Let us assume that politicians want to be in office and to maximize their autonomy in decision making. On the other side, citizens want to avoid abuses by politicians. Citizens have two instruments to protect them: first, to throw the rulers out of office at election time; second, to enforce, through institutions, legal limits to the political discretion of incumbents between elections. The first protection is provided by democracy; the second, by the rule of law. Prima facie they complement each other. Citizens are not just interested in electing politicians who, once in office, are controlled only by the prospect of future elections; they are not interested either in unelected, nonrepresentative rulers, even if bounded by laws passed by an undemocratic assembly.

I use here a minimalist definition of the rule of law. It consists of the enforcement of laws that have been publicly promulgated and passed in a preestablished manner; are prospective (nulla poena sine lege), general (like cases are treated alike), stable, clear, and hierarchically ordered (the more particular norms conform to the more general ones); and are applied to particular cases by courts independent from the political rulers and open to all, whose decisions respond to procedural requirements, and that establish guilt through the ordinary trial process. This definition makes no reference to fundamental rights, democracy, equality, or justice: it corresponds to what Dworkin (1985: 9–32) has termed the “rulebook” conception of the rule of law.

The nature of the law – in the sense of both customary human practices and positive, written law – is a central theme in Montaigne's Essais, though, like any other major subject in his work, it is difficult to present systematically without imposing a somewhat arbitrary order on the text. There were basically three dimensions in which Montaigne developed this subject: the first one was the broad anthropological reflection on the nature of social norms and moral conventions within human societies, no doubt the best-known aspect of his contribution to this issue and one that is generally regarded as distinctly representative of his skeptical approach. The second dimension was the devastating critique of contemporary French legislation and of the judicial machinery responsible for administering it – a reality Montaigne was intimately associated with in his capacity of conseiller first, between 1554 and 1557, at the Cour des Aides de Périgueux, and then, from 1557, at the Parlement of Bordeaux until his decision to sell his post in 1570. Finally, the third dimension was the evaluation of the impact of the Reformation and of religious conflict upon the French legal order, a retrospective assessment that Montaigne attempted at various points in the Essais but which was more fully developed in the 1588 edition of the work (the one that contains book III), published when the prospect of peace and of a durable settlement with the Protestants under Henry of Navarre was finally in sight.

Maravall and Przeworski open this book with a difficult question: why do governments act according to laws? The fact that so many governments, both contemporary and historical, have difficulty doing just this indicates that the answer is not obvious. The principal argument of this book is that the force of law is not normative – citizens and political officials do not obey law because of a duty to obey law. Instead, political officials obey the law because they have incentives to do so.

Maravall and Przeworski fill out this logic. For example, they suggest that the constitution is important for the rule of law. “But the constitution matters not because governments feel a duty to obey it. Rather, it serves as a focal device, enabling particular individuals to guess what others will consider as major transgressions and thus to agree when to act.” To police the behavior of government officials, “Actions of groups with different interests must be coordinated.”

Maravall and Przeworski emphasize another important aspect of the rule of law, that “laws inform people what to expect of others…. At the same time, [laws] facilitate coordination of sanctions against a government that deviates from its own announcements. In this sense, publicly promulgated rules provide an equilibrium manual.” They note that “laws indicate to citizens when to act against the government. By coordinating expectations, they facilitate collection actions that impose sanctions on governments.”

In short, one answer to Maravall and Przeworski's question is that political officials obey the law because not doing so puts their political future at risk.

Should we associate the rule of law only with democratic legal systems or can we conceive of the rule of laws as an independent phenomenon that may equally be associated with other forms of regime? In particular, can we speak of an autocratic or dictatorial rule of law? In this chapter, I discuss two notions of the rule of law and argue that in principle, under specific conditions, both are compatible with nondemocratic forms of rule. Although this association may not be historically all that common, I analyze one case, the military dictatorship in Chile (1973–90), and try to show that a form of rule of law was operative within the regime, particularly during the last nine years of military rule. In developing this argument, I hope to elucidate some general properties of the rule of law and specify conditions under which rules can have force even upon their own makers.

Two Notions of the Rule of Law

The term “rule of law” is used quite widely in contemporary theoretical and political discussions. Nevertheless two broad conceptions are prominent. One, variously referred to as a “narrow,” “formal,” or “instrumental” conception of the rule of law, examines the formal characteristics that law must have if a legal system is to provide a nonarbitrary framework around which subjects can form expectations and live their lives. This notion of the rule of law essentially concerns the character of law as a mechanism of mediation between state authorities and social actors.

Rule of law and democracy are both desirable attributes of a political system. Scholars writing of democratic transitions from authoritarian rule usually argue that the goal of such a transition is the establishment of democracy with the rule of law, implying that both may be achieved simultaneously. Perhaps that is so. What is often meant by rule of law is no more than the notion that government should work its will through general legislation, legislation to which the governors themselves are subject, rather than through irregular decrees and ad hominem proclamations. But rule of law may require more than this: it may require that people are able to foresee accurately the legal consequences of their actions and not be subject to sudden surprises whether or not these take the form of legistation, or perhaps that the law contain, or at least not violate, certain substantive principles and rights.

Democratic rule minimally requires government by the people or their representatives, elected on a broad franchise. But, in some conceptions, it too many require more than that. Perhaps, democracy demands that the range of choice open to government be broad and not constricted by externally imposed restraints (such as legal protections for minorities). We expect, for example, or hope, that our government can correct inequities arising from markets or social interactions. Such interventions can involve confiscatory taxes or draconian regulations, either of which can threaten claims for minority rights. Or perhaps, democracy requires that the people be regularly and genuinely consulted on fundamental legal changes so that institutions or practices of deliberation and consultation are in place and functioning.

If democracy is to exist, at least one rule must be observed, namely the rule that specifies which of the political parties should occupy the office of government. My purpose is to investigate under what conditions political parties competing in elections obey their results.

According to one view, people obey laws when they share a particular kind of culture. This culture may value the rule of law per se, regardless of the outcomes it generates. It may impose on people the duty to obey outcomes resulting from rules to which they agreed. Or it may foster temperamental characteristics predisposing them to obey laws. But whatever the specific features, rule of law can be sustained if and only if a society is characterized by a particular culture.

An alternative theory sees the emergence of a rule of laws as an outcome of conflicts of interests. While situations in which everyone acts according to law can be described in cultural terms, this theory claims that such situations arise when the conflicting political forces find it in their best interest to act in conformity with some laws, given what everyone else is doing.

The controversy concerns the causal mechanisms that generate these situations: are they an expression of some antecedent cultural patterns or are they an effect of pursuit of interests? Both views arrive at the same conclusion – namely, that there are some situations in which political actors act in conformity with laws.

In this chapter I challenge one common view of the “rule of law” and, particularly, of what institutional arrangements put the rule of law at risk. According to this perspective, the rule of law is closely connected with (what I call) liberal political systems (which, synthetically, are characterized by a system of checks and balances and entrenched individual rights protected by an independent judiciary). In addition, many among those who defend this view tend to evaluate all moves towards a more majoritarian democracy as threats to the rule of law. Normally, they assume that a majoritarian democracy necessarily results in arbitrary government. This arbitrariness would be the result of basically two features that they associate with the majoritarian democracy: its tendency to produce “hasty” decisions and its inherent incapacity to establish adequate institutional controls, that is, to establish controls over the will of the majority. Majoritarian democracy, thus, is directly associated with an “unchecked majority,” that is, a “populist” regime – in the end, the breakdown of the rule of law.

In what follows, I try to show that it is possible to defend a more majoritarian political system without renouncing, at the same time, the ideal of the rule of law. In this sense, I affirm that there is ample space between the liberal system of government and a “populist” system – enough so to make it possible to defend a reasonable majoritarian system. In the first section of my work, I show how we began to correlate the notion of a majoritarian system with violations of the rule of law.

General Pinochet warned in October 1989, a few weeks before Chile's first democratic elections after the 1988 referendum, that “if someone touches one of my men, the rule of law is over” (si me tocan a uno de mis hombres, se acabó el Estado de Derecho). There seems to be something profoundly paradoxical about the general's subtle warning. It implies that the existence of the rule of law depends on the will of a single person, but part of the meaning of the rule of law is precisely that the institutional order is something other than the product of a single will.

This is usually misunderstood by those who discuss the rule of law. They often affirm that the point is to institute “a government of laws, not of men.” Yet this statement is at best ambiguous. A government cannot consist of laws. A government of laws can only mean that the rulers are bound by what the law establishes, that is, that a government of men complies with the laws. The underlying confusion is also apparent in other, equally misleading phrases that people link to the rule of law, such as “the sovereignty of the law” or “the supremacy of the law.” All this is empty rhetoric. The law, being a human creation, must necessarily be subject to human will. In fact, the very term “the rule of law” is in itself rhetorical. The law cannot rule. Ruling is an activity, and laws cannot act.

The problem of obedience may look strange to a legal scholar. Lawyers deal not with obedience but with obligation. The question they ask is not, What behavior does effectively take place? but What behavior ought to take place? According to leading theories, actual behavior has no effect on the validity of a rule. In other words, a rule is binding or not binding independently of the fact that it is being obeyed (or disobeyed). This is just another way of expressing the difference between is and ought and the law deals only with what ought to be, not with what actually is or will be. Moreover, it is generally agreed that there can be no causal relation between what is and what will be. The fact that something ought to happen does not cause it to happen and the fact that something actually happens has no influence on its being mandatory or forbidden. In other words, the validity of a rule does not depend on its efficacy. Indeed it can be said that the specificity of a rule, as distinguished from a law of nature, lies in its capacity to be violated.

Lawyers and legal scholars therefore leave to sociologists and psychologists the question why men obey the law. True, those who make the rules must have some idea of what makes men obey. If they make and publish those rules, they must assume that knowing the existence of a rule will have some sort of psychological effect on actual behavior and they draft them according to the behavior they expect.

Let us begin with a definition that is widely agreed upon in the literature: we are in the presence of the “rule of law” when the rules defining permitted and forbidden actions are not discretionary decisions of an individual, but rather take the form of laws that discipline every citizen, regardless of his or her power or status. From this starting definition, it can be inferred that the “rule of law” faces an important political problem. This problem is not related to the difficulties rulers have in imposing and enforcing rules upon individual or collective actors. Indeed, with different degrees of efficiency, rulers have a varied and diverse repertoire of instruments they can use to make subjects obey the law (force, incentives, socialization, etc.). Rather, the central problem with the “rule of law” derives from the difficulties subjects have in making those who rule obey the law. Why should we expect that those who rule – that is, those who control the state resources – will obey the law? Why shouldn't we expect that they would use these resources to disobey the law? Therefore, an analysis of the rule of law implies answering what determines rulers' obedience to the law.

Authors such as Weingast, Przeworski, and Hardin have asserted that the rule of law will rule only if rulers and subjects conclude that it is in their interest to obey the law. That is, the rule of law will be successful if it becomes self-enforcing.

This chapter elaborates a highly stylized and simplified account of the emergence of two features of the rule of law as commonly understood: predictability and equality. Legal historians would stress the role of economic, demographic, technological, scientific, religious, and cultural factors in bringing about and stabilizing institutional innovations as startlingly novel as legal certainty and equality before the law. When describing the role of important social actors in promoting or inhibiting such developments, they would weave into their story a variety of factors, including ideology, irrational passions, improvisation within inherited institutions, and the unexpected consequences of habitual behavior in a changed setting. My objective, in what follows, is both more modest and more theoretical.

I aim to clarify the reasons why powerful political actors might furiously resist or warmly embrace the rule of law. We cannot explain why the rule of law does or does not emerge in a specific historical context by invoking nothing but the strategic calculations of powerful political actors. But the self-interested reasons why powerful members of a society might encourage or discourage such a development are undoubtedly relevant and deserve a focused treatment.

I ask, first, why governments, with the means of repression in their hands, might be induced to make their own behavior predictable. For help in answering this question, I turn to Machiavelli. His thesis, essentially, is that governments are driven to make their own behavior predictable for the sake of cooperation.

Today, the traditional view according to which democracy implies majority rule in the form of parliamentary supremacy has come under growing criticism. In Europe since World War II, democratic regimes have increasingly incorporated substantive constraints to what the parliamentary majority can do. Not only must public authority be exerted within general rules, but citizens are deemed to be entitled to fundamental rights, whose exercise must remain outside the will of the majority. Therefore, submitting the performance of public functions to the scrutiny of independent judges becomes an effective and essential check on the exercise of political power, ensures the supremacy of the law, and guarantees citizens' rights (Stone 2000).

There is some ambiguity in the concept of judicial independence (Russell 2001). On one hand, judicial independence is understood as institutional independence, that is, as the guarantees judges enjoy vis-à-vis the political branches of government. On the other, the term refers to the behavior of the judges, that is, to their independence on the bench. However, if, as a rule, in order to behave independently, a judge needs to be independent from the parties at the case (and, therefore, also from the executive), it does not follow that institutionally independent judges will automatically behave in an independent way, a point to which we return later.

Historically, in democratic countries, the level of judicial independence as well as the role played by courts in the political system has varied.

In a society subject to the rule of law, a state of emergency proclaimed under existing law enables the government to resort to measures of an exceptional and temporary nature in order to protect the essential fabric of that society. International Covenant on Civil and Political Rights (ICCPR) 4, European Convention on Human Rights (ECHR) 15 and American Convention on Human Rights (ACHR) 27 specify the circumstances under which, in a state of emergency, a state may derogate from its obligations under the relevant instrument, the conditions under which measures derogating from its obligations may be taken, and the notification that is required to be submitted thereon.

When ICCPR 4 was being drafted, it was argued that the eventualities for which it was proposed and the rights to which it might apply were sufficiently covered by the relevant limitation clauses. For instance, it was thought that the concept of ‘national security’ or of ‘public order’, already included in a number of articles of the covenant could be invoked to deal with situations which might arise in time of war or national emergency. In reply it was contended that in time of war, for example, states could not be strictly bound by obligations assumed under a convention unless the convention contained provisions to the contrary. There might also be instances of extraordinary peril or crisis, not necessarily in time of war, when derogation from obligations assumed under a convention would become essential for the safety of the people and the existence of the nation.