manifests not even the most abstract concern for fairness. It does identify, however, practices that support and enhance some widely held moral ideals. Little more can be expected of a mere political idea.
NOTES
1. see Gottfried Dietze, Two Concepts of the Rule of Law, Indianapolis (The Liberty Fund) 1973.
2. Rousseau, Contrat Social, I.3.
3. In some cases laws could be reformulated or interpreted to avoid excessive consequences. We might do away with age requirements for driver's licences, and stick with tests. We might also suppose that age requirements are not really what they seem to be, but a kind of shorthand for "belonging to a group which will not pass the tests".
4. One would do well to restrain oneself from making Kantian comparisons here. Though both the rule of law and Kantian morality have in common their interest in the very idea of a rule, they share little else. Kantian morality holds that conformity to the idea of a rule for all rational beings is the necessary and sufficient condition for the adequacy of a moral principle. The rule of law holds that conformity to its constraints is a necessary and sufficient condition of something's being a law, but claims no relation to morality. Moreover, the rule of law does not, of course, require that its principles be such as can be accepted by all rational beings. It requires at most that its principles can be understood by all such beings.